experiment: zync: an rsync-like tool for ZFS datasets (re-uses zrepl abstractions)

Local sync and sync over SSH seems to work. The sshdirect transport might be interesting / re-usable for zrepl as well, although we'd still need some kind of locking in that case.
docs: fix typos
2021-01-14 01:33:24 +01:00 · 2020-12-17 12:00:29 +01:00 · 2020-11-11 16:06:54 +01:00 · 2020-11-04 17:53:54 +01:00 · 2020-11-01 16:10:48 +01:00 · 2020-11-01 14:19:03 +01:00
123 changed files with 4762 additions and 1557 deletions
@@ -1,74 +1,79 @@
-version: 2.0
-workflows:
-  version: 2
-  build:
-    jobs:
-      - build-1.11
-      - build-1.12
-      - build-1.13
-      - build-1.14
-      - build-latest
-      - test-build-in-docker
-jobs:
+version: 2.1

-  # build-latest serves as the template
-  # we use YAML anchors & aliases to exchange the docker image (and hence Go version used for the build)
-  build-latest: &build-latest
-    description: Builds zrepl
-    parameters:
-      image:
-        description: "the docker image that the job should use"
-        type: string
-    docker:
-      - image: circleci/golang:latest
-    environment:
-      # required by lazy.sh
-      TERM: xterm
-    working_directory: /go/src/github.com/zrepl/zrepl
+commands:
+  setup-home-local-bin:
    steps:
      - run:
-          name: Setup environment variables
+          shell: /bin/bash -euo pipefail
          command: |
-            # used by pip (for docs)
-            echo 'export PATH="$HOME/.local/bin:$PATH"' >> $BASH_ENV
-            # we use modules
-            echo 'export GO111MODULE=on' >> $BASH_ENV
+            mkdir -p "$HOME/.local/bin"
+            line='export PATH="$HOME/.local/bin:$PATH"'
+            if grep "$line" $BASH_ENV >/dev/null; then
+              echo "$line" >> $BASH_ENV
+            fi

+  invoke-lazy-sh:
+    parameters:
+      subcommand:
+        type: string
+    steps:
+      - run:
+          environment:
+            TERM: xterm
+          command: ./lazy.sh <<parameters.subcommand>>
+
+  apt-update-and-install-common-deps:
+    steps:
+      - run: sudo apt update && sudo apt install gawk make
+
+  restore-cache-gomod:
+    steps:
      - restore_cache:
-          keys:
-            - source
-            - protobuf
-
-      - checkout
-
+          key: go-mod-v4-{{ checksum "go.sum" }}
+  save-cache-gomod:
+    steps:
      - save_cache:
-          key: source
+          key: go-mod-v4-{{ checksum "go.sum" }}
          paths:
-            - ".git"
-      
-      # install deps
-      - run: wget https://github.com/protocolbuffers/protobuf/releases/download/v3.6.1/protoc-3.6.1-linux-x86_64.zip
-      - run: echo "6003de742ea3fcf703cfec1cd4a3380fd143081a2eb0e559065563496af27807  protoc-3.6.1-linux-x86_64.zip" | sha256sum -c
-      - run: sudo unzip -d /usr protoc-3.6.1-linux-x86_64.zip
+            - "/go/pkg/mod"
+
+  install-godep:
+    steps:
+      - apt-update-and-install-common-deps
+      - invoke-lazy-sh:
+          subcommand: godep
+
+  install-docdep:
+    steps:
+      - apt-update-and-install-common-deps
+      - run: sudo apt install python3 python3-pip libgirepository1.0-dev
+      - invoke-lazy-sh:
+          subcommand: docdep
+
+  download-and-install-minio-client:
+    steps:
+      - setup-home-local-bin
+      - restore_cache:
+          key: minio-client-v2
+      - run:
+          shell: /bin/bash -eo pipefail
+          command: |
+            if which mc; then exit 0; fi
+            sudo curl -sSL https://dl.min.io/client/mc/release/linux-amd64/archive/mc.RELEASE.2020-08-20T00-23-01Z \
+              -o "$HOME/.local/bin/mc"
+            sudo chmod +x "$HOME/.local/bin/mc"
      - save_cache:
-          key: protobuf
+          key: minio-client-v2
          paths:
-            - "/usr/include/google/protobuf"
-            
-      - run: sudo apt update && sudo apt install python3 python3-pip libgirepository1.0-dev gawk
-      - run: ./lazy.sh devsetup
-
-      - run: make zrepl-bin
-      - run: make vet
-      - run: make lint
-      - run: make release
-      - run: make test-go
-      # cannot run test-platform because circle-ci runs in linux containers
-
-      - store_artifacts:
-          path: ./artifacts/release
-          when: always
+            - "$HOME/.local/bin/mc"

+  upload-minio:
+    parameters:
+      src:
+        type: string
+      dst:
+        type: string
+    steps:
      - run:
          shell: /bin/bash -eo pipefail
          when: always
@@ -78,78 +83,296 @@ jobs:
              exit 0
            fi
            set -u # from now on
-            
-            # Download and install minio
-            curl -sSL https://dl.minio.io/client/mc/release/linux-amd64/mc -o ${GOPATH}/bin/mc
-            chmod +x ${GOPATH}/bin/mc
-            mc config host add --api s3v4 zrepl-minio https://minio.cschwarz.com ${MINIO_ACCESS_KEY} ${MINIO_SECRET_KEY}
-            
-            # Upload artifacts
-            echo "$CIRCLE_BUILD_URL" > ./artifacts/release/cirlceci_build_url
-            mc cp -r artifacts/release "zrepl-minio/zrepl-ci-artifacts/${CIRCLE_SHA1}/${CIRCLE_JOB}/"

+            mc config host add --api s3v4 zrepl-minio https://minio.cschwarz.com ${MINIO_ACCESS_KEY} ${MINIO_SECRET_KEY}
+
+            # keep in sync with set-github-minio-status
+            jobprefix=zrepl-ci-artifacts/${CIRCLE_SHA1}-pipeline-<<pipeline.number>>/${CIRCLE_JOB}
+
+            # Upload artifacts
+            mkdir -p ./artifacts
+            mc cp -r <<parameters.src>> "zrepl-minio/$jobprefix/<<parameters.dst>>"
+
+  set-github-minio-status:
+    parameters:
+      context:
+        type: string
+      description:
+        type: string
+      minio-dst:
+        type: string
+    steps:
+      - run:
+          shell: /bin/bash -eo pipefail
+          command: |
+            if [ -n "$CIRCLE_PR_NUMBER" ]; then # CIRCLE_PR_NUMBER is guaranteed to be only present in forked PRs (external)
+              echo "Forked PR detected. Sry, can't trust you with credentials to external artifact store, use CircleCI's instead."
+              exit 0
+            fi
+            set -u # from now on
+
+            # keep in sync with with upload-minio command
+            jobprefix=zrepl-ci-artifacts/${CIRCLE_SHA1}-pipeline-<<pipeline.number>>/${CIRCLE_JOB}
            # Push Artifact Link to GitHub
            REPO="zrepl/zrepl"
            COMMIT="${CIRCLE_SHA1}"
            JOB_NAME="${CIRCLE_JOB}"
+            CONTEXT="<<parameters.context>>"
+            DESCRIPTION="<<parameters.description>>"
+            TARGETURL=https://minio.cschwarz.com/minio/"$jobprefix"/"<<parameters.minio-dst>>"
            curl "https://api.github.com/repos/$REPO/statuses/$COMMIT" \
              -H "Content-Type: application/json" \
              -H "Authorization: token $GITHUB_COMMIT_STATUS_TOKEN" \
              -X POST \
-              -d '{"context":"zrepl/publish-ci-artifacts", "state": "success", "description":"CI Build Artifacts for '"$JOB_NAME"'", "target_url":"https://minio.cschwarz.com/minio/zrepl-ci-artifacts/'"$COMMIT"'/"}'
+              -d '{"context":"'"$CONTEXT"'", "state": "success", "description":"'"$DESCRIPTION"'", "target_url":"'"$TARGETURL"'"}'

-      - run:
-          shell: /bin/bash -euo pipefail
-          command: |
-            # Trigger Debian Package Build
-            curl -v -X POST https://api.github.com/repos/zrepl/debian-binary-packaging/dispatches \
-              -H 'Accept: application/vnd.github.v3+json' \
-              -H "Authorization: token $ZREPL_DEBIAN_BINARYPACKAGIN_TRIGGER_BUILD_GITHUB_TOKEN" \
-              --data '{"event_type": "push", "client_payload": { "zrepl_main_repo_commit": "'"$CIRCLE_SHA1"'", "go_version": "'"${CIRCLE_JOB##build-}"'" }}'

-  build-1.11:
-    <<: *build-latest
-    docker:
-    - image: circleci/golang:1.11
-
-  build-1.12:
-    <<: *build-latest
-    docker:
-    - image: circleci/golang:1.12
-
-  build-1.13:
-    <<: *build-latest
-    docker:
-    - image: circleci/golang:1.13
-
-  build-1.14:
-    <<: *build-latest
-    docker:
-    - image: circleci/golang:1.14
-
-  # this job tries to mimic the build-in-docker instructions
-  # given in docs/installation.rst
-  #
-  # However, CircleCI doesn't support volume mounts, so we have to copy
-  # the source into the build-container by modifying the Dockerfile
-  test-build-in-docker:
-    description: Check that build-in-docker works
-    docker:
-      - image: circleci/golang:latest
-    environment:
-    working_directory: /go/src/github.com/zrepl/zrepl
+  trigger-pipeline:
+    parameters:
+      body_no_shell_subst:
+        type: string
    steps:
-    - checkout
-    - setup_remote_docker
-    - run:
-        name: (hacky) circleci doesn't allow volume mounts, so copy src to container
-        command: echo "ADD . /src" >> build.Dockerfile
-    - run:
-        name: (hacky) commit modified Dockerfile to avoid failing git clean check in Makefile
-        command: git -c user.name='circleci' -c user.email='circleci@localhost' commit -m 'CIRCLECI modified Dockerfile with zrepl src' --author 'autoauthor <circleci@localhost>' -- build.Dockerfile
-    - run:
-        name: build the build image (build deps)
-        command: docker build -t zrepl_build -f build.Dockerfile .
-    - run:
-        name: try compiling
-        command: docker run -it zrepl_build make release
+      - run: |
+          curl -X POST https://circleci.com/api/v2/project/github/zrepl/zrepl/pipeline \
+            -H 'Content-Type: application/json' \
+            -H 'Accept: application/json' \
+            -H "Circle-Token: $ZREPL_BOT_CIRCLE_TOKEN" \
+            --data '<<parameters.body_no_shell_subst>>'
+
+parameters:
+  do_ci:
+    type: boolean
+    default: true
+
+  do_release:
+    type: boolean
+    default: false
+
+  release_docker_baseimage_tag:
+    type: string
+    default: "1.15"
+
+workflows:
+  version: 2
+
+  ci:
+    when: << pipeline.parameters.do_ci >>
+    jobs:
+      - quickcheck-docs
+      - quickcheck-go: &quickcheck-go-smoketest
+          name: quickcheck-go-amd64-linux-1.15
+          goversion: &latest-go-release "1.15"
+          goos: linux
+          goarch: amd64
+      - test-go-on-latest-go-release:
+          goversion: *latest-go-release
+      - quickcheck-go:
+          requires:
+            - quickcheck-go-amd64-linux-1.15 #quickcheck-go-smoketest.name
+          matrix: &quickcheck-go-matrix
+            alias: quickcheck-go-matrix
+            parameters:
+              goversion: [*latest-go-release, "1.11"]
+              goos: ["linux", "freebsd"]
+              goarch: ["amd64", "arm64"]
+            exclude:
+              # don't re-do quickcheck-go-smoketest
+              - goversion: *latest-go-release
+                goos: linux
+                goarch: amd64
+              # not supported by Go 1.11
+              - goversion: "1.11"
+                goos: freebsd
+                goarch: arm64
+
+  release:
+    when: << pipeline.parameters.do_release >>
+    jobs:
+      - release-build
+      - release-deb:
+          requires:
+            - release-build
+      - release-rpm:
+          requires:
+            - release-build
+      - release-upload:
+          requires:
+            - release-build
+            - release-deb
+            - release-rpm
+
+  periodic:
+    triggers:
+      - schedule:
+          cron: "00 17 * * *"
+          filters:
+            branches:
+              only:
+                - master
+                - stable
+                - problame/circleci-build
+    jobs:
+      - periodic-full-pipeline-run
+
+  zrepl.github.io:
+    jobs:
+      - publish-zrepl-github-io:
+          filters:
+            branches:
+              only:
+                - stable
+
+jobs:
+  quickcheck-docs:
+    docker:
+      - image: cimg/base:2020.08
+    steps:
+      - checkout
+      - install-docdep
+      - run: make docs
+
+      - store_artifacts:
+          path: artifacts
+      - download-and-install-minio-client
+      - upload-minio:
+          src: artifacts
+          dst: ""
+      - set-github-minio-status:
+          context: artifacts/${CIRCLE_JOB}
+          description: artifacts of CI job ${CIRCLE_JOB}
+          minio-dst: ""
+
+  quickcheck-go:
+    parameters:
+      goversion:
+        type: string
+      goos:
+        type: string
+      goarch:
+        type: string
+    docker:
+      - image: circleci/golang:<<parameters.goversion>>
+    environment:
+      GOOS: <<parameters.goos>>
+      GOARCH: <<parameters.goarch>>
+
+    steps:
+      - checkout
+
+      - restore-cache-gomod
+      - run: go mod download
+      - run: cd build && go mod download
+      - save-cache-gomod
+
+      - install-godep
+      - run: make formatcheck
+      - run: make generate-platform-test-list
+      - run: make zrepl-bin test-platform-bin
+      - run: make vet
+      - run: make lint
+
+      - download-and-install-minio-client
+      - run: rm -f artifacts/generate-platform-test-list
+      - store_artifacts:
+          path: artifacts
+      - upload-minio:
+          src: artifacts
+          dst: ""
+      - set-github-minio-status:
+          context: artifacts/${CIRCLE_JOB}
+          description: artifacts of CI job ${CIRCLE_JOB}
+          minio-dst: ""
+
+  test-go-on-latest-go-release:
+    parameters:
+      goversion:
+        type: string
+    docker:
+      - image: circleci/golang:<<parameters.goversion>>
+    steps:
+      - checkout
+      - restore-cache-gomod
+      - run: make test-go
+      # don't save-cache-gomod here, test-go doesn't pull all the dependencies
+
+  release-build:
+    machine: true
+    steps:
+      - checkout
+      - run: make release-docker RELEASE_DOCKER_BASEIMAGE_TAG=<<pipeline.parameters.release_docker_baseimage_tag>>
+      - persist_to_workspace:
+          root: .
+          paths: [.]
+  release-deb:
+    machine: true
+    steps:
+      - attach_workspace:
+          at: .
+      - run: make debs-docker
+      - persist_to_workspace:
+          root: .
+          paths:
+            - "artifacts/*.deb"
+
+  release-rpm:
+    machine: true
+    steps:
+      - attach_workspace:
+          at: .
+      - run: make rpms-docker
+      - persist_to_workspace:
+          root: .
+          paths:
+            - "artifacts/*.rpm"
+
+  release-upload:
+    docker:
+      - image: cimg/base:2020.08
+    steps:
+      - attach_workspace:
+          at: .
+      - store_artifacts:
+          path: artifacts
+      - download-and-install-minio-client
+      - upload-minio:
+          src: artifacts
+          dst: ""
+      - set-github-minio-status:
+          context: artifacts/release
+          description: CI-generated release artifacts
+          minio-dst: ""
+
+  periodic-full-pipeline-run:
+    docker:
+      - image: cimg/base:2020.08
+    steps:
+      - trigger-pipeline:
+          body_no_shell_subst: '{"branch":"<<pipeline.git.branch>>", "parameters": { "do_ci": true, "do_release": true }}'
+
+  publish-zrepl-github-io:
+    docker:
+      - image: cimg/python:3.7
+    steps:
+      - checkout
+      - invoke-lazy-sh:
+          subcommand: docdep
+      - run:
+          command: |
+            git config --global user.email "me@cschwarz.com"
+            git config --global user.name "zrepl-github-io-ci"
+
+      # https://circleci.com/docs/2.0/add-ssh-key/#adding-multiple-keys-with-blank-hostnames
+      - run: ssh-add -D
+      # the default circleci ssh config only additional ssh keys for Host !github.com
+      - run:
+          command: |
+            cat > ~/.ssh/config \<<EOF
+            Host *
+              IdentityFile /home/circleci/.ssh/id_rsa_458e62c517f6c480e40452126ce47421
+            EOF
+      - add_ssh_keys:
+          fingerprints:
+            # deploy key for zrepl.github.io
+            - "45:8e:62:c5:17:f6:c4:80:e4:04:52:12:6c:e4:74:21"
+
+      - run: bash -x docs/publish.sh -c -a
@@ -0,0 +1,10 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+COMMIT="$1"
+GO_VERSION="$2"
+
+curl -v -X POST https://api.github.com/repos/zrepl/debian-binary-packaging/dispatches \
+    -H 'Accept: application/vnd.github.v3+json' \
+    -H "Authorization: token $GITHUB_ACCESS_TOKEN" \
+    --data '{"event_type": "push", "client_payload": { "zrepl_main_repo_commit": "'"$COMMIT"'", "go_version": "'"$GO_VERSION"'" }}'
@@ -1,70 +0,0 @@
-dist: xenial
-services:
- docker
-
-env: # for allow_failures: https://docs.travis-ci.com/user/customizing-the-build/
-matrix:
-  include:
-
-  - language: go
-    name: "Build in Docker (docs/installation.rst)"
-    script:
-    - sudo docker build -t zrepl_build -f build.Dockerfile .
-    - |
-      sudo docker run -it --rm \
-        -v "${PWD}:/go/src/github.com/zrepl/zrepl" \
-        --user "$(id -u):$(id -g)" \
-        zrepl_build make vendordeps release
-
-  - &zrepl_build_template
-    language: go
-    go_import_path: github.com/zrepl/zrepl
-    before_install:
-        - wget https://github.com/protocolbuffers/protobuf/releases/download/v3.6.1/protoc-3.6.1-linux-x86_64.zip
-        - echo "6003de742ea3fcf703cfec1cd4a3380fd143081a2eb0e559065563496af27807  protoc-3.6.1-linux-x86_64.zip" | sha256sum -c
-        - sudo unzip -d /usr protoc-3.6.1-linux-x86_64.zip
-        - ./lazy.sh godep
-        - make vendordeps
-    script:
-      - make
-      - make vet
-      - make test
-      - make lint
-      - make artifacts/zrepl-freebsd-amd64
-      - make artifacts/zrepl-linux-amd64
-      - make artifacts/zrepl-darwin-amd64
-    go:
-    - "1.11"
-
-  - <<: *zrepl_build_template
-    go:
-    - "1.12"
-
-  - <<: *zrepl_build_template
-    go:
-    - "master"
-
-  - &zrepl_docs_template 
-    language: python
-    python:
-    - "3.4"
-    install:
-    - sudo apt-get install libgirepository1.0-dev
-    - pip install -r docs/requirements.txt
-    script:
-    - make docs
-  - <<: *zrepl_docs_template
-    python:
-    - "3.5"
-  - <<: *zrepl_docs_template
-    python:
-    - "3.6"
-  - <<: *zrepl_docs_template
-    python:
-    - "3.7"
-  
-
-  allow_failures:
-  - <<: *zrepl_build_template
-    go:
-    - "master"
@@ -28,6 +28,8 @@ GO_BUILDFLAGS := $(GO_MOD_READONLY) $(GO_EXTRA_BUILDFLAGS)
 GO_BUILD := $(GO_ENV_VARS) $(GO) build $(GO_BUILDFLAGS) -ldflags $(GO_LDFLAGS)
 GOLANGCI_LINT := golangci-lint
 GOCOVMERGE := gocovmerge
+RELEASE_DOCKER_BASEIMAGE_TAG ?= 1.15
+RELEASE_DOCKER_BASEIMAGE ?= golang:$(RELEASE_DOCKER_BASEIMAGE_TAG)

 ifneq ($(GOARM),)
 	ZREPL_TARGET_TUPLE := $(GOOS)-$(GOARCH)v$(GOARM)
@@ -57,6 +59,85 @@ ifeq (SIGN, 1)
 endif
 	@echo "ZREPL RELEASE ARTIFACTS AVAILABLE IN artifacts/release"

+release-docker: $(ARTIFACTDIR)
+	sed 's/FROM.*!SUBSTITUTED_BY_MAKEFILE/FROM $(RELEASE_DOCKER_BASEIMAGE)/' build.Dockerfile > artifacts/release-docker.Dockerfile
+	docker build -t zrepl_release --pull -f artifacts/release-docker.Dockerfile .
+	docker run --rm -i -v $(CURDIR):/src -u $$(id -u):$$(id -g) \
+		zrepl_release \
+		make release GOOS=$(GOOS) GOARCH=$(GOARCH) GOARM=$(GOARM)
+
+debs-docker:
+	$(MAKE) _debs_or_rpms_docker _DEB_OR_RPM=deb
+rpms-docker:
+	$(MAKE) _debs_or_rpms_docker _DEB_OR_RPM=rpm
+_debs_or_rpms_docker: # artifacts/_zrepl.zsh_completion artifacts/bash_completion docs zrepl-bin
+	$(MAKE) $(_DEB_OR_RPM)-docker GOOS=linux GOARCH=amd64
+	$(MAKE) $(_DEB_OR_RPM)-docker GOOS=linux GOARCH=arm64
+	$(MAKE) $(_DEB_OR_RPM)-docker GOOS=linux GOARCH=arm GOARM=7
+	$(MAKE) $(_DEB_OR_RPM)-docker GOOS=linux GOARCH=386
+
+rpm: $(ARTIFACTDIR) # artifacts/_zrepl.zsh_completion artifacts/bash_completion docs zrepl-bin
+	$(eval _ZREPL_RPM_VERSION := $(subst -,.,$(_ZREPL_VERSION)))
+	$(eval _ZREPL_RPM_TOPDIR_ABS := $(CURDIR)/$(ARTIFACTDIR)/rpmbuild)
+	rm -rf "$(_ZREPL_RPM_TOPDIR_ABS)"
+	mkdir "$(_ZREPL_RPM_TOPDIR_ABS)"
+	mkdir -p "$(_ZREPL_RPM_TOPDIR_ABS)"/{SPECS,RPMS,BUILD,BUILDROOT}
+	sed "s/^Version:.*/Version:          $(_ZREPL_RPM_VERSION)/g" \
+	    packaging/rpm/zrepl.spec >  $(_ZREPL_RPM_TOPDIR_ABS)/SPECS/zrepl.spec
+
+	# see /usr/lib/rpm/platform
+ifeq ($(GOARCH),amd64)
+	$(eval _ZREPL_RPMBUILD_TARGET := x86_64)
+else ifeq ($(GOARCH), 386)
+	$(eval _ZREPL_RPMBUILD_TARGET := i386)
+else ifeq ($(GOARCH), arm64)
+	$(eval _ZREPL_RPMBUILD_TARGET := aarch64)
+else ifeq ($(GOARCH), arm)
+	$(eval _ZREPL_RPMBUILD_TARGET := armv7hl)
+else
+	$(eval _ZREPL_RPMBUILD_TARGET := $(GOARCH))
+endif
+	rpmbuild \
+		--build-in-place \
+		--define "_sourcedir $(CURDIR)" \
+		--define "_topdir $(_ZREPL_RPM_TOPDIR_ABS)" \
+		--define "_zrepl_binary_filename zrepl-$(ZREPL_TARGET_TUPLE)" \
+		--target $(_ZREPL_RPMBUILD_TARGET) \
+		-bb "$(_ZREPL_RPM_TOPDIR_ABS)"/SPECS/zrepl.spec
+	cp "$(_ZREPL_RPM_TOPDIR_ABS)"/RPMS/$(_ZREPL_RPMBUILD_TARGET)/zrepl-$(_ZREPL_RPM_VERSION)*.rpm $(ARTIFACTDIR)/
+
+rpm-docker:
+	docker build -t zrepl_rpm_pkg --pull -f packaging/rpm/Dockerfile .
+	docker run --rm -i -v $(CURDIR):/build/src -u $$(id -u):$$(id -g) \
+		zrepl_rpm_pkg \
+		make rpm GOOS=$(GOOS) GOARCH=$(GOARCH) GOARM=$(GOARM)
+
+deb: $(ARTIFACTDIR) # artifacts/_zrepl.zsh_completion artifacts/bash_completion docs zrepl-bin
+
+	cp packaging/deb/debian/changelog.template packaging/deb/debian/changelog
+	sed -i 's/DATE_DASH_R_OUTPUT/$(shell date -R)/' packaging/deb/debian/changelog
+	VERSION="$(subst -,.,$(_ZREPL_VERSION))"; \
+		export VERSION="$${VERSION#v}"; \
+		sed -i 's/VERSION/'"$$VERSION"'/' packaging/deb/debian/changelog
+
+ifeq ($(GOARCH), arm)
+	$(eval  DEB_HOST_ARCH := armhf)
+else ifeq ($(GOARCH), 386)
+	$(eval DEB_HOST_ARCH := i386)
+else
+	$(eval DEB_HOST_ARCH := $(GOARCH))
+endif
+
+	export ZREPL_DPKG_ZREPL_BINARY_FILENAME=zrepl-$(ZREPL_TARGET_TUPLE); \
+		dpkg-buildpackage -b --no-sign --host-arch $(DEB_HOST_ARCH)
+	cp ../*.deb artifacts/
+
+deb-docker:
+	docker build -t zrepl_debian_pkg --pull -f packaging/deb/Dockerfile .
+	docker run --rm -i -v $(CURDIR):/build/src -u $$(id -u):$$(id -g) \
+		zrepl_debian_pkg \
+		make deb GOOS=$(GOOS) GOARCH=$(GOARCH) GOARM=$(GOARM)
+
 # expects `release` target to have run before
 NOARCH_TARBALL := $(ARTIFACTDIR)/zrepl-noarch.tar
 wrapup-and-checksum:
@@ -147,26 +228,41 @@ zrepl-bin:
 generate-platform-test-list:
 	$(GO_BUILD) -o $(ARTIFACTDIR)/generate-platform-test-list ./platformtest/tests/gen

-test-platform-bin:
+COVER_PLATFORM_BIN_PATH := $(ARTIFACTDIR)/platformtest-cover-$(ZREPL_TARGET_TUPLE)
+cover-platform-bin:
 	$(GO_ENV_VARS) $(GO) test $(GO_BUILDFLAGS) \
-		-c -o "$(ARTIFACTDIR)/platformtest-$(ZREPL_TARGET_TUPLE)" \
+		-c -o "$(COVER_PLATFORM_BIN_PATH)" \
 		-covermode=atomic -cover -coverpkg github.com/zrepl/zrepl/... \
 		./platformtest/harness
+cover-platform:
+	# do not track dependency on cover-platform-bin to allow build of binary outside of test VM
+	export _TEST_PLATFORM_CMD="$(COVER_PLATFORM_BIN_PATH) \
+		-test.coverprofile \"$(ARTIFACTDIR)/platformtest.cover\" \
+		-test.v \
+		__DEVEL--i-heard-you-like-tests"; \
+		$(MAKE) _test-or-cover-platform-impl
+
+TEST_PLATFORM_BIN_PATH := $(ARTIFACTDIR)/platformtest-$(ZREPL_TARGET_TUPLE)
+test-platform-bin:
+	$(GO_BUILD) -o "$(TEST_PLATFORM_BIN_PATH)" ./platformtest/harness
+test-platform:
+	export _TEST_PLATFORM_CMD="\"$(TEST_PLATFORM_BIN_PATH)\""; \
+		$(MAKE) _test-or-cover-platform-impl

 ZREPL_PLATFORMTEST_POOLNAME := zreplplatformtest
 ZREPL_PLATFORMTEST_IMAGEPATH := /tmp/zreplplatformtest.pool.img
 ZREPL_PLATFORMTEST_MOUNTPOINT := /tmp/zreplplatformtest.pool
 ZREPL_PLATFORMTEST_ZFS_LOG := /tmp/zreplplatformtest.zfs.log
 # ZREPL_PLATFORMTEST_STOP_AND_KEEP := -failure.stop-and-keep-pool
-ZREPL_PLATFORMTEST_ARGS := 
-test-platform: $(ARTIFACTDIR) # do not track dependency on test-platform-bin to allow build of platformtest outside of test VM
+ZREPL_PLATFORMTEST_ARGS :=
+_test-or-cover-platform-impl: $(ARTIFACTDIR)
+ifndef _TEST_PLATFORM_CMD
+	$(error _TEST_PLATFORM_CMD is undefined, caller 'cover-platform' or 'test-platform' should have defined it)
+endif
 	rm -f "$(ZREPL_PLATFORMTEST_ZFS_LOG)"
 	rm -f "$(ARTIFACTDIR)/platformtest.cover"
 	platformtest/logmockzfs/logzfsenv "$(ZREPL_PLATFORMTEST_ZFS_LOG)" `which zfs` \
-	"$(ARTIFACTDIR)/platformtest-$(ZREPL_TARGET_TUPLE)" \
-		-test.coverprofile "$(ARTIFACTDIR)/platformtest.cover" \
-		-test.v \
-		__DEVEL--i-heard-you-like-tests \
+		$(_TEST_PLATFORM_CMD) \
 		-poolname "$(ZREPL_PLATFORMTEST_POOLNAME)" \
 		-imagepath "$(ZREPL_PLATFORMTEST_IMAGEPATH)" \
 		-mountpoint "$(ZREPL_PLATFORMTEST_MOUNTPOINT)" \
@@ -178,22 +274,31 @@ cover-merge: $(ARTIFACTDIR)
 cover-html: cover-merge
 	$(GO) tool cover -html "$(ARTIFACTDIR)/merged.cover" -o "$(ARTIFACTDIR)/merged.cover.html"

-test-full:
+cover-full:
 	test "$$(id -u)" = "0" || echo "MUST RUN AS ROOT" 1>&2
 	$(MAKE) test-go COVER=1
-	$(MAKE) test-platform
+	$(MAKE) cover-platform-bin
+	$(MAKE) cover-platform
 	$(MAKE) cover-html

 ##################### DEV TARGETS #####################
 # not part of the build, must do that manually
-.PHONY: generate format
+.PHONY: generate formatcheck format

 generate: generate-platform-test-list
 	protoc -I=replication/logic/pdu --go_out=plugins=grpc:replication/logic/pdu replication/logic/pdu/pdu.proto
 	$(GO_ENV_VARS) $(GO) generate $(GO_BUILDFLAGS) -x ./...

+GOIMPORTS := goimports -srcdir . -local 'github.com/zrepl/zrepl'
+FINDSRCFILES := find . -type f -name '*.go' -not -path "./vendor/*" -not -name '*.pb.go' -not -name '*_enumer.go'
+
+formatcheck:
+	@# goimports doesn't have a knob to exit with non-zero status code if formatting is needed
+	@# see https://go-review.googlesource.com/c/tools/+/237378
+	@ affectedfiles=$$($(GOIMPORTS) -l $(shell $(FINDSRCFILES)) | tee /dev/stderr | wc -l); test "$$affectedfiles" = 0
+
 format:
-	goimports -srcdir . -local 'github.com/zrepl/zrepl' -w $(shell find . -type f -name '*.go' -not -path "./vendor/*" -not -name '*.pb.go' -not -name '*_enumer.go')
+	@ $(GOIMPORTS) -w -d $(shell  $(FINDSRCFILES))

 ##################### NOARCH #####################
 .PHONY: noarch $(ARTIFACTDIR)/bash_completion $(ARTIFACTDIR)/_zrepl.zsh_completion $(ARTIFACTDIR)/go_env.txt docs docs-clean
@@ -217,11 +322,14 @@ $(ARTIFACTDIR)/_zrepl.zsh_completion:

 $(ARTIFACTDIR)/go_env.txt:
 	$(GO_ENV_VARS) $(GO) env > $@
+	$(GO) version >> $@

 docs: $(ARTIFACTDIR)/docs
+	# https://www.sphinx-doc.org/en/master/man/sphinx-build.html
 	make -C docs \
 		html \
 		BUILDDIR=../artifacts/docs \
+		SPHINXOPTS="-W --keep-going -n"

 docs-clean:
 	make -C docs \
@@ -1,7 +1,7 @@
 [![GitHub license](https://img.shields.io/github/license/zrepl/zrepl.svg)](https://github.com/zrepl/zrepl/blob/master/LICENSE)
 [![Language: Go](https://img.shields.io/badge/language-Go-6ad7e5.svg)](https://golang.org/)
 [![User Docs](https://img.shields.io/badge/docs-web-blue.svg)](https://zrepl.github.io)
-[![Donate via Patreon](https://img.shields.io/endpoint.svg?url=https%3A%2F%2Fshieldsio-patreon.herokuapp.com%2Fzrepl%2Fpledges&style=flat&color=yellow)](https://www.patreon.com/zrepl)
+[![Support me on Patreon](https://img.shields.io/badge/dynamic/json?color=yellow&label=Patreon&query=data.attributes.patron_count&suffix=%20patrons&url=https%3A%2F%2Fwww.patreon.com%2Fapi%2Fcampaigns%2F3095079)](https://patreon.com/zrepl)
 [![Donate via GitHub Sponsors](https://img.shields.io/static/v1?label=Sponsor&message=%E2%9D%A4&logo=GitHub&style=flat&color=yellow)](https://github.com/sponsors/problame)
 [![Donate via Liberapay](https://img.shields.io/liberapay/patrons/zrepl.svg?logo=liberapay)](https://liberapay.com/zrepl/donate)
 [![Donate via PayPal](https://img.shields.io/badge/donate-paypal-yellow.svg)](https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=R5QSXJVYHGX96)
@@ -21,7 +21,7 @@ zrepl is a one-stop ZFS backup & replication solution.

 ## Feature Requests

-1. Does you feature request require default values / some kind of configuration?
+1. Does your feature request require default values / some kind of configuration?
   If so, think of an expressive configuration example.
 2. Think of at least one use case that generalizes from your concrete application.
 3. Open an issue on GitHub with example conf & use case attached.
@@ -30,20 +30,12 @@ zrepl is a one-stop ZFS backup & replication solution.
 The above does not apply if you already implemented everything.
 Check out the *Coding Workflow* section below for details.

-## Package Maintainer Information
+## Building, Releasing, Downstream-Packaging

-* Follow the steps in `docs/installation.rst -> Compiling from Source` and read the Makefile / shell scripts used in this process.
-* Make sure your distro is compatible with the paths in `docs/installation.rst`.
-* Ship a default config that adheres to your distro's `hier` and logging system.
-* Ship a service manager file and _please_ try to upstream it to this repository.
-  * `dist/systemd` contains a Systemd unit template.
-* Ship other material provided in `./dist`, e.g. in `/usr/share/zrepl/`.
-* Use `make release ZREPL_VERSION='mydistro-1.2.3_1'`
-    * Your distro's name and any versioning supplemental to zrepl's (e.g. package revision) should be in this string
-* Use `sudo make test-platform` **on a test system** to validate that zrepl's abstractions on top of ZFS work with the system ZFS.
-* Make sure you are informed about new zrepl versions, e.g. by subscribing to GitHub's release RSS feed.
+This section provides an overview of the zrepl build & release process.
+Check out `docs/installation/compile-from-source.rst` for build-from-source instructions.

-## Developer Documentation
+### Overview

 zrepl is written in [Go](https://golang.org) and uses [Go modules](https://github.com/golang/go/wiki/Modules) to manage dependencies.
 The documentation is written in [ReStructured Text](http://docutils.sourceforge.net/rst.html) using the [Sphinx](https://www.sphinx-doc.org) framework.
@@ -60,60 +52,61 @@ An HTML report can be generated using `make cover-html`.

 **Code generation** is triggered by `make generate`. Generated code is committed to the source tree.

-### Project Structure
+### Build & Release Process

-```
-├── artifacts               # build artifcats generate by make
-├── cli                     # wrapper around CLI package cobra
-├── client                  # all subcommands that are not `daemon`
-├── config                  # config data types (=> package yaml-config)
-│   └── samples
-├── daemon                  # the implementation of `zrepl daemon` subcommand
-│   ├── filters
-│   ├── hooks               # snapshot hooks
-│   ├── job                 # job implementations
-│   ├── logging             # logging outlets + formatters
-│   ├── nethelpers
-│   ├── prometheus
-│   ├── pruner              # pruner implementation
-│   ├── snapper             # snapshotter implementation
-├── dist                    # supplemental material for users & package maintainers
-├── docs                    # sphinx-based documentation
-│   ├── **/*.rst            # documentation in reStructuredText
-│   ├── sphinxconf
-│   │   └── conf.py         # sphinx config (see commit 445a280 why its not in docs/)
-│   ├── requirements.txt    # pip3 requirements to build documentation
-│   ├── publish.sh          # shell script for automated rendering & deploy to zrepl.github.io repo
-│   └── public_git          # checkout of zrepl.github.io managed by above shell script
-├── endpoint                # implementation of replication endpoints (=> package replication)
-├── logger                  # our own logger package
-├── platformtest            # test suite for our zfs abstractions (error classification, etc)
-├── pruning                 # pruning rules (the logic, not the actual execution)
-│   └── retentiongrid
-├── replication
-│   ├── driver              # the driver of the replication logic (status reporting, error handling)
-│   ├── logic               # planning & executing replication steps via rpc
-|   |   └── pdu             # the generated gRPC & protobuf code used in replication (and endpoints)
-│   └── report              # the JSON-serializable report datastructures exposed to the client
-├── rpc                     # the hybrid gRPC + ./dataconn RPC client: connects to a remote replication.Endpoint
-│   ├── dataconn            # Bulk data-transfer RPC protocol
-│   ├── grpcclientidentity  # adaptor to inject package transport's 'client identity' concept into gRPC contexts
-│   ├── netadaptor          # adaptor to convert a package transport's Connecter and Listener into net.* primitives
-│   ├── transportmux        # TCP connecter and listener used to split control & data traffic
-│   └── versionhandshake    # replication protocol version handshake perfomed on newly established connections
-├── tlsconf                 # abstraction for Go TLS server + client config
-├── transport               # transport implementations
-│   ├── fromconfig
-│   ├── local
-│   ├── ssh
-│   ├── tcp
-│   └── tls
-├── util
-├── version                 # abstraction for versions (filled during build by Makefile)
-└── zfs                     # zfs(8) wrappers
-```
+**The `Makefile` is catering to the needs of developers & CI, not distro packagers**.
+It provides phony targets for
+* local development (building, running tests, etc)
+* building a release in Docker (used by the CI & release management)
+* building .deb and .rpm packages out of the release artifacts.

-### Coding Workflow
+**Build tooling & dependencies** are documented as code in `lazy.sh`.
+Go dependencies are then fetched by the go command and pip dependencies are pinned through a `requirements.txt`.
+
+**We use CircleCI for continuous integration**.
+There are two workflows:
+
+* `ci` runs for every commit / branch / tag pushed to GitHub.
+  It is supposed to run very fast (<5min and provides quick feedback to developers).
+  It runs formatting checks, lints and tests on the most important OSes / architectures.
+  Artifacts are published to minio.cschwarz.com (see GitHub Commit Status).
+
+* `release` runs
+  * on manual triggers through the CircleCI API (in order to produce a release)
+  * periodically on `master`
+  Artifacts are published to minio.cschwarz.com (see GitHub Commit Status).
+
+**Releases** are issued via Git tags + GitHub Releases feature.
+The procedure to issue a release is as follows:
+* Issue the source release:
+  * Git tag the release on the `master` branch.
+  * Push the tag.
+  * Run `./docs/publish.sh` to re-build & push zrepl.github.io.
+* Issue the official binary release:
+  * Run the `release` pipeline (triggered via CircleCI API)
+  * Download the artifacts to the release manager's machine.
+  * Create a GitHub release, edit the changelog, upload all the release artifacts, including .rpm and .deb files.
+  * Issue the GitHub release.
+  * Add the .rpm and .deb files to the official zrepl repos, publish those.
+
+**Official binary releases are not re-built when Go receives an update. If the Go update is critical to zrepl (e.g. a Go security update that affects zrepl), we'd issue a new source release**.
+The rationale for this is that whereas distros provide a mechanism for this (`$zrepl_source_release-$distro_package_revision`), GitHub Releases doesn't which means we'd need to update the existing GitHub release's assets, which nobody would notice (no RSS feed updates, etc.).
+Downstream packagers can read the changelog to determine whether they want to push that minor release into their distro or simply skip it.
+
+### Additional Notes to Distro Package Maintainers
+
+* Use `sudo make test-platform-bin && sudo make test-platform` **on a test system** to validate that zrepl's abstractions on top of ZFS work with the system ZFS.
+* Ship a default config that adheres to your distro's `hier` and logging system.
+* Ship a service manager file and _please_ try to upstream it to this repository.
+  * `dist/systemd` contains a Systemd unit template.
+* Ship other material provided in `./dist`, e.g. in `/usr/share/zrepl/`.
+* Have a look at the `Makefile`'s `ZREPL_VERSION` variable and how it passed to Go's `ldFlags`.
+  This is how `zrepl version` knows what version number to show.
+  Your build system should set the `ldFlags` flags appropriately and add a prefix or suffix that indicates that the given zrepl binary is a distro build, not an official one.
+* Make sure you are informed about new zrepl versions, e.g. by subscribing to GitHub's release RSS feed.
+
+
+## Contributing Code

 * Open an issue when starting to hack on a new feature
 * Commits should reference the issue they are related to
@@ -123,9 +116,6 @@ An HTML report can be generated using `make cover-html`.

 Backward-incompatible changes must be documented in the git commit message and are listed in `docs/changelog.rst`.

-* Config-breaking changes must contain a line `BREAK CONFIG` in the commit message
-* Other breaking changes must contain a line `BREAK` in the commit message
-
 ### Glossary & Naming Inconsistencies

 In ZFS, *dataset* refers to the objects *filesystem*, *ZVOL* and *snapshot*. <br />
@@ -142,16 +132,3 @@ variables and types are often named *dataset* when they in fact refer to a *file
 There will not be a big refactoring (an attempt was made, but it's destroying too much history without much gain).

 However, new contributions & patches should fix naming without further notice in the commit message.
-
-### RPC debugging
-
-Optionally, there are various RPC-related environment variables, that if set to something != `""` will produce additional debug output on stderr:
-
-https://github.com/zrepl/zrepl/blob/master/rpc/rpc_debug.go#L11
-
-https://github.com/zrepl/zrepl/blob/master/rpc/dataconn/dataconn_debug.go#L11
-
-https://github.com/zrepl/zrepl/blob/master/rpc/dataconn/stream/stream_debug.go#L11
-
-https://github.com/zrepl/zrepl/blob/master/rpc/dataconn/heartbeatconn/heartbeatconn_debug.go#L11
-
@@ -1,4 +1,4 @@
-FROM golang:latest
+FROM !SUBSTITUTED_BY_MAKEFILE

 RUN apt-get update && apt-get install -y \
    python3-pip \
@@ -7,6 +7,7 @@ import (

 	"github.com/spf13/cobra"
 	"github.com/spf13/pflag"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"

 	"github.com/zrepl/zrepl/config"
@@ -480,6 +480,11 @@ func (t *tui) renderReplicationReport(rep *report.Report, history *bytesProgress
 			t.newline()
 		}

+		if len(latest.Filesystems) == 0 {
+			t.write("NOTE: no filesystems were considered for replication!")
+			t.newline()
+		}
+
 		var maxFSLen int
 		for _, fs := range latest.Filesystems {
 			if len(fs.Info.Name) > maxFSLen {
@@ -63,7 +63,7 @@ func doZabsList(ctx context.Context, sc *cli.Subcommand, args []string) error {
 				defer line.Lock().Unlock()
 				if zabsListFlags.Json {
 					enc.SetIndent("", "  ")
-					if err := enc.Encode(abstractions); err != nil {
+					if err := enc.Encode(a); err != nil {
 						panic(err)
 					}
 					fmt.Println()
@@ -52,11 +52,12 @@ func (j JobEnum) Name() string {
 }

 type ActiveJob struct {
-	Type    string                `yaml:"type"`
-	Name    string                `yaml:"name"`
-	Connect ConnectEnum           `yaml:"connect"`
-	Pruning PruningSenderReceiver `yaml:"pruning"`
-	Debug   JobDebugSettings      `yaml:"debug,optional"`
+	Type        string                `yaml:"type"`
+	Name        string                `yaml:"name"`
+	Connect     ConnectEnum           `yaml:"connect"`
+	Pruning     PruningSenderReceiver `yaml:"pruning"`
+	Debug       JobDebugSettings      `yaml:"debug,optional"`
+	Replication *Replication          `yaml:"replication,optional,fromdefaults"`
 }

 type PassiveJob struct {
@@ -76,18 +77,7 @@ type SnapJob struct {
 }

 type SendOptions struct {
-	Encrypted bool                 `yaml:"encrypted"`
-	StepHolds SendOptionsStepHolds `yaml:"step_holds,optional"`
-}
-
-type SendOptionsStepHolds struct {
-	DisableIncremental bool `yaml:"disable_incremental,optional"`
-}
-
-var _ yaml.Defaulter = (*SendOptions)(nil)
-
-func (l *SendOptions) SetDefault() {
-	*l = SendOptions{Encrypted: false}
+	Encrypted bool `yaml:"encrypted,optional,default=false"`
 }

 type RecvOptions struct {
@@ -98,10 +88,13 @@ type RecvOptions struct {
 	// Reencrypt bool `yaml:"reencrypt"`
 }

-var _ yaml.Defaulter = (*RecvOptions)(nil)
+type Replication struct {
+	Protection *ReplicationOptionsProtection `yaml:"protection,optional,fromdefaults"`
+}

-func (l *RecvOptions) SetDefault() {
-	*l = RecvOptions{}
+type ReplicationOptionsProtection struct {
+	Initial     string `yaml:"initial,optional,default=guarantee_resumability"`
+	Incremental string `yaml:"incremental,optional,default=guarantee_resumability"`
 }

 type PushJob struct {
@@ -111,6 +104,9 @@ type PushJob struct {
 	Send         *SendOptions      `yaml:"send,fromdefaults,optional"`
 }

+func (j *PushJob) GetFilesystems() FilesystemsFilter { return j.Filesystems }
+func (j *PushJob) GetSendOptions() *SendOptions      { return j.Send }
+
 type PullJob struct {
 	ActiveJob `yaml:",inline"`
 	RootFS    string                   `yaml:"root_fs"`
@@ -118,6 +114,10 @@ type PullJob struct {
 	Recv      *RecvOptions             `yaml:"recv,fromdefaults,optional"`
 }

+func (j *PullJob) GetRootFS() string             { return j.RootFS }
+func (j *PullJob) GetAppendClientIdentity() bool { return false }
+func (j *PullJob) GetRecvOptions() *RecvOptions  { return j.Recv }
+
 type PositiveDurationOrManual struct {
 	Interval time.Duration
 	Manual   bool
@@ -155,6 +155,10 @@ type SinkJob struct {
 	Recv       *RecvOptions `yaml:"recv,optional,fromdefaults"`
 }

+func (j *SinkJob) GetRootFS() string             { return j.RootFS }
+func (j *SinkJob) GetAppendClientIdentity() bool { return true }
+func (j *SinkJob) GetRecvOptions() *RecvOptions  { return j.Recv }
+
 type SourceJob struct {
 	PassiveJob   `yaml:",inline"`
 	Snapshotting SnapshottingEnum  `yaml:"snapshotting"`
@@ -162,6 +166,9 @@ type SourceJob struct {
 	Send         *SendOptions      `yaml:"send,optional,fromdefaults"`
 }

+func (j *SourceJob) GetFilesystems() FilesystemsFilter { return j.Filesystems }
+func (j *SourceJob) GetSendOptions() *SendOptions      { return j.Send }
+
 type FilesystemsFilter map[string]bool

 type SnapshottingEnum struct {
@@ -317,6 +324,7 @@ type PruneKeepNotReplicated struct {
 type PruneKeepLastN struct {
 	Type  string `yaml:"type"`
 	Count int    `yaml:"count"`
+	Regex string `yaml:"regex,optional"`
 }

 type PruneKeepRegex struct { // FIXME rename to KeepRegex
@@ -0,0 +1,90 @@
+package config
+
+import (
+	"fmt"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+	"github.com/zrepl/yaml-config"
+)
+
+type A struct {
+	B  *B     `yaml:"b,optional,fromdefaults"`
+	A1 string `yaml:"a1,optional"`
+}
+
+type B struct {
+	C *C     `yaml:"c,optional,fromdefaults"`
+	D string `yaml:"d,default=ddd"`
+	E string `yaml:"e,optional"`
+}
+
+type C struct {
+	Q string `yaml:"q,optional"`
+	R string `yaml:"r,default=r"`
+}
+
+func TestDepFromDefaults(t *testing.T) {
+
+	type testcase struct {
+		name   string
+		yaml   string
+		expect *A
+	}
+
+	tcs := []testcase{
+		{
+			name: "empty",
+			yaml: `{}`,
+			expect: &A{
+				B: &B{
+					C: &C{
+						R: "r",
+					},
+					D: "ddd",
+				},
+			},
+		},
+		{
+			name: "a1 set",
+			yaml: `{"a1":"blah"}`,
+			expect: &A{
+				A1: "blah",
+				B: &B{
+					C: &C{
+						R: "r",
+					},
+					D: "ddd",
+				},
+			},
+		},
+		{
+			name: "D set",
+			yaml: `
+b:
+  d: 4d
+`,
+			expect: &A{
+				B: &B{
+					D: "4d",
+					C: &C{
+						R: "r",
+					},
+				},
+			},
+		},
+	}
+
+	for tci := range tcs {
+		t.Run(fmt.Sprintf("%d-%s", tci, tcs[tci].name), func(t *testing.T) {
+			tc := tcs[tci]
+
+			var a A
+			err := yaml.UnmarshalStrict([]byte(tc.yaml), &a)
+			require.NoError(t, err)
+
+			require.Equal(t, tc.expect, &a)
+		})
+	}
+
+}
@@ -60,9 +60,9 @@ jobs:
 	})

 	t.Run("encrypted_unspecified", func(t *testing.T) {
-		c, err := testConfig(t, fill(encrypted_unspecified))
-		assert.Error(t, err)
-		assert.Nil(t, c)
+		c = testValidConfig(t, fill(encrypted_unspecified))
+		encrypted := c.Jobs[0].Ret.(*PushJob).Send.Encrypted
+		assert.Equal(t, false, encrypted)
 	})

 	t.Run("send_not_specified", func(t *testing.T) {
@@ -37,7 +37,7 @@ func (t *RetentionIntervalList) UnmarshalYAML(u func(interface{}, bool) error) (
 		return err
 	}

-	intervals, err := parseRetentionGridIntervalsString(in)
+	intervals, err := ParseRetentionIntervalSpec(in)
 	if err != nil {
 		return err
 	}
@@ -102,7 +102,7 @@ func parseRetentionGridIntervalString(e string) (intervals []RetentionInterval,

 }

-func parseRetentionGridIntervalsString(s string) (intervals []RetentionInterval, err error) {
+func ParseRetentionIntervalSpec(s string) (intervals []RetentionInterval, err error) {

 	ges := strings.Split(s, "|")
 	intervals = make([]RetentionInterval, 0, 7*len(ges))
@@ -52,12 +52,15 @@ jobs:
  }
  send:
    encrypted: true
-    # disable incremental step holds so that
-    # - we can yank out the backup drive during replication
-    # - thereby sacrificing resumability
-    # - in exchange for the replicating snapshot not sticking around until we reconnect the backup drive
-    step_holds:
-      disable_incremental: true
+  replication:
+    protection:
+      initial: guarantee_resumability
+      # Downgrade protection to guarantee_incremental which uses zfs bookmarks instead of zfs holds.
+      # Thus, when we yank out the backup drive during replication
+      # - we might not be able to resume the interrupted replication step because the partially received `to` snapshot of a `from`->`to` step may be pruned any time
+      # - but in exchange we get back the disk space allocated by `to` when we prune it
+      # - and because we still have the bookmarks created by `guarantee_incremental`, we can still do incremental replication of `from`->`to2` in the future
+      incremental: guarantee_incremental
  snapshotting:
    type: manual
  pruning:
@@ -12,6 +12,7 @@ import (

 	"github.com/pkg/errors"
 	"github.com/prometheus/client_golang/prometheus"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"
 	"github.com/zrepl/zrepl/endpoint"
 	"github.com/zrepl/zrepl/util/envconst"
@@ -10,6 +10,7 @@ import (
 	"text/template"

 	"github.com/stretchr/testify/require"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"

 	"github.com/zrepl/zrepl/config"
@@ -9,10 +9,10 @@ import (
 	"github.com/pkg/errors"
 	"github.com/prometheus/client_golang/prometheus"
 	"github.com/prometheus/common/log"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"

 	"github.com/zrepl/zrepl/config"
-	"github.com/zrepl/zrepl/daemon/filters"
 	"github.com/zrepl/zrepl/daemon/job/reset"
 	"github.com/zrepl/zrepl/daemon/job/wakeup"
 	"github.com/zrepl/zrepl/daemon/pruner"
@@ -35,9 +35,10 @@ type ActiveSide struct {

 	prunerFactory *pruner.PrunerFactory

-	promRepStateSecs    *prometheus.HistogramVec // labels: state
-	promPruneSecs       *prometheus.HistogramVec // labels: prune_side
-	promBytesReplicated *prometheus.CounterVec   // labels: filesystem
+	promRepStateSecs      *prometheus.HistogramVec // labels: state
+	promPruneSecs         *prometheus.HistogramVec // labels: prune_side
+	promBytesReplicated   *prometheus.CounterVec   // labels: filesystem
+	promReplicationErrors prometheus.Gauge

 	tasksMtx sync.Mutex
 	tasks    activeSideTasks
@@ -145,23 +146,24 @@ func (m *modePush) ResetConnectBackoff() {

 func modePushFromConfig(g *config.Global, in *config.PushJob, jobID endpoint.JobID) (*modePush, error) {
 	m := &modePush{}
+	var err error

-	fsf, err := filters.DatasetMapFilterFromConfig(in.Filesystems)
+	m.senderConfig, err = buildSenderConfig(in, jobID)
 	if err != nil {
-		return nil, errors.Wrap(err, "cannot build filesystem filter")
+		return nil, errors.Wrap(err, "sender config")
 	}

-	m.senderConfig = &endpoint.SenderConfig{
-		FSF:                         fsf,
-		Encrypt:                     &zfs.NilBool{B: in.Send.Encrypted},
-		DisableIncrementalStepHolds: in.Send.StepHolds.DisableIncremental,
-		JobID:                       jobID,
+	replicationConfig, err := logic.ReplicationConfigFromConfig(in.Replication)
+	if err != nil {
+		return nil, errors.Wrap(err, "field `replication`")
 	}
+
 	m.plannerPolicy = &logic.PlannerPolicy{
-		EncryptedSend: logic.TriFromBool(in.Send.Encrypted),
+		EncryptedSend:     logic.TriFromBool(in.Send.Encrypted),
+		ReplicationConfig: *replicationConfig,
 	}

-	if m.snapper, err = snapper.FromConfig(g, fsf, in.Snapshotting); err != nil {
+	if m.snapper, err = snapper.FromConfig(g, m.senderConfig.FSF, in.Snapshotting); err != nil {
 		return nil, errors.Wrap(err, "cannot build snapper")
 	}

@@ -173,7 +175,6 @@ type modePull struct {
 	receiver       *endpoint.Receiver
 	receiverConfig endpoint.ReceiverConfig
 	sender         *rpc.Client
-	rootFS         *zfs.DatasetPath
 	plannerPolicy  *logic.PlannerPolicy
 	interval       config.PositiveDurationOrManual
 }
@@ -247,26 +248,19 @@ func modePullFromConfig(g *config.Global, in *config.PullJob, jobID endpoint.Job
 	m = &modePull{}
 	m.interval = in.Interval

-	m.rootFS, err = zfs.NewDatasetPath(in.RootFS)
+	replicationConfig, err := logic.ReplicationConfigFromConfig(in.Replication)
 	if err != nil {
-		return nil, errors.New("RootFS is not a valid zfs filesystem path")
-	}
-	if m.rootFS.Length() <= 0 {
-		return nil, errors.New("RootFS must not be empty") // duplicates error check of receiver
+		return nil, errors.Wrap(err, "field `replication`")
 	}

 	m.plannerPolicy = &logic.PlannerPolicy{
-		EncryptedSend: logic.DontCare,
+		EncryptedSend:     logic.DontCare,
+		ReplicationConfig: *replicationConfig,
 	}

-	m.receiverConfig = endpoint.ReceiverConfig{
-		JobID:                      jobID,
-		RootWithoutClientComponent: m.rootFS,
-		AppendClientIdentity:       false, // !
-		UpdateLastReceivedHold:     true,
-	}
-	if err := m.receiverConfig.Validate(); err != nil {
-		return nil, errors.Wrap(err, "cannot build receiver config")
+	m.receiverConfig, err = buildReceiverConfig(in, jobID)
+	if err != nil {
+		return nil, err
 	}

 	return m, nil
@@ -307,6 +301,14 @@ func activeSide(g *config.Global, in *config.ActiveJob, configJob interface{}) (
 		ConstLabels: prometheus.Labels{"zrepl_job": j.name.String()},
 	}, []string{"filesystem"})

+	j.promReplicationErrors = prometheus.NewGauge(prometheus.GaugeOpts{
+		Namespace:   "zrepl",
+		Subsystem:   "replication",
+		Name:        "filesystem_errors",
+		Help:        "number of filesystems that failed replication in the latest replication attempt, or -1 if the job failed before enumerating the filesystems",
+		ConstLabels: prometheus.Labels{"zrepl_job": j.name.String()},
+	})
+
 	j.connecter, err = fromconfig.ConnecterFromConfig(g, in.Connect)
 	if err != nil {
 		return nil, errors.Wrap(err, "cannot build client")
@@ -331,6 +333,7 @@ func (j *ActiveSide) RegisterMetrics(registerer prometheus.Registerer) {
 	registerer.MustRegister(j.promRepStateSecs)
 	registerer.MustRegister(j.promPruneSecs)
 	registerer.MustRegister(j.promBytesReplicated)
+	registerer.MustRegister(j.promReplicationErrors)
 }

 func (j *ActiveSide) Name() string { return j.name.String() }
@@ -365,7 +368,7 @@ func (j *ActiveSide) OwnedDatasetSubtreeRoot() (rfs *zfs.DatasetPath, ok bool) {
 		_ = j.mode.(*modePush) // make sure we didn't introduce a new job type
 		return nil, false
 	}
-	return pull.rootFS.Copy(), true
+	return pull.receiverConfig.RootWithoutClientComponent.Copy(), true
 }

 func (j *ActiveSide) SenderConfig() *endpoint.SenderConfig {
@@ -463,6 +466,10 @@ func (j *ActiveSide) do(ctx context.Context) {
 		GetLogger(ctx).Info("start replication")
 		repWait(true) // wait blocking
 		repCancel()   // always cancel to free up context resources
+
+		replicationReport := j.tasks.replicationReport()
+		j.promReplicationErrors.Set(float64(replicationReport.GetFailedFilesystemsCountInLatestAttempt()))
+
 		endSpan()
 	}

@@ -5,6 +5,7 @@ import (

 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
+
 	"github.com/zrepl/zrepl/endpoint"
 	"github.com/zrepl/zrepl/transport"
 )
@@ -0,0 +1,56 @@
+package job
+
+import (
+	"github.com/pkg/errors"
+
+	"github.com/zrepl/zrepl/config"
+	"github.com/zrepl/zrepl/daemon/filters"
+	"github.com/zrepl/zrepl/endpoint"
+	"github.com/zrepl/zrepl/zfs"
+)
+
+type SendingJobConfig interface {
+	GetFilesystems() config.FilesystemsFilter
+	GetSendOptions() *config.SendOptions // must not be nil
+}
+
+func buildSenderConfig(in SendingJobConfig, jobID endpoint.JobID) (*endpoint.SenderConfig, error) {
+
+	fsf, err := filters.DatasetMapFilterFromConfig(in.GetFilesystems())
+	if err != nil {
+		return nil, errors.Wrap(err, "cannot build filesystem filter")
+	}
+
+	return &endpoint.SenderConfig{
+		FSF:     fsf,
+		Encrypt: &zfs.NilBool{B: in.GetSendOptions().Encrypted},
+		JobID:   jobID,
+	}, nil
+}
+
+type ReceivingJobConfig interface {
+	GetRootFS() string
+	GetAppendClientIdentity() bool
+	GetRecvOptions() *config.RecvOptions
+}
+
+func buildReceiverConfig(in ReceivingJobConfig, jobID endpoint.JobID) (rc endpoint.ReceiverConfig, err error) {
+	rootFs, err := zfs.NewDatasetPath(in.GetRootFS())
+	if err != nil {
+		return rc, errors.New("root_fs is not a valid zfs filesystem path")
+	}
+	if rootFs.Length() <= 0 {
+		return rc, errors.New("root_fs must not be empty") // duplicates error check of receiver
+	}
+
+	rc = endpoint.ReceiverConfig{
+		JobID:                      jobID,
+		RootWithoutClientComponent: rootFs,
+		AppendClientIdentity:       in.GetAppendClientIdentity(),
+	}
+	if err := rc.Validate(); err != nil {
+		return rc, errors.Wrap(err, "cannot build receiver config")
+	}
+
+	return rc, nil
+}
@@ -2,12 +2,16 @@ package job

 import (
 	"fmt"
+	"path"
+	"path/filepath"
 	"testing"

+	"github.com/kr/pretty"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"

 	"github.com/zrepl/zrepl/config"
+	"github.com/zrepl/zrepl/transport/tls"
 )

 func TestValidateReceivingSidesDoNotOverlap(t *testing.T) {
@@ -108,3 +112,35 @@ jobs:
 	}

 }
+
+func TestSampleConfigsAreBuiltWithoutErrors(t *testing.T) {
+	paths, err := filepath.Glob("../../config/samples/*")
+	if err != nil {
+		t.Errorf("glob failed: %+v", err)
+	}
+
+	for _, p := range paths {
+
+		if path.Ext(p) != ".yml" {
+			t.Logf("skipping file %s", p)
+			continue
+		}
+
+		t.Run(p, func(t *testing.T) {
+			c, err := config.ParseConfig(p)
+			if err != nil {
+				t.Errorf("error parsing %s:\n%+v", p, err)
+			}
+
+			t.Logf("file: %s", p)
+			t.Log(pretty.Sprint(c))
+
+			tls.FakeCertificateLoading(t)
+			jobs, err := JobsFromConfig(c)
+			t.Logf("jobs: %#v", jobs)
+			assert.NoError(t, err)
+		})
+
+	}
+
+}
@@ -6,10 +6,10 @@ import (

 	"github.com/pkg/errors"
 	"github.com/prometheus/client_golang/prometheus"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"

 	"github.com/zrepl/zrepl/config"
-	"github.com/zrepl/zrepl/daemon/filters"
 	"github.com/zrepl/zrepl/daemon/logging"
 	"github.com/zrepl/zrepl/daemon/snapper"
 	"github.com/zrepl/zrepl/endpoint"
@@ -48,19 +48,9 @@ func (m *modeSink) SnapperReport() *snapper.Report { return nil }
 func modeSinkFromConfig(g *config.Global, in *config.SinkJob, jobID endpoint.JobID) (m *modeSink, err error) {
 	m = &modeSink{}

-	rootDataset, err := zfs.NewDatasetPath(in.RootFS)
+	m.receiverConfig, err = buildReceiverConfig(in, jobID)
 	if err != nil {
-		return nil, errors.New("root dataset is not a valid zfs filesystem path")
-	}
-
-	m.receiverConfig = endpoint.ReceiverConfig{
-		JobID:                      jobID,
-		RootWithoutClientComponent: rootDataset,
-		AppendClientIdentity:       true, // !
-		UpdateLastReceivedHold:     true,
-	}
-	if err := m.receiverConfig.Validate(); err != nil {
-		return nil, errors.Wrap(err, "cannot build receiver config")
+		return nil, err
 	}

 	return m, nil
@@ -74,18 +64,13 @@ type modeSource struct {
 func modeSourceFromConfig(g *config.Global, in *config.SourceJob, jobID endpoint.JobID) (m *modeSource, err error) {
 	// FIXME exact dedup of modePush
 	m = &modeSource{}
-	fsf, err := filters.DatasetMapFilterFromConfig(in.Filesystems)
+
+	m.senderConfig, err = buildSenderConfig(in, jobID)
 	if err != nil {
-		return nil, errors.Wrap(err, "cannot build filesystem filter")
-	}
-	m.senderConfig = &endpoint.SenderConfig{
-		FSF:                         fsf,
-		Encrypt:                     &zfs.NilBool{B: in.Send.Encrypted},
-		DisableIncrementalStepHolds: in.Send.StepHolds.DisableIncremental,
-		JobID:                       jobID,
+		return nil, errors.Wrap(err, "send options")
 	}

-	if m.snapper, err = snapper.FromConfig(g, fsf, in.Snapshotting); err != nil {
+	if m.snapper, err = snapper.FromConfig(g, m.senderConfig.FSF, in.Snapshotting); err != nil {
 		return nil, errors.Wrap(err, "cannot build snapper")
 	}

@@ -7,6 +7,7 @@ import (

 	"github.com/pkg/errors"
 	"github.com/prometheus/client_golang/prometheus"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"

 	"github.com/zrepl/zrepl/config"
@@ -175,8 +176,6 @@ func (j *SnapJob) doPrune(ctx context.Context) {
 		FSF:   j.fsfilter,
 		// FIXME encryption setting is irrelevant for SnapJob because the endpoint is only used as pruner.Target
 		Encrypt: &zfs.NilBool{B: true},
-		// FIXME DisableIncrementalStepHolds setting is irrelevant for SnapJob because the endpoint is only used as pruner.Target
-		DisableIncrementalStepHolds: false,
 	})
 	j.pruner = j.prunerFactory.BuildLocalPruner(ctx, sender, alwaysUpToDateReplicationCursorHistory{sender})
 	log.Info("start pruning")
@@ -9,6 +9,7 @@ import (

 	"github.com/mattn/go-isatty"
 	"github.com/pkg/errors"
+
 	"github.com/zrepl/zrepl/config"
 	"github.com/zrepl/zrepl/daemon/logging/trace"
 	"github.com/zrepl/zrepl/logger"
@@ -93,11 +93,15 @@ package trace
 import (
 	"context"
 	"fmt"
+	"os"
+	runtimedebug "runtime/debug"
 	"strings"
 	"time"

+	"github.com/kr/pretty"
 	"github.com/pkg/errors"
 	"github.com/prometheus/client_golang/prometheus"
+
 	"github.com/zrepl/zrepl/util/chainlock"
 )

@@ -120,15 +124,19 @@ func RegisterMetrics(r prometheus.Registerer) {
 }

 type traceNode struct {
-	id         string
-	annotation string
-	parentTask *traceNode
+	// NOTE: members with prefix debug are only valid if the package variable debugEnabled is true.
+
+	id                 string
+	annotation         string
+	parentTask         *traceNode
+	debugCreationStack string // debug: stack trace of when the traceNode was created

 	mtx chainlock.L

-	activeChildTasks int32 // only for task nodes, insignificant for span nodes
-	parentSpan       *traceNode
-	activeChildSpan  *traceNode // nil if task or span doesn't have an active child span
+	activeChildTasks      int32               // only for task nodes, insignificant for span nodes
+	debugActiveChildTasks map[*traceNode]bool // debug: set of active child tasks
+	parentSpan            *traceNode
+	activeChildSpan       *traceNode // nil if task or span doesn't have an active child span

 	startedAt time.Time
 	endedAt   time.Time
@@ -137,6 +145,14 @@ type traceNode struct {
 func (s *traceNode) StartedAt() time.Time { return s.startedAt }
 func (s *traceNode) EndedAt() time.Time   { return s.endedAt }

+// caller must hold mtx
+func (s *traceNode) debugString() string {
+	if !debugEnabled {
+		return "<debugEnabled=false>"
+	}
+	return pretty.Sprint(s)
+}
+
 // Returned from WithTask or WithSpan.
 // Must be called once the task or span ends.
 // See package-level docs for nesting rules.
@@ -196,8 +212,16 @@ func WithTask(ctx context.Context, taskName string) (context.Context, DoneFunc)
 		endedAt:   time.Time{},
 	}

+	if debugEnabled {
+		this.debugCreationStack = string(runtimedebug.Stack())
+		this.debugActiveChildTasks = map[*traceNode]bool{}
+	}
+
 	if parentTask != nil {
 		this.parentTask.activeChildTasks++
+		if debugEnabled {
+			this.parentTask.debugActiveChildTasks[this] = true
+		}
 		parentTask.mtx.Unlock()
 	}

@@ -218,7 +242,11 @@ func WithTask(ctx context.Context, taskName string) (context.Context, DoneFunc)
 			defer this.mtx.Lock().Unlock()

 			if this.activeChildTasks != 0 {
-				panic(errors.Wrapf(ErrTaskStillHasActiveChildTasks, "end task: %v active child tasks", this.activeChildTasks))
+				if debugEnabled {
+					// the debugString can be quite long and panic won't print it completely
+					fmt.Fprintf(os.Stderr, "going to panic due to activeChildTasks:\n%s\n", this.debugString())
+				}
+				panic(errors.WithMessagef(ErrTaskStillHasActiveChildTasks, "end task: %v active child tasks\n", this.activeChildTasks))
 			}

 			// support idempotent task ends
@@ -229,8 +257,15 @@ func WithTask(ctx context.Context, taskName string) (context.Context, DoneFunc)

 			if this.parentTask != nil {
 				this.parentTask.activeChildTasks--
+				if debugEnabled {
+					delete(this.parentTask.debugActiveChildTasks, this)
+				}
 				if this.parentTask.activeChildTasks < 0 {
-					panic("impl error: parent task with negative activeChildTasks count")
+					if debugEnabled {
+						// the debugString can be quite long and panic won't print it completely
+						fmt.Fprintf(os.Stderr, "going to panic due to activeChildTasks < 0:\n%s\n", this.parentTask.debugString())
+					}
+					panic(fmt.Sprintf("impl error: parent task with negative activeChildTasks count: %v", this.parentTask.activeChildTasks))
 				}
 			}
 			return false
@@ -279,6 +314,10 @@ func WithSpan(ctx context.Context, annotation string) (context.Context, DoneFunc
 		endedAt:   time.Time{},
 	}

+	if debugEnabled {
+		this.debugCreationStack = string(runtimedebug.Stack())
+	}
+
 	parentSpan.mtx.HoldWhile(func() {
 		if parentSpan.activeChildSpan != nil {
 			panic(ErrAlreadyActiveChildSpan)
@@ -3,9 +3,11 @@ package trace
 import (
 	"fmt"
 	"os"
+
+	"github.com/zrepl/zrepl/util/envconst"
 )

-const debugEnabled = false
+var debugEnabled = envconst.Bool("ZREPL_TRACE_DEBUG_ENABLED", false)

 func debug(format string, args ...interface{}) {
 	if !debugEnabled {
@@ -8,10 +8,10 @@ import (
 	"time"

 	"github.com/pkg/errors"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"

 	"github.com/zrepl/zrepl/config"
-	"github.com/zrepl/zrepl/daemon/filters"
 	"github.com/zrepl/zrepl/daemon/hooks"
 	"github.com/zrepl/zrepl/daemon/logging"
 	"github.com/zrepl/zrepl/logger"
@@ -49,7 +49,7 @@ type args struct {
 	ctx            context.Context
 	prefix         string
 	interval       time.Duration
-	fsf            *filters.DatasetMapFilter
+	fsf            zfs.DatasetFilter
 	snapshotsTaken chan<- struct{}
 	hooks          *hooks.List
 	dryRun         bool
@@ -109,7 +109,7 @@ func getLogger(ctx context.Context) Logger {
 	return logging.GetLogger(ctx, logging.SubsysSnapshot)
 }

-func PeriodicFromConfig(g *config.Global, fsf *filters.DatasetMapFilter, in *config.SnapshottingPeriodic) (*Snapper, error) {
+func PeriodicFromConfig(g *config.Global, fsf zfs.DatasetFilter, in *config.SnapshottingPeriodic) (*Snapper, error) {
 	if in.Prefix == "" {
 		return nil, errors.New("prefix must not be empty")
 	}
@@ -383,7 +383,7 @@ func wait(a args, u updater) state {
 	}
 }

-func listFSes(ctx context.Context, mf *filters.DatasetMapFilter) (fss []*zfs.DatasetPath, err error) {
+func listFSes(ctx context.Context, mf zfs.DatasetFilter) (fss []*zfs.DatasetPath, err error) {
 	return zfs.ZFSListMapping(ctx, mf)
 }

@@ -5,7 +5,7 @@ import (
 	"fmt"

 	"github.com/zrepl/zrepl/config"
-	"github.com/zrepl/zrepl/daemon/filters"
+	"github.com/zrepl/zrepl/zfs"
 )

 // FIXME: properly abstract snapshotting:
@@ -32,7 +32,7 @@ func (s *PeriodicOrManual) Report() *Report {
 	return nil
 }

-func FromConfig(g *config.Global, fsf *filters.DatasetMapFilter, in config.SnapshottingEnum) (*PeriodicOrManual, error) {
+func FromConfig(g *config.Global, fsf zfs.DatasetFilter, in config.SnapshottingEnum) (*PeriodicOrManual, error) {
 	switch v := in.Ret.(type) {
 	case *config.SnapshottingPeriodic:
 		snapper, err := PeriodicFromConfig(g, fsf, v)
@@ -0,0 +1 @@
+packaging/deb/debian
@@ -69,6 +69,139 @@
            "title": "Panel Title",
            "type": "text"
        },
+        {
+            "cacheTimeout": null,
+            "colorBackground": true,
+            "colorPostfix": false,
+            "colorPrefix": false,
+            "colorValue": false,
+            "colors": [
+                "#bf1b00",
+                "#508642",
+                "#bf1b00"
+            ],
+            "datasource": "${DS_PROMETHEUS}",
+            "description": "Number of filesystems that failed replications",
+            "format": "none",
+            "gauge": {
+                "maxValue": 100,
+                "minValue": 0,
+                "show": false,
+                "thresholdLabels": false,
+                "thresholdMarkers": true
+            },
+            "gridPos": {
+                "h": 3,
+                "w": 24,
+                "x": 0,
+                "y": 10
+            },
+            "id": 50,
+            "interval": null,
+            "links": [],
+            "mappingType": 1,
+            "mappingTypes": [
+                {
+                    "name": "value to text",
+                    "value": 1
+                },
+                {
+                    "name": "range to text",
+                    "value": 2
+                }
+            ],
+            "maxDataPoints": 100,
+            "nullPointMode": "connected",
+            "nullText": null,
+            "postfix": "",
+            "postfixFontSize": "50%",
+            "prefix": "",
+            "prefixFontSize": "50%",
+            "rangeMaps": [
+                {
+                    "from": "",
+                    "text": "",
+                    "to": ""
+                }
+            ],
+            "repeat": "zrepl_job_name",
+            "repeatDirection": "h",
+            "scopedVars": {
+                "zrepl_job_name": {
+                    "selected": false,
+                    "text": "desktop_to_homesrv",
+                    "value": "desktop_to_homesrv"
+                }
+            },
+            "sparkline": {
+                "fillColor": "rgba(31, 118, 189, 0.18)",
+                "full": true,
+                "lineColor": "rgb(31, 120, 193)",
+                "show": true
+            },
+            "tableColumn": "__name__",
+            "targets": [
+                {
+                    "expr": "zrepl_replication_filesystem_errors{job=\"$prom_job_name\",zrepl_job=\"$zrepl_job_name\"}",
+                    "format": "time_series",
+                    "groupBy": [
+                        {
+                            "params": [
+                                "$__interval"
+                            ],
+                            "type": "time"
+                        },
+                        {
+                            "params": [
+                                "null"
+                            ],
+                            "type": "fill"
+                        }
+                    ],
+                    "instant": true,
+                    "interval": "",
+                    "intervalFactor": 1,
+                    "legendFormat": "",
+                    "orderByTime": "ASC",
+                    "policy": "default",
+                    "refId": "A",
+                    "resultFormat": "time_series",
+                    "select": [
+                        [
+                            {
+                                "params": [
+                                    "value"
+                                ],
+                                "type": "field"
+                            },
+                            {
+                                "params": [],
+                                "type": "mean"
+                            }
+                        ]
+                    ],
+                    "tags": []
+                }
+            ],
+            "thresholds": "0,1",
+            "title": "Failed replications $zrepl_job_name",
+            "transparent": false,
+            "type": "singlestat",
+            "valueFontSize": "80%",
+            "valueMaps": [
+                {
+                    "op": "=",
+                    "text": "All failed",
+                    "value": "-1"
+                },
+                {
+                    "op": "=",
+                    "text": "All OK",
+                    "value": "0"
+                }
+            ],
+            "valueName": "avg"
+        },
        {
            "aliasColors": {},
            "bars": false,
@@ -87,7 +220,7 @@
                "h": 4,
                "w": 12,
                "x": 0,
-                "y": 10
+                "y": 13
            },
            "hiddenSeries": false,
            "id": 48,
@@ -181,7 +314,7 @@
                "h": 5,
                "w": 12,
                "x": 12,
-                "y": 10
+                "y": 13
            },
            "hiddenSeries": false,
            "id": 44,
@@ -273,7 +406,7 @@
                "h": 5,
                "w": 12,
                "x": 0,
-                "y": 14
+                "y": 17
            },
            "hiddenSeries": false,
            "id": 42,
@@ -373,7 +506,7 @@
                "h": 4,
                "w": 12,
                "x": 12,
-                "y": 15
+                "y": 18
            },
            "hiddenSeries": false,
            "id": 22,
@@ -465,7 +598,7 @@
                "h": 5,
                "w": 12,
                "x": 0,
-                "y": 19
+                "y": 22
            },
            "hiddenSeries": false,
            "id": 33,
@@ -573,7 +706,7 @@
                "h": 5,
                "w": 12,
                "x": 12,
-                "y": 19
+                "y": 22
            },
            "hiddenSeries": false,
            "id": 23,
@@ -665,7 +798,7 @@
                "h": 5,
                "w": 12,
                "x": 0,
-                "y": 24
+                "y": 27
            },
            "hiddenSeries": false,
            "id": 41,
@@ -758,7 +891,7 @@
                "h": 5,
                "w": 12,
                "x": 12,
-                "y": 24
+                "y": 27
            },
            "hiddenSeries": false,
            "id": 47,
@@ -787,7 +920,7 @@
            "stack": false,
            "steppedLine": false,
            "targets": [{
-                "expr": "zrepl_endpoint_send_abstractions_cache_entry_count",
+                "expr": "zrepl_endpoint_abstractions_cache_entry_count",
                "format": "time_series",
                "intervalFactor": 1,
                "refId": "A"
@@ -796,7 +929,7 @@
            "timeFrom": null,
            "timeRegions": [],
            "timeShift": null,
-            "title": "send abstractions cache entry count (should not be zero and not grow unboundedly)",
+            "title": "zfs abstractions cache entry count (should not be zero and not grow unboundedly)",
            "tooltip": {
                "shared": true,
                "sort": 0,
@@ -850,7 +983,7 @@
                "h": 5,
                "w": 12,
                "x": 0,
-                "y": 29
+                "y": 32
            },
            "hiddenSeries": false,
            "id": 17,
@@ -943,7 +1076,7 @@
                "h": 5,
                "w": 12,
                "x": 0,
-                "y": 34
+                "y": 37
            },
            "hiddenSeries": false,
            "id": 19,
@@ -1044,6 +1177,33 @@
            "tagsQuery": "",
            "type": "query",
            "useTags": false
+        },
+        {
+            "allValue": null,
+            "current": {
+                "text": "All",
+                "value": [
+                    "$__all"
+                ]
+            },
+            "datasource": "${DS_PROMETHEUS}",
+            "definition": "label_values(zrepl_replication_filesystem_errors{job=\"$prom_job_name\"}, zrepl_job)",
+            "hide": 2,
+            "includeAll": true,
+            "label": "Zrepl Job Name",
+            "multi": true,
+            "name": "zrepl_job_name",
+            "options": [],
+            "query": "label_values(zrepl_replication_filesystem_errors{job=\"$prom_job_name\"}, zrepl_job)",
+            "refresh": 1,
+            "regex": "",
+            "skipUrlSync": false,
+            "sort": 1,
+            "tagValuesQuery": "",
+            "tags": [],
+            "tagsQuery": "",
+            "type": "query",
+            "useTags": false
        }]
    },
    "time": {
@@ -1076,6 +1236,6 @@
    },
    "timezone": "",
    "title": "zrepl 0.3",
-    "uid": "etJuvBmGz",
-    "version": 6
-}
+    "uid": "etQuvBnGz",
+    "version": 7
+}
@@ -27,6 +27,33 @@ We use the following annotations for classifying changes:
 * |bugfix| Change that fixes a bug, no regressions or incompatibilities expected.
 * |docs| Change to the documentation.

+0.3.1
+-----
+
+Mostly a bugfix release for :ref:`zrepl 0.3 <release-0.3>`.
+
+* |feature| pruning: add optional ``regex`` field to ``last_n`` rule
+* |docs| pruning: ``grid`` : improve documentation and add an example
+* |bugfix| pruning: ``grid``:  add all snapshots that do not match the regex to the rule's destroy list.
+  This brings the implementation in line with the docs.
+* |bugfix| ``easyrsa`` script in docs
+* |bugfix| platformtest: fix skipping encryption-only tests on systems that don't support encryption
+* |bugfix| replication: report AttemptDone if no filesystems are replicated
+* |feature| status + replication: warning if replication succeeeded without any filesystem being replicated
+* |docs| update multi-job & multi-host setup section
+* RPM Packaging
+* CI infrastructure rework
+* Continuous deployment of that new `stable` branch to zrepl.github.io.
+
+.. NOTE::
+   |  zrepl is a spare-time project primarily developed by `Christian Schwarz <https://cschwarz.com>`_.
+   |  You can support maintenance and feature development through one of the following services:
+   |  |Donate via Patreon| |Donate via GitHub Sponsors| |Donate via Liberapay| |Donate via PayPal|
+   |  Note that PayPal processing fees are relatively high for small donations.
+   |  For SEPA wire transfer and **commercial support**, please `contact Christian directly <https://cschwarz.com>`_.
+
+.. _release-0.3:
+
 0.3
 ---

@@ -34,27 +61,38 @@ This is a big one! Headlining features:

 * **Resumable Send & Recv Support**
  No knobs required, automatically used where supported.
-* **Hold-Protected Send & Recv**
-  Automatic ZFS holds to ensure that we can always use resumable send&recv for a replication step.
 * **Encrypted Send & Recv Support** for OpenZFS native encryption,
  :ref:`configurable <job-send-options>` at the job level, i.e., for all filesystems a job is responsible for.
-* **Receive-side hold on last received dataset**
-  The counterpart to the replication cursor bookmark on the send-side.
-  Ensures that incremental replication will always be possible between a sender and receiver.
+* **Replication Guarantees**
+  Automatic use of ZFS holds and bookmarks to protect a replicated filesystem from losing synchronization between sender and receiver.
+  By default, zrepl guarantees that incremental replication will always be possible and interrupted steps will always be resumable.

 .. TIP::

-   We highly recommend studying the :ref:`overview section of the configuration chapter <overview-how-replication-works>` to understand how replication works.
+   We highly recommend studying the updated :ref:`overview section of the configuration chapter <overview-how-replication-works>` to understand how replication works.
+
+.. TIP::
+
+   Go 1.15 changed the default TLS validation policy to **require Subject Alternative Names (SAN) in certificates**.
+   The openssl commands we provided in the quick-start guides up to and including the zrepl 0.3 docs seem not to work properly.
+   If you encounter certificate validation errors regarding SAN and wish to continue to use your old certificates, start the zrepl daemon with env var ``GODEBUG=x509ignoreCN=0``.
+   Alternatively, generate new certificates with SANs (see :ref:`both options int the TLS transport docs <transport-tcp+tlsclientauth-certgen>` ).
+
+Quick-start guides:
+
+* We have added :ref:`another quick-start guide for a typical workstation use case for zrepl <quickstart-backup-to-external-disk>`.
+  Check it out to learn how you can use zrepl to back up your workstation's OpenZFS natively-encrypted root filesystem to an external disk.

 Additional changelog:

+* |break| Go 1.15 TLS changes mentioned above.
 * |break| |break_config| **more restrictive job names than in prior zrepl versions**
-  Starting with this version, job names are going to be embedded into ZFS holds and bookmark names (see :ref:`here<replication-cursor-and-last-received-hold>` and :ref:`here<step-holds-and-bookmarks>`).
+  Starting with this version, job names are going to be embedded into ZFS holds and bookmark names (see :ref:`this section for details <zrepl-zfs-abstractions>`).
  Therefore you might need to adjust your job names.
-  **Note that jobs** :issue:`cannot be renamed easily` **once you start using zrepl 0.3.**
+  **Note that jobs** cannot be renamed easily **once you start using zrepl 0.3.**
 * |break| |mig| replication cursor representation changed

-  * zrepl now manages the :ref:`replication cursor bookmark <replication-cursor-and-last-received-hold>` per job-filesystem tuple instead of a single replication cursor per filesystem.
+  * zrepl now manages the :ref:`replication cursor bookmark <zrepl-zfs-abstractions>` per job-filesystem tuple instead of a single replication cursor per filesystem.
    In the future, this will permit multiple sending jobs to send from the same filesystems.
  * ZFS does not allow bookmark renaming, thus we cannot migrate the old replication cursors.
  * zrepl 0.3 will automatically create cursors in the new format for new replications, and warn if it still finds ones in the old format.
@@ -62,6 +100,7 @@ Additional changelog:
    The migration will ensure that only those old-format cursors are destroyed that have been superseeded by new-format cursors.

 * |feature| New option ``listen_freebind`` (tcp, tls, prometheus listener)
+* |feature| :issue:`341` Prometheus metric for failing replications + corresponding Grafana panel
 * |feature| :issue:`265` transport/tcp: support for CIDR masks in client IP whitelist
 * |feature| documented subcommand to generate ``bash`` and ``zsh`` completions
 * |feature| :issue:`307` ``chrome://trace`` -compatible activity tracing of zrepl daemon activity
@@ -77,13 +116,6 @@ Additional changelog:
 * **[MAINTAINER NOTICE]** New platform tests in this version, please make sure you run them for your distro!
 * **[MAINTAINER NOTICE]** Please add the shell completions to the zrepl packages.

-.. NOTE::
-   |  zrepl is a spare-time project primarily developed by `Christian Schwarz <https://cschwarz.com>`_.
-   |  You can support maintenance and feature development through one of the following services:
-   |  |Donate via Patreon| |Donate via GitHub Sponsors| |Donate via Liberapay| |Donate via PayPal|
-   |  Note that PayPal processing fees are relatively high for small donations.
-   |  For SEPA wire transfer and **commercial support**, please `contact Christian directly <https://cschwarz.com>`_.
-
 0.2.1
 -----

@@ -12,6 +12,7 @@ Configuration
    configuration/transports
    configuration/filter_syntax
    configuration/sendrecvoptions
+    configuration/replication
    configuration/snapshotting
    configuration/prune
    configuration/logging
@@ -59,8 +59,8 @@ Note that snapshot-creation denoted by "(snap)" is orthogonal to whether a job i
 | Pull mode             | ``pull``     | ``source``                       | * Central backup-server for many nodes                                             |
 |                       |              | (snap)                           | * Remote server to NAS behind NAT                                                  |
 +-----------------------+--------------+----------------------------------+------------------------------------------------------------------------------------+
-| Local replication     | | ``push`` + ``sink`` in one config             | * Backup FreeBSD boot pool                                                         |
-|                       | | with :ref:`local transport <transport-local>` |                                                                                    |
+| Local replication     | | ``push`` + ``sink`` in one config             | * Backup to :ref:`locally attached disk <quickstart-backup-to-external-disk>`      |
+|                       | | with :ref:`local transport <transport-local>` | * Backup FreeBSD boot pool                                                         |
 +-----------------------+--------------+----------------------------------+------------------------------------------------------------------------------------+
 | Snap & prune-only     | ``snap``     | N/A                              | * | Snapshots & pruning but no replication                                         |
 |                       | (snap)       |                                  |   | required                                                                       |
@@ -80,16 +80,19 @@ The active side (:ref:`push <job-push>` and :ref:`pull <job-pull>` job) executes
 .. TIP::
  The progress of the active side can be watched live using the ``zrepl status`` subcommand.

+.. _overview-passive-side--client-identity:
+
 How the Passive Side Works
 --------------------------

 The passive side (:ref:`sink <job-sink>` and :ref:`source <job-source>`) waits for connections from the corresponding active side,
 using the transport listener type specified in the ``serve`` field of the job configuration.
-Each transport listener provides a client's identity to the passive side job.
-It uses the client identity for access control:
+When a client connects, the transport listener performS listener-specific access control (cert validation, IP ACLs, etc)
+and determines the *client identity*.
+The passive side job then uses this client identity as follows:

 * The ``sink`` job maps requests from different client identities to their respective sub-filesystem tree ``root_fs/${client_identity}``.
-* The ``source`` job has a whitelist of client identities that are allowed pull access.
+* The ``source`` might, in the future, embed the client identity in :ref:`zrepl's ZFS abstraction names <zrepl-zfs-abstractions>` in order to support multi-host replication.

 .. TIP::
   The implementation of the ``sink`` job requires that the connecting client identities be a valid ZFS filesystem name components.
@@ -132,23 +135,10 @@ The following high-level steps take place during replication and can be monitore
   
 The idea behind the execution order of replication steps is that if the sender snapshots all filesystems simultaneously at fixed intervals, the receiver will have all filesystems snapshotted at time ``T1`` before the first snapshot at ``T2 = T1 + $interval`` is replicated.

-Placeholder Filesystems
-^^^^^^^^^^^^^^^^^^^^^^^
-.. _replication-placeholder-property:
-
-**Placeholder filesystems** on the receiving side are regular ZFS filesystems with the placeholder property ``zrepl:placeholder=on``.
-Placeholders allow the receiving side to mirror the sender's ZFS dataset hierarchy without replicating every filesystem at every intermediary dataset path component.
-Consider the following example: ``S/H/J`` shall be replicated to ``R/sink/job/S/H/J``, but neither ``S/H`` nor ``S`` shall be replicated.
-ZFS requires the existence of ``R/sink/job/S`` and ``R/sink/job/S/H`` in order to receive into ``R/sink/job/S/H/J``.
-Thus, zrepl creates the parent filesystems as placeholders on the receiving side.
-If at some point ``S/H`` and ``S`` shall be replicated, the receiving side invalidates the placeholder flag automatically.
-The ``zrepl test placeholder`` command can be used to check whether a filesystem is a placeholder.
-
 ZFS Background Knowledge
 ^^^^^^^^^^^^^^^^^^^^^^^^
-
-This section gives some background knowledge about ZFS features that zrepl uses to guarantee that
-**incremental replication is always possible and that started replication steps can always be resumed if they are interrupted.**
+This section gives some background knowledge about ZFS features that zrepl uses to provide guarantees for a replication filesystem.
+Specifically, zrepl guarantees by default that **incremental replication is always possible and that started replication steps can always be resumed if they are interrupted.**

 **ZFS Send Modes & Bookmarks**
 ZFS supports full sends (``zfs send fs@to``) and incremental sends (``zfs send -i @from fs@to``).
@@ -166,41 +156,56 @@ An incremental send can only be resumed if ``@to`` still exists *and* either ``@

 **ZFS Holds**
 ZFS holds prevent a snapshot from being deleted through ``zfs destroy``, letting the destroy fail with a ``datset is busy`` error.
-Holds are created and referred to by a user-defined *tag*. They can be thought of as a named, persistent lock on the snapshot.
+Holds are created and referred to by a *tag*. They can be thought of as a named, persistent lock on the snapshot.
+
+
+.. _zrepl-zfs-abstractions:
+
+ZFS Abstractions Managed By zrepl
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+With the background knowledge from the previous paragraph, we now summarize the different on-disk ZFS objects that zrepl manages to provide its functionality.
+
+.. _replication-placeholder-property:
+
+**Placeholder filesystems** on the receiving side are regular ZFS filesystems with the ZFS property ``zrepl:placeholder=on``.
+Placeholders allow the receiving side to mirror the sender's ZFS dataset hierarchy without replicating every filesystem at every intermediary dataset path component.
+Consider the following example: ``S/H/J`` shall be replicated to ``R/sink/job/S/H/J``, but neither ``S/H`` nor ``S`` shall be replicated.
+ZFS requires the existence of ``R/sink/job/S`` and ``R/sink/job/S/H`` in order to receive into ``R/sink/job/S/H/J``.
+Thus, zrepl creates the parent filesystems as placeholders on the receiving side.
+If at some point ``S/H`` and ``S`` shall be replicated, the receiving side invalidates the placeholder flag automatically.
+The ``zrepl test placeholder`` command can be used to check whether a filesystem is a placeholder.

 .. _replication-cursor-and-last-received-hold:

-Guaranteeing That Incremental Sends Are Always Possible
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-
-**Replication cursor** bookmark and **last-received-hold** are managed by zrepl to ensure that future replications can always be done incrementally.
+The **replication cursor** bookmark and **last-received-hold** are managed by zrepl to ensure that future replications can always be done incrementally.
 The replication cursor is a send-side bookmark of the most recent successfully replicated snapshot,
 and the last-received-hold is a hold of that snapshot on the receiving side.
-Both are moved aomically after the receiving side has confirmed that a replication step is complete.
+Both are moved atomically after the receiving side has confirmed that a replication step is complete.

 The replication cursor has the format ``#zrepl_CUSOR_G_<GUID>_J_<JOBNAME>``.
 The last-received-hold tag has the format ``zrepl_last_received_J_<JOBNAME>``.
 Encoding the job name in the names ensures that multiple sending jobs can replicate the same filesystem to different receivers without interference.

+.. _tentative-replication-cursor-bookmarks:
+
+**Tentative replication cursor bookmarks** are short-lived bookmarks that protect the atomic moving-forward of the replication cursor and last-received-hold (see :issue:`this issue <340>`).
+They are only necessary if step holds are not used as per the :ref:`replication.protection <replication-option-protection>` setting.
+The tentative replication cursor has the format ``#zrepl_CUSORTENTATIVE_G_<GUID>_J_<JOBNAME>``.
 The ``zrepl zfs-abstraction list`` command provides a listing of all bookmarks and holds managed by zrepl.

-.. _step-holds-and-bookmarks:
-
-Guaranteeing That Sends Are Always Resumable
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+.. _step-holds:

 **Step holds** are zfs holds managed by zrepl to ensure that a replication step can always be resumed if it is interrupted, e.g., due to network outage.
 zrepl creates step holds before it attempts a replication step and releases them after the receiver confirms that the replication step is complete.
 For an initial replication ``full @initial_snap``, zrepl puts a zfs hold on ``@initial_snap``.
 For an incremental send ``@from -> @to``, zrepl puts a zfs hold on both ``@from`` and ``@to``.
 Note that ``@from`` is not strictly necessary for resumability -- a bookmark on the sending side would be sufficient --, but size-estimation in currently used OpenZFS versions only works if ``@from`` is a snapshot.
-
 The hold tag has the format ``zrepl_STEP_J_<JOBNAME>``.
 A job only ever has one active send per filesystem.
 Thus, there are never more than two step holds for a given pair of ``(job,filesystem)``.

 **Step bookmarks** are zrepl's equivalent for holds on bookmarks (ZFS does not support putting holds on bookmarks).
-They are intended for a situation where a replication step uses a bookmark ``#bm`` as incremental ``from`` that is not managed by zrepl.
+They are intended for a situation where a replication step uses a bookmark ``#bm`` as incremental ``from`` where ``#bm`` is not managed by zrepl.
 To ensure resumability, zrepl copies ``#bm`` to step bookmark ``#zrepl_STEP_G_<GUID>_J_<JOBNAME>``.
 If the replication is interrupted and ``#bm`` is deleted by the user, the step bookmark remains as an incremental source for the resumable send.
 Note that zrepl does not yet support creating step bookmarks because the `corresponding ZFS feature for copying bookmarks <https://github.com/openzfs/zfs/pull/9571>`_ is not yet widely available .
@@ -227,39 +232,53 @@ Limitations
 Multiple Jobs & More than 2 Machines
 ------------------------------------

+The quick-start guides focus on simple setups with a single sender and a single receiver.
+This section documents considerations for more complex setups.
+
 .. ATTENTION::

-  When using multiple jobs across single or multiple machines, the following rules are critical to avoid race conditions & data loss:
+   Before you continue, make sure you have a working understanding of :ref:`how zrepl works <overview-how-replication-works>`
+   and :ref:`what zrepl does to ensure <zrepl-zfs-abstractions>` that replication between sender and receiver is always
+   possible without conflicts.
+   This will help you understand why certain kinds of multi-machine setups do not (yet) work.

-  1. The sets of ZFS filesystems matched by the ``filesystems`` filter fields must be disjoint across all jobs configured on a machine.
-  2. The ZFS filesystem subtrees of jobs with ``root_fs`` must be disjoint.
-  3. Across all zrepl instances on all machines in the replication domain, there must be a 1:1 correspondence between active and passive jobs.
+.. NOTE::

-  Explanations & exceptions to above rules are detailed below.
+   If you can't find your desired configuration, have questions or would like to see improvements to multi-job setups, please `open an issue on GitHub <https://github.com/zrepl/zrepl/issues/new>`_.

-If you would like to see improvements to multi-job setups, please `open an issue on GitHub <https://github.com/zrepl/zrepl/issues/new>`_.
+Multiple Jobs on one Machine
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+As a general rule, multiple jobs configured on one machine **must operate on disjoint sets of filesystems**.
+Otherwise, concurrently running jobs might interfere when operating on the same filesystem.

-No Overlapping
-^^^^^^^^^^^^^^
+On your setup, ensure that

-Jobs run independently of each other.
-If two jobs match the same filesystem with their ``filesystems`` filter, they will operate on that filesystem independently and potentially in parallel.
-For example, if job A prunes snapshots that job B is planning to replicate, the replication will fail because B assumed the snapshot to still be present.
-However, the next replication attempt will re-examine the situation from scratch and should work.
+* all ``filesystems`` filter specifications are disjoint
+* no ``root_fs`` is a prefix or equal to another ``root_fs``
+* no ``filesystems`` filter matches any ``root_fs``

-N push jobs to 1 sink
-^^^^^^^^^^^^^^^^^^^^^
+**Exceptions to the rule**:

-The :ref:`sink job <job-sink>` namespaces by client identity.
-It is thus safe to push to one sink job with different client identities.
-If the push jobs have the same client identity, the filesystems matched by the push jobs must be disjoint to avoid races.
+* A ``snap`` and ``push`` job on the same machine can match the same ``filesystems``.
+  To avoid interference, only one of the jobs should be pruning snapshots on the sender, the other one should keep all snapshots.
+  Since the jobs won't coordinate, errors in the log are to be expected, but :ref:`zrepl's ZFS abstractions <zrepl-zfs-abstractions>` ensure that ``push`` and ``sink`` can always replicate incrementally.
+  This scenario is detailed in one of the :ref:`quick-start guides <quickstart-backup-to-external-disk>`.

-N pull jobs from 1 source
-^^^^^^^^^^^^^^^^^^^^^^^^^

-Multiple pull jobs pulling from the same source have potential for race conditions during pruning:
-each pull job prunes the source side independently, causing replication-prune and prune-prune races.
+More Than 2 Machines
+^^^^^^^^^^^^^^^^^^^^

-There is currently no way for a pull job to filter which snapshots it should attempt to replicate.
-Thus, it is not possible to just manually assert that the prune rules of all pull jobs are disjoint to avoid replication-prune and prune-prune races.
+This section might be relevant to users who wish to *fan-in* (N machines replicate to 1) or *fan-out* (replicate 1 machine to N machines).
+
+**Working setups**:
+
+* N ``push`` identities, 1 ``sink`` (as long as the different push jobs have a different :ref:`client identity <overview-passive-side--client-identity>`)
+
+  * ``sink`` constrains each client to a disjoint sub-tree of the sink-side dataset hierarchy ``${root_fs}/${client_identity}``.
+    Therefore, the different clients cannot interfere.
+
+
+**Setups that do not work**:
+
+* N ``pull`` identities, 1 ``source`` job. Tracking :issue:`380`.

@@ -46,6 +46,8 @@ Example Configuration:
             grid: 1x1h(keep=all) | 24x1h | 35x1d | 6x30d
             regex: "^zrepl_.*"
           # manually created snapshots will be kept forever on receiver
+           - type: regex
+             regex: "^manual_.*"

 .. DANGER::
    You might have **existing snapshots** of filesystems affected by pruning which you want to keep, i.e. not be destroyed by zrepl.
@@ -82,34 +84,82 @@ Policy ``grid``
        - type: grid
          regex: "^zrepl_.*"
          grid: 1x1h(keep=all) | 24x1h | 35x1d | 6x30d
-                │                │
-                └─ one hour interval
-                                 │
-                                 └─ 24 adjacent one-hour intervals
+                │                │               │
+                └─ 1 repetition of a one-hour interval with keep=all
+                                 │               │
+                                 └─ 24 repetitions of a one-hour interval with keep=1
+                                                 │
+                                                 └─ 6 repetitions of a 30-day interval with keep=1
      ...

-The retention grid can be thought of as a time-based sieve:
-The ``grid`` field specifies a list of adjacent time intervals:
-the left edge of the leftmost (first) interval is the ``creation`` date of the youngest snapshot.
-All intervals to its right describe time intervals further in the past.
-
-Each interval carries a maximum number of snapshots to keep.
-It is specified via ``(keep=N)``, where ``N`` is either ``all`` (all snapshots are kept) or a positive integer.
-The default value is **keep=1**.
+The retention grid can be thought of as a time-based sieve that thins out snapshots as they get older.

+The ``grid`` field specifies a list of adjacent time intervals.
+Each interval is a bucket with a maximum capacity of ``keep`` snapshots.
 The following procedure happens during pruning:

 #. The list of snapshots is filtered by the regular expression in ``regex``.
-   Only snapshots names that match the regex are considered for this rule, all others are not affected.
-#. The filtered list of snapshots is sorted by ``creation``
-#. The left edge of the first interval is aligned to the ``creation`` date of the youngest snapshot
-#. A list of buckets is created, one for each interval
-#. The list of snapshots is split up into the buckets.
-#. For each bucket
+   Only snapshots names that match the regex are considered for this rule, all others will be pruned unless another rule keeps them.
+#. The snapshots that match ``regex`` are placed onto a time axis according to their ``creation`` date.
+   The youngest snapshot is on the left, the oldest on the right.
+#. The first buckets are placed "under" that axis so that the ``grid`` spec's first bucket's left edge aligns with youngest snapshot.
+#. All subsequent buckets are placed adjacent to their predecessor bucket.
+#. Now each snapshot on the axis either falls into one bucket or it is older than our rightmost bucket.
+   Buckets are left-inclusive and right-exclusive which means that a snapshot on the edge of bucket will always 'fall into the right one'.
+#. Snapshots older than the rightmost bucket **not kept** by this gridspec.
+#. For each bucket, we only keep the ``keep`` oldest snapshots.

-   #. the contained snapshot list is sorted by creation.
-   #. snapshots from the list, oldest first, are destroyed until the specified ``keep`` count is reached.
-   #. all remaining snapshots on the list are kept.
+The syntax to describe the bucket list is as follows:
+
+::
+
+     Repeat x Duration (keep=all)
+
+* The **duration** specifies the length of the interval.
+* The **keep** count specifies the number of snapshots that fit into the bucket.
+  It can be either a positive integer or ``all`` (all snapshots are kept).
+* The **repeat** count repeats the bucket definition for the specified number of times.
+
+**Example**:
+
+::
+
+          This grid spec produces the following list of adjacent buckets. For the sake of simplicity,
+          we subject all snapshots to the grid pruning policy by settings `regex: .*`.
+
+          `
+           grid: 1x1h(keep=all) | 2x2h | 1x3h
+           regex: .*
+          `
+
+          0h        1h        2h        3h        4h        5h        6h        7h        8h        9h
+          |         |         |         |         |         |         |         |         |         |
+          |-Bucket1-|-----Bucket 2------|------Bucket 3-----|-----------Bucket 4----------|
+          | keep=all|      keep=1       |       keep=1      |            keep=1           |
+
+
+
+          Let us consider the following set of snapshots @a-zA-C:
+
+
+          |  a  b  c  d  e  f  g  h  i  j  k  l  m  n  o  p  q  r  s  t  u  v  w  x  y  z  A  B  C  D        |
+
+          The `grid` algorithm maps them to their respective buckets:
+
+          Bucket 1: a, b, c
+          Bucket 2: d,e,f,g,h,i,j
+          Bucket 3: k,l,m,n,o,p
+          Bucket 4: q,r, q,r,s,t,u,v,w,x,y,z
+          None:     A,B,C,D
+
+          It then applies the per-bucket pruning logic described above which resulting in the
+          following list of remaining snapshots.
+
+          |  a  b  c                    j                 p                             z                    |
+
+          Note that it only makes sense to grow (not shorten) the interval duration for buckets
+          further in the past since each bucket acts like a low-pass filter for incoming snapshots
+          and adding a less-low-pass-filter after a low-pass one has no effect.


 .. _prune-keep-last-n:
@@ -125,9 +175,11 @@ Policy ``last_n``
         keep_receiver:
         - type: last_n
           count: 10
+           regex: ^zrepl_.*$ # optional
     ...

-``last_n`` keeps the last ``count`` snapshots (last = youngest = most recent creation date).
+``last_n`` filters the snapshot list by ``regex``, then keeps the last ``count`` snapshots in that list (last = youngest = most recent creation date)
+All snapshots that don't match ``regex`` or exceed ``count`` in the filtered list are destroyed unless matched by other rules.

 .. _prune-keep-regex:

@@ -0,0 +1,47 @@
+.. include:: ../global.rst.inc
+
+
+Replication Options
+===================
+
+
+::
+   
+   jobs:
+   - type: push
+     filesystems: ...
+     replication:
+       protection:
+         initial:     guarantee_resumability # guarantee_{resumability,incremental,nothing}
+         incremental: guarantee_resumability # guarantee_{resumability,incremental,nothing}
+     ...
+
+.. _replication-option-protection:
+
+``protection`` option
+--------------------------
+
+The ``protection`` variable controls the degree to which a replicated filesystem is protected from getting out of sync through a zrepl pruner or external tools that destroy snapshots.
+zrepl can guarantee :ref:`resumability <step-holds>` or just :ref:`incremental replication <replication-cursor-and-last-received-hold>`.
+
+``guarantee_resumability`` is the **default** value and guarantees that a replication step is always resumable and that incremental replication will always be possible.
+The implementation uses replication cursors, last-received-hold and step holds.
+
+``guarantee_incremental`` only guarantees that incremental replication will always be possible.
+If a step ``from -> to`` is interrupted and its `to` snapshot is destroyed, zrepl will remove the half-received ``to``'s resume state and start a new step ``from -> to2``.
+The implementation uses replication cursors, tentative replication cursors and last-received-hold.
+
+``guarantee_nothing`` does not make any guarantees with regards to keeping sending and receiving side in sync.
+No bookmarks or holds are created to protect sender and receiver from diverging.
+
+**Tradeoffs**
+
+Using ``guarantee_incremental`` instead of ``guarantee_resumability`` obviously removes the resumability guarantee.
+This means that replication progress is no longer monotonic which might lead to a replication setup that never makes progress if mid-step interruptions are too frequent (e.g. frequent network outages).
+However, the advantage and :issue:`reason for existence <288>` of the ``incremental`` mode is that it allows the pruner to delete snapshots of interrupted replication steps
+which is useful if replication happens so rarely (or fails so frequently) that the amount of disk space exclusively referenced by the step's snapshots becomes intolerable.
+
+.. NOTE::
+
+   When changing this flag, obsoleted zrepl-managed bookmarks and holds will be destroyed on the next replication step that is attempted for each filesystem.
+
@@ -16,8 +16,6 @@ Send Options
     filesystems: ...
     send:
       encrypted: true
-       step_holds:
-         disable_incremental: false
     ...

 :ref:`Source<job-source>` and :ref:`push<job-push>` jobs have an optional ``send`` configuration section.
@@ -36,24 +34,6 @@ Filesystems matched by ``filesystems`` that are not encrypted are not sent and w

 If ``encryption=false``, zrepl expects that filesystems matching ``filesystems`` are not encrypted or have loaded encryption keys.

-.. _job-send-option-step-holds-disable-incremental:
-
-``step_holds.disable_incremental`` option
-----------------------------------------
-
-The ``step_holds.disable_incremental`` variable controls whether the creation of :ref:`step holds <step-holds-and-bookmarks>` should be disabled for incremental replication.
-The default value is ``false``.
-
-Disabling step holds has the disadvantage that steps :ref:`might not be resumable <step-holds-and-bookmarks>` if interrupted.
-Non-resumability means that replication progress is no longer monotonic which might result in a replication setup that never makes progress if mid-step interruptions are too frequent (e.g. frequent network outages).
-
-However, the advantage and :issue:`reason for existence <288>` of this flag is that it allows the pruner to delete snapshots of interrupted replication steps
-which is useful if replication happens so rarely (or fails so frequently) that the amount of disk space exclusively referenced by the step's snapshots becomes intolerable.
-
-.. NOTE::
-
-   When setting this flag to ``true``, existing step holds for the job will be destroyed on the next replication attempt.
-
 .. _job-recv-options:

 Recv Options
@@ -131,7 +131,7 @@ The following environment variables are set:
 * ``ZREPL_SNAPNAME``: the zrepl-generated snapshot name (e.g. ``zrepl_20380119_031407_000``)
 * ``ZREPL_DRYRUN``: set to ``"true"`` if a dry run is in progress so scripts can print, but not run, their commands

-An empty template hook can be found in :sampleconf:`hooks/template.sh`.
+An empty template hook can be found in :sampleconf:`/hooks/template.sh`.

 .. _job-hook-type-postgres-checkpoint:

@@ -91,7 +91,8 @@ The ``tls`` transport uses TCP + TLS with client authentication using client cer
 The client identity is the common name (CN) presented in the client certificate.

 It is recommended to set up a dedicated CA infrastructure for this transport, e.g. using OpenVPN's `EasyRSA <https://github.com/OpenVPN/easy-rsa>`_.
-For a simple 2-machine setup, see the :ref:`instructions below<transport-tcp+tlsclientauth-2machineopenssl>`.
+For a simple 2-machine setup, mutual TLS might also be sufficient.
+We provide :ref:`copy-pastable instructions to generate the certificates below <transport-tcp+tlsclientauth-certgen>`.

 The implementation uses `Go's TLS library <https://golang.org/pkg/crypto/tls/>`_.
 Since Go binaries are statically linked, you or your distribution need to recompile zrepl when vulnerabilities in that library are disclosed.
@@ -103,6 +104,16 @@ If intermediate CAs are used, the **full chain** must be present in either in th
 Regardless, the client's certificate must be first in the ``cert`` file, with each following certificate directly certifying the one preceding it (see `TLS's specification <https://tools.ietf.org/html/rfc5246#section-7.4.2>`_).
 This is the common default when using a CA management tool.

+.. NOTE::
+
+   As of Go 1.15 (zrepl 0.3.0 and newer), the Go TLS / x509 library **requrires Subject Alternative Names**
+   be present in certificates. You might need to re-generate your certificates using one of the :ref:`two alternatives
+   provided below<transport-tcp+tlsclientauth-certgen>`.
+
+   Note further that zrepl continues to use the CommonName field to assign client identities.
+   Hence, we recommend to keep the Subject Alternative Name and the CommonName in sync.
+
+
 Serve
 ~~~~~

@@ -147,45 +158,25 @@ The ``server_cn`` specifies the expected common name (CN) of the server's certif
 It overrides the hostname specified in ``address``.
 The connection fails if either do not match.

+.. _transport-tcp+tlsclientauth-certgen:
+
 .. _transport-tcp+tlsclientauth-2machineopenssl:

-Self-Signed Certificates
-~~~~~~~~~~~~~~~~~~~~~~~~
+Mutual-TLS between Two Machines
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-Tools like `EasyRSA <https://github.com/OpenVPN/easy-rsa>`_ make it easy to manage CA infrastructure for multiple clients, e.g. a central zrepl backup server (in sink mode).
 However, for a two-machine setup, self-signed certificates distributed using an out-of-band mechanism will also work just fine:

 Suppose you have a push-mode setup, with `backups.example.com` running the :ref:`sink job <job-sink>`, and `prod.example.com` running the :ref:`push job <job-push>`.
 Run the following OpenSSL commands on each host, substituting HOSTNAME in both filenames and the interactive input prompt by OpenSSL:

 .. code-block:: bash
-   :emphasize-lines: 1-5,24

-   openssl req -x509 -sha256 -nodes \
-      -newkey rsa:4096 \
-      -days 365 \
-      -keyout HOSTNAME.key \
-      -out HOSTNAME.crt
-
-   #Generating a 4096 bit RSA private key
-   #................++++
-   #.++++
-   #writing new private key to 'backups.key'
-   #-----
-   #You are about to be asked to enter information that will be incorporated
-   #into your certificate request.
-   #What you are about to enter is what is called a Distinguished Name or a DN.
-   #There are quite a few fields but you can leave some blank
-   #For some fields there will be a default value,
-   #If you enter '.', the field will be left blank.
-   #-----
-   #Country Name (2 letter code) [XX]:
-   #State or Province Name (full name) []:
-   #Locality Name (eg, city) [Default City]:
-   #Organization Name (eg, company) [Default Company Ltd]:
-   #Organizational Unit Name (eg, section) []:
-   #Common Name (eg, your name or your server's hostname) []:HOSTNAME
-   #Email Address []:
+   (name=HOSTNAME; openssl req -x509 -sha256 -nodes \
+    -newkey rsa:4096 \
+    -days 365 \
+    -keyout $name.key \
+    -out $name.crt -addext "subjectAltName = DNS:$name" -subj "/CN=$name")

 Now copy each machine's ``HOSTNAME.crt`` to the other machine's ``/etc/zrepl/HOSTNAME.crt``, for example using `scp`.
 The serve & connect configuration will thus look like the following:
@@ -216,6 +207,36 @@ The serve & connect configuration will thus look like the following:
     ...


+Certificate Authority using EasyRSA
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+For more than two machines, it might make sense to set up a CA infrastructure.
+Tools like `EasyRSA <https://github.com/OpenVPN/easy-rsa>`_ make this very easy:
+
+::
+
+     #!/usr/bin/env bash
+     set -euo pipefail
+
+     HOSTS=(backupserver prod1 prod2 prod3)
+
+     curl -L https://github.com/OpenVPN/easy-rsa/releases/download/v3.0.7/EasyRSA-3.0.7.tgz > EasyRSA-3.0.7.tgz
+     echo "157d2e8c115c3ad070c1b2641a4c9191e06a32a8e50971847a718251eeb510a8  EasyRSA-3.0.7.tgz" | sha256sum -c
+     rm -rf EasyRSA-3.0.7
+     tar -xf EasyRSA-3.0.7.tgz
+     cd EasyRSA-3.0.7
+     ./easyrsa
+     ./easyrsa init-pki
+     ./easyrsa build-ca nopass
+
+     for host in "${HOSTS[@]}"; do
+         ./easyrsa build-serverClient-full $host nopass
+         echo cert for host $host available at pki/issued/$host.crt
+         echo key for host $host available at pki/private/$host.key
+     done
+     echo ca cert available at pki/ca.crt
+
+
 .. _transport-ssh+stdinserver:

 ``ssh+stdinserver`` Transport
@@ -58,13 +58,11 @@ for latest_patch in latest_by_major_minor:
    cmdline.append("--whitelist-tags")
    cmdline.append(f"^{re.escape(latest_patch.orig)}$")

-# we want to render the latest non-rc version as the default page
-# (latest_by_major_minor is already sorted)
-default_version = latest_by_major_minor[-1]
-for tag in reversed(latest_by_major_minor):
-    if tag.rc == 0:
-        default_version = tag
-        break
+# we want flexibility to update docs for the latest stable release
+# => we have a branch for that, called `stable` which we move manually
+# TODO: in the future, have f"stable-{latest_by_major_minor[-1]}"
+default_version = "stable"
+cmdline.extend(["--whitelist-branches", default_version])

 cmdline.extend(["--root-ref", f"{default_version}"])
 cmdline.extend(["--banner-main-ref", f"{default_version}"])
@@ -73,4 +71,4 @@ cmdline.extend(["--sort", "semver"])

 cmdline.extend(["--whitelist-branches", "master"])

-print(" ".join(cmdline))
+print(" ".join(cmdline))
@@ -13,7 +13,7 @@
   :target: https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=R5QSXJVYHGX96
 .. |Donate via Liberapay| image:: https://img.shields.io/liberapay/patrons/zrepl.svg?logo=liberapay
   :target: https://liberapay.com/zrepl/donate
-.. |Donate via Patreon| image:: https://img.shields.io/endpoint.svg?url=https%3A%2F%2Fshieldsio-patreon.herokuapp.com%2Fzrepl%2Fpledges&style=flat&color=yellow
+.. |Donate via Patreon| image:: https://img.shields.io/badge/dynamic/json?color=yellow&label=Patreon&query=data.attributes.patron_count&suffix=%20patrons&url=https%3A%2F%2Fwww.patreon.com%2Fapi%2Fcampaigns%2F3095079
   :target: https://www.patreon.com/zrepl
 .. |Donate via GitHub Sponsors| image:: https://img.shields.io/static/v1?label=Sponsor&message=%E2%9D%A4&logo=GitHub&style=flat&color=yellow
   :target: https://github.com/sponsors/problame
@@ -14,6 +14,7 @@ Installation
    installation/user-privileges
    installation/packages
    installation/apt-repos
+    installation/rpm-repos
    installation/compile-from-source
    installation/freebsd-jail-with-iocage
    installation/what-next
@@ -7,7 +7,7 @@ Debian / Ubuntu APT repositories
 We maintain APT repositories for Debian, Ubuntu and derivatives.
 The fingerprint of the signing key is ``E101 418F D3D6 FBCB 9D65  A62D 7086 99FC 5F2E BF16``.
 It is available at `<https://zrepl.cschwarz.com/apt/apt-key.asc>`_ .
-Please open an issue `in the packaging repository <https://github.com/zrepl/debian-binary-packaging>`_ if you encounter any issues with the repository.
+Please open an issue in on GitHub if you encounter any issues with the repository.

 The following snippet configure the repository for your Debian or Ubuntu release:

@@ -24,6 +24,6 @@ The following snippet configure the repository for your Debian or Ubuntu release

 .. NOTE::

-   Until zrepl reaches 1.0, all APT repositories will be updated to the latest zrepl release immediately.
+   Until zrepl reaches 1.0, the repositories will be updated to the latest zrepl release immediately.
   This includes breaking changes between zrepl versions.
   Use ``apt-mark hold zrepl`` to prevent upgrades of zrepl.
@@ -9,37 +9,35 @@ Producing a release requires **Go 1.11** or newer and **Python 3** + **pip3** +
 A tutorial to install Go is available over at `golang.org <https://golang.org/doc/install>`_.
 Python and pip3 should probably be installed via your distro's package manager.

+::
+   cd to/your/zrepl/checkout
+   python3 -m venv3
+   source venv3/bin/activate
+   ./lazy.sh devsetup
+   make release
+   # build artifacts are available in ./artifacts/release
+
+The Python venv is used for the documentation build dependencies.
+If you just want to build the zrepl binary, leave it out and use `./lazy.sh godep` instead.
+
 Alternatively, you can use the Docker build process:
 it is used to produce the official zrepl `binary releases`_
 and serves as a reference for build dependencies and procedure:

 ::

-    git clone https://github.com/zrepl/zrepl.git && \
-    cd zrepl && \
-    sudo docker build -t zrepl_build -f build.Dockerfile . && \
-    sudo docker run -it --rm \
-        -v "${PWD}:/src" \
-        --user "$(id -u):$(id -g)" \
-        zrepl_build make release
+    cd to/your/zrepl/checkout
+    # make sure your user has access to the docker socket
+    make release-docker
+    # if you want .deb or .rpm packages, invoke the follwoing
+    # targets _after_ you invoked release-docker
+    make deb-docker
+    make rpm-docker
+    # build artifacts are available in ./artifacts/release
+    # packages are available in ./artifacts

-Alternatively, you can install build dependencies on your local system and then build in your ``$GOPATH``:
-
-::
-
-    mkdir -p "${GOPATH}/src/github.com/zrepl/zrepl"
-    git clone https://github.com/zrepl/zrepl.git "${GOPATH}/src/github.com/zrepl/zrepl"
-    cd "${GOPATH}/src/github.com/zrepl/zrepl"
-    python3 -m venv3
-    source venv3/bin/activate
-    ./lazy.sh devsetup
-    make release
-
-The Python venv is used for the documentation build dependencies.
-If you just want to build the zrepl binary, leave it out and use `./lazy.sh godep` instead.
-Either way, all build results are located in the ``artifacts/`` directory.

 .. NOTE::

-    It is your job to install the appropriate binary in the zrepl users's ``$PATH``, e.g. ``/usr/local/bin/zrepl``.
+    It is your job to install the built binary in the zrepl users's ``$PATH``, e.g. ``/usr/local/bin/zrepl``.
    Otherwise, the examples in the :ref:`quick-start guides <quickstart-toc>` may need to be adjusted.
@@ -30,15 +30,12 @@ The following list may be incomplete, feel free to submit a PR with an update:
    * - Arch Linux
      - ``yay install zrepl``
      - Available on `AUR <https://aur.archlinux.org/packages/zrepl>`_
-    * - Fedora
+    * - Fedora, CentOS, RHEL, OpenSUSE
      - ``dnf install zrepl``
-      - Available on `COPR <https://copr.fedorainfracloud.org/coprs/poettlerric/zrepl/>`_
-    * - CentOS/RHEL
-      - ``yum install zrepl``
-      - Available on `COPR <https://copr.fedorainfracloud.org/coprs/poettlerric/zrepl/>`_
+      - :ref:`RPM repository config <installation-rpm-repos>`
    * - Debian + Ubuntu
      - ``apt install zrepl``
-      - APT repository config :ref:`see below <installation-apt-repos>`
+      - :ref:`APT repository config <installation-apt-repos>`
    * - OmniOS
      - ``pkg install zrepl``
      - Available since `r151030 <https://pkg.omniosce.org/r151030/extra/en/search.shtml?token=zrepl&action=Search>`_
@@ -0,0 +1,30 @@
+.. _installation-rpm-repos:
+
+RPM repositories
+~~~~~~~~~~~~~~~~
+
+We provide a single RPM repository for all RPM-based Linux distros.
+The zrepl binary in the repo is the same as the one published to GitHub.
+Since Go binaries are statically linked, the RPM should work about everywhere.
+
+The fingerprint of the signing key is ``F6F6 E8EA 6F2F 1462 2878 B5DE 50E3 4417 826E 2CE6``.
+It is available at `<https://zrepl.cschwarz.com/rpm/rpm-key.asc>`_ .
+Please open an issue on GitHub if you encounter any issues with the repository.
+
+Copy-paste the following snippet into your shell to set up the zrepl repository.
+Then ``dnf install zrepl`` and make sure to confirm that the signing key matches the one shown above.
+
+::
+
+    cat > /etc/yum.repos.d/zrepl.repo <<EOF
+    [zrepl]
+    name = zrepl
+    baseurl = https://zrepl.cschwarz.com/rpm/repo
+    gpgkey = https://zrepl.cschwarz.com/rpm/rpm-key.asc
+    EOF
+
+.. NOTE::
+
+   Until zrepl reaches 1.0, the repository will be updated to the latest zrepl release immediately.
+   This includes breaking changes between zrepl versions.
+   If that bothers you, use the `dnf versionlock plugin <https://dnf-plugins-core.readthedocs.io/en/latest/versionlock.html>`_ to pin the version of zrepl on your system.
@@ -1,14 +1,30 @@
 #!/bin/bash
 set -eo pipefail

+
+NON_INTERACTIVE=false
+DO_CLONE=false
+while getopts "ca" arg; do
+    case "$arg" in
+        "a")
+            NON_INTERACTIVE=true
+            ;;
+        "c")
+            DO_CLONE=true
+            ;;
+        *)
+            echo invalid option
+            exit 1
+            ;;
+    esac
+done
+
 GHPAGESREPO="git@github.com:zrepl/zrepl.github.io.git"
 SCRIPTDIR=$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )
 PUBLICDIR="${SCRIPTDIR}/public_git"

 checkout_repo_msg() {
    echo "clone ${GHPAGESREPO} to ${PUBLICDIR}:"
-    echo "	git clone ${GHPAGESREPO} ${PUBLICDIR}"
-    git clone "${GHPAGESREPO}" "${PUBLICDIR}"
 }

 exit_msg() {
@@ -25,11 +41,19 @@ cd "$SCRIPTDIR"

 if [ ! -d "$PUBLICDIR" ]; then
    checkout_repo_msg
-    exit 1
+    if $DO_CLONE; then
+        git clone "${GHPAGESREPO}" "${PUBLICDIR}"
+    else
+        exit 1
+    fi
 fi

-echo -n "PRESS ENTER to confirm you commited and pushed docs changes and tags to the zrepl repo"
-read
+if $NON_INTERACTIVE; then
+    echo "non-interactive mode"
+else
+    echo -n "PRESS ENTER to confirm you commited and pushed docs changes and tags to the zrepl repo"
+    read -r
+fi

 pushd "$PUBLICDIR" 

@@ -52,9 +76,10 @@ popd

 echo "building site"

+flags="$(python3 gen-sphinx-versioning-flags.py)"
 set -e
 sphinx-versioning build \
-   $(python3 gen-sphinx-versioning-flags.py) \
+   $flags \
   docs ./public_git \
   -- -c sphinxconf # older conf.py throw errors because they used
                    # version = subprocess.show_output(["git", "describe"])
@@ -66,7 +91,7 @@ git status --porcelain
 if [[ "$(git status --porcelain)" != "" ]]; then
    CURRENT_COMMIT="${CURRENT_COMMIT}(dirty)" 
 fi
-COMMIT_MSG="sphinx-versioning render from publish.sh - `date -u` - ${CURRENT_COMMIT}"
+COMMIT_MSG="sphinx-versioning render from publish.sh - $(date -u) - ${CURRENT_COMMIT}"

 pushd "$PUBLICDIR"

@@ -25,7 +25,7 @@ A few additional requirements:
 * We want to be able to put off the backups for more than three weeks, i.e., longer than the lifetime of the automatically created snapshots on our workstation.
  **zrepl should use bookmarks and holds to achieve this goal**.
 * When we yank out the drive during replication and go on a long vacation, we do *not* want the partially replicated snapshot to stick around as it would hold on to too much disk space over time.
-  Therefore, we want zrepl to deviate from its :ref:`default step-hold behavior <step-holds-and-bookmarks>` and sacrifice resumability, but nonetheless retain the ability to do incremental replication once we return from our vacation.
+  Therefore, we want zrepl to deviate from its :ref:`default behavior <replication-option-protection>` and sacrifice resumability, but nonetheless retain the ability to do incremental replication once we return from our vacation.
  **zrepl should provide an easy config knob to disable step holds for incremental replication**.

 The following config snippet implements the setup described above.
@@ -51,21 +51,18 @@ To get things going quickly, we skip setting up a CA and generate two self-signe
 For convenience, we generate the key pairs on our local machine and distribute them using ssh:

 .. code-block:: bash
-   :emphasize-lines: 6,13

-   openssl req -x509 -sha256 -nodes \
-      -newkey rsa:4096 \
-      -days 365 \
-      -keyout backups.key \
-      -out backups.crt
-   # ... and use "backups" as Common Name (CN)
+   (name=backups; openssl req -x509 -sha256 -nodes \
+    -newkey rsa:4096 \
+    -days 365 \
+    -keyout $name.key \
+    -out $name.crt -addext "subjectAltName = DNS:$name" -subj "/CN=$name")

-   openssl req -x509 -sha256 -nodes \
-      -newkey rsa:4096 \
-      -days 365 \
-      -keyout prod.key \
-      -out prod.crt
-   # ... and use "prod" as Common Name (CN)
+   (name=prod; openssl req -x509 -sha256 -nodes \
+    -newkey rsa:4096 \
+    -days 365 \
+    -keyout $name.key \
+    -out $name.crt -addext "subjectAltName = DNS:$name" -subj "/CN=$name")

   ssh root@backups "mkdir /etc/zrepl"
   scp  backups.key backups.crt prod.crt root@backups:/etc/zrepl
@@ -94,4 +91,4 @@ We define a corresponding **sink job** named ``sink`` in ``/etc/zrepl/zrepl.yml`
 Go Back To Quickstart Guide
 ---------------------------

-:ref:`Click here <quickstart-apply-config>` to go back to the quickstart guide.
+:ref:`Click here <quickstart-apply-config>` to go back to the quickstart guide.
@@ -32,6 +32,11 @@ We would like to thank the following people and organizations for supporting zre

   <div class="fa fa-code" style="width: 1em;"></div>

+* |supporter-gold| Cyberiada GmbH
+* |supporter-std| `Gordon Schulz <https://github.com/azmodude>`_
+* |supporter-std| `@jwittlincohen <https://github.com/jwittlincohen>`_
+* |supporter-std| `Michael D. Schmitt <https://waterbendingscroll.dancingdragons.org>`_
+* |supporter-std| `Hans Schulz <https://github.com/schulzh>`_
 * |supporter-std| Henning Kessler
 * |supporter-std| `John Ramsden <https://github.com/johnramsden>`_
 * |supporter-std| `DrLuke <https://github.com/drluke>`_
@@ -43,7 +48,7 @@ We would like to thank the following people and organizations for supporting zre
 * |supporter-gold| `MNX.io <https://mnx.io>`_
 * |supporter-std| `Marshall Clyburn <https://github.com/mdclyburn>`_
 * |supporter-code| `Ross Williams <https://github.com/overhacked>`_
-* |supporter-std| Mike T.
+* |supporter-gold| Mike T.
 * |supporter-code| `Justin Scholz <https://github.com/JMoVS>`_
 * |supporter-code| `InsanePrawn <https://github.com/InsanePrawn>`_
 * |supporter-code| `Ben Woods <https://www.freshports.org/sysutils/zrepl/>`_
@@ -10,6 +10,7 @@ import (

 	"github.com/kr/pretty"
 	"github.com/pkg/errors"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"

 	"github.com/zrepl/zrepl/replication/logic/pdu"
@@ -21,10 +22,9 @@ import (
 )

 type SenderConfig struct {
-	FSF                         zfs.DatasetFilter
-	Encrypt                     *zfs.NilBool
-	DisableIncrementalStepHolds bool
-	JobID                       JobID
+	FSF     zfs.DatasetFilter
+	Encrypt *zfs.NilBool
+	JobID   JobID
 }

 func (c *SenderConfig) Validate() error {
@@ -40,10 +40,9 @@ func (c *SenderConfig) Validate() error {

 // Sender implements replication.ReplicationEndpoint for a sending side
 type Sender struct {
-	FSFilter                    zfs.DatasetFilter
-	encrypt                     *zfs.NilBool
-	disableIncrementalStepHolds bool
-	jobId                       JobID
+	FSFilter zfs.DatasetFilter
+	encrypt  *zfs.NilBool
+	jobId    JobID
 }

 func NewSender(conf SenderConfig) *Sender {
@@ -51,10 +50,9 @@ func NewSender(conf SenderConfig) *Sender {
 		panic("invalid config" + err.Error())
 	}
 	return &Sender{
-		FSFilter:                    conf.FSF,
-		encrypt:                     conf.Encrypt,
-		disableIncrementalStepHolds: conf.DisableIncrementalStepHolds,
-		jobId:                       conf.JobID,
+		FSFilter: conf.FSF,
+		encrypt:  conf.Encrypt,
+		jobId:    conf.JobID,
 	}
 }

@@ -211,37 +209,27 @@ func (s *Sender) Send(ctx context.Context, r *pdu.SendReq) (*pdu.SendRes, io.Rea
 		return res, nil, nil
 	}

-	// create a replication cursor for `From` (usually an idempotent no-op because SendCompleted already created it before)
-	var fromReplicationCursor Abstraction
-	if sendArgs.From != nil {
-		// For all but the first replication, this should always be a no-op because SendCompleted already moved the cursor
-		fromReplicationCursor, err = CreateReplicationCursor(ctx, sendArgs.FS, *sendArgs.FromVersion, s.jobId) // no shadow
-		if err == zfs.ErrBookmarkCloningNotSupported {
-			getLogger(ctx).Debug("not creating replication cursor from bookmark because ZFS does not support it")
-			// fallthrough
-		} else if err != nil {
-			return nil, nil, errors.Wrap(err, "cannot set replication cursor to `from` version before starting send")
-		}
+	// create holds or bookmarks of `From` and `To` to guarantee one of the following:
+	// - that the replication step can always be resumed (`holds`),
+	// - that the replication step can be interrupted and a future replication
+	//   step with same or different `To` but same `From` is still possible (`bookmarks`)
+	// - nothing (`none`)
+	//
+	// ...
+	//
+	// ... actually create the abstractions
+	replicationGuaranteeOptions, err := replicationGuaranteeOptionsFromPDU(r.GetReplicationConfig().Protection)
+	if err != nil {
+		return nil, nil, err
 	}
-
-	takeStepHolds := sendArgs.FromVersion == nil || !s.disableIncrementalStepHolds
-
-	var fromHold, toHold Abstraction
-	// make sure `From` doesn't go away in order to make this step resumable
-	if sendArgs.From != nil && takeStepHolds {
-		fromHold, err = HoldStep(ctx, sendArgs.FS, *sendArgs.FromVersion, s.jobId) // no shadow
-		if err == zfs.ErrBookmarkCloningNotSupported {
-			getLogger(ctx).Debug("not creating step bookmark because ZFS does not support it")
-			// fallthrough
-		} else if err != nil {
-			return nil, nil, errors.Wrapf(err, "cannot hold `from` version %q before starting send", *sendArgs.FromVersion)
-		}
+	replicationGuaranteeStrategy := replicationGuaranteeOptions.Strategy(sendArgs.From != nil)
+	liveAbs, err := replicationGuaranteeStrategy.SenderPreSend(ctx, s.jobId, &sendArgs)
+	if err != nil {
+		return nil, nil, err
 	}
-	if takeStepHolds {
-		// make sure `To` doesn't go away in order to make this step resumable
-		toHold, err = HoldStep(ctx, sendArgs.FS, sendArgs.ToVersion, s.jobId)
-		if err != nil {
-			return nil, nil, errors.Wrapf(err, "cannot hold `to` version %q before starting send", sendArgs.ToVersion)
+	for _, a := range liveAbs {
+		if a != nil {
+			abstractionsCacheSingleton.Put(a)
 		}
 	}

@@ -261,7 +249,6 @@ func (s *Sender) Send(ctx context.Context, r *pdu.SendReq) (*pdu.SendRes, io.Rea
 	//
 	// Note further that a resuming send, due to the idempotent nature of func CreateReplicationCursor and HoldStep,
 	// will never lose its step holds because we just (idempotently re-)created them above, before attempting the cleanup.
-	liveAbs := []Abstraction{fromHold, toHold, fromReplicationCursor}
 	func() {
 		ctx, endSpan := trace.WithSpan(ctx, "cleanup-stale-abstractions")
 		defer endSpan()
@@ -283,22 +270,29 @@ func (s *Sender) Send(ctx context.Context, r *pdu.SendReq) (*pdu.SendRes, io.Rea
 			for _, staleVersion := range obsoleteAbs {
 				for _, mustLiveVersion := range mustLiveVersions {
 					isSendArg := zfs.FilesystemVersionEqualIdentity(mustLiveVersion, staleVersion.GetFilesystemVersion())
-					isStepHoldWeMightHaveCreatedWithCurrentValueOf_takeStepHolds :=
-						takeStepHolds && staleVersion.GetType() == AbstractionStepHold
-					if isSendArg && isStepHoldWeMightHaveCreatedWithCurrentValueOf_takeStepHolds {
+					stepHoldBasedGuaranteeStrategy := false
+					k := replicationGuaranteeStrategy.Kind()
+					switch k {
+					case ReplicationGuaranteeKindResumability:
+						stepHoldBasedGuaranteeStrategy = true
+					case ReplicationGuaranteeKindIncremental:
+					case ReplicationGuaranteeKindNone:
+					default:
+						panic(fmt.Sprintf("this is supposed to be an exhaustive match, got %v", k))
+					}
+					isSnapshot := mustLiveVersion.IsSnapshot()
+					if isSendArg && (!isSnapshot || stepHoldBasedGuaranteeStrategy) {
 						panic(fmt.Sprintf("impl error: %q would be destroyed because it is considered stale but it is part of of sendArgs=%s", mustLiveVersion.String(), pretty.Sprint(sendArgs)))
 					}
 				}
 			}
 		}
-		sendAbstractionsCacheSingleton.TryBatchDestroy(ctx, s.jobId, sendArgs.FS, keep, check)
-	}()
-	// now add the newly created abstractions to the cleaned-up cache
-	for _, a := range liveAbs {
-		if a != nil {
-			sendAbstractionsCacheSingleton.Put(a)
+		destroyTypes := AbstractionTypeSet{
+			AbstractionStepHold:                           true,
+			AbstractionTentativeReplicationCursorBookmark: true,
 		}
-	}
+		abstractionsCacheSingleton.TryBatchDestroy(ctx, s.jobId, sendArgs.FS, destroyTypes, keep, check)
+	}()

 	sendStream, err := zfs.ZFSSend(ctx, sendArgs)
 	if err != nil {
@@ -332,36 +326,32 @@ func (p *Sender) SendCompleted(ctx context.Context, r *pdu.SendCompletedReq) (*p
 		return nil, errors.Wrap(err, "validate `to` exists")
 	}

-	log := func(ctx context.Context) Logger {
-		log := getLogger(ctx).WithField("to_guid", to.Guid).
-			WithField("fs", fs).
-			WithField("to", to.RelName)
-		if from != nil {
-			log = log.WithField("from", from.RelName).WithField("from_guid", from.Guid)
-		}
-		return log
-	}
-
-	toReplicationCursor, err := CreateReplicationCursor(ctx, fs, to, p.jobId)
+	replicationGuaranteeOptions, err := replicationGuaranteeOptionsFromPDU(orig.GetReplicationConfig().Protection)
 	if err != nil {
-		if err == zfs.ErrBookmarkCloningNotSupported {
-			log(ctx).Debug("not setting replication cursor, bookmark cloning not supported")
-		} else {
-			msg := "cannot move replication cursor, keeping hold on `to` until successful"
-			log(ctx).WithError(err).Error(msg)
-			err = errors.Wrap(err, msg)
-			// it is correct to not destroy from and to step holds if we can't move the cursor!
-			return &pdu.SendCompletedRes{}, err
+		return nil, err
+	}
+	liveAbs, err := replicationGuaranteeOptions.Strategy(from != nil).SenderPostRecvConfirmed(ctx, p.jobId, fs, to)
+	if err != nil {
+		return nil, err
+	}
+	for _, a := range liveAbs {
+		if a != nil {
+			abstractionsCacheSingleton.Put(a)
 		}
-	} else {
-		sendAbstractionsCacheSingleton.Put(toReplicationCursor)
-		log(ctx).WithField("to_cursor", toReplicationCursor.String()).Info("successfully created `to` replication cursor")
 	}
-
-	keep := func(a Abstraction) bool {
-		return AbstractionEquals(a, toReplicationCursor)
+	keep := func(a Abstraction) (keep bool) {
+		keep = false
+		for _, k := range liveAbs {
+			keep = keep || AbstractionEquals(a, k)
+		}
+		return keep
 	}
-	sendAbstractionsCacheSingleton.TryBatchDestroy(ctx, p.jobId, fs, keep, nil)
+	destroyTypes := AbstractionTypeSet{
+		AbstractionStepHold:                           true,
+		AbstractionTentativeReplicationCursorBookmark: true,
+		AbstractionReplicationCursorBookmarkV2:        true,
+	}
+	abstractionsCacheSingleton.TryBatchDestroy(ctx, p.jobId, fs, destroyTypes, keep, nil)

 	return &pdu.SendCompletedRes{}, nil

@@ -437,8 +427,6 @@ type ReceiverConfig struct {

 	RootWithoutClientComponent *zfs.DatasetPath // TODO use
 	AppendClientIdentity       bool
-
-	UpdateLastReceivedHold bool
 }

 func (c *ReceiverConfig) copyIn() {
@@ -679,7 +667,7 @@ func (s *Receiver) Receive(ctx context.Context, req *pdu.ReceiveReq, receive io.
 		f := zfs.NewDatasetPathForest()
 		f.Add(lp)
 		getLogger(ctx).Debug("begin tree-walk")
-		f.WalkTopDown(func(v zfs.DatasetPathVisit) (visitChildTree bool) {
+		f.WalkTopDown(func(v *zfs.DatasetPathVisit) (visitChildTree bool) {
 			if v.Path.Equal(lp) {
 				return false
 			}
@@ -707,7 +695,7 @@ func (s *Receiver) Receive(ctx context.Context, req *pdu.ReceiveReq, receive io.
 				}
 				l := getLogger(ctx).WithField("placeholder_fs", v.Path)
 				l.Debug("create placeholder filesystem")
-				err := zfs.ZFSCreatePlaceholderFilesystem(ctx, v.Path)
+				err := zfs.ZFSCreatePlaceholderFilesystem(ctx, v.Path, v.Parent.Path)
 				if err != nil {
 					l.WithError(err).Error("cannot create placeholder filesystem")
 					visitErr = err
@@ -863,12 +851,38 @@ func (s *Receiver) Receive(ctx context.Context, req *pdu.ReceiveReq, receive io.
 		return nil, errors.Wrap(err, msg)
 	}

-	if s.conf.UpdateLastReceivedHold {
-		log.Debug("move last-received-hold")
-		if err := MoveLastReceivedHold(ctx, lp.ToString(), toRecvd, s.conf.JobID); err != nil {
-			return nil, errors.Wrap(err, "cannot move last-received-hold")
+	replicationGuaranteeOptions, err := replicationGuaranteeOptionsFromPDU(req.GetReplicationConfig().Protection)
+	if err != nil {
+		return nil, err
+	}
+	replicationGuaranteeStrategy := replicationGuaranteeOptions.Strategy(ph.FSExists)
+	liveAbs, err := replicationGuaranteeStrategy.ReceiverPostRecv(ctx, s.conf.JobID, lp.ToString(), toRecvd)
+	if err != nil {
+		return nil, err
+	}
+	for _, a := range liveAbs {
+		if a != nil {
+			abstractionsCacheSingleton.Put(a)
 		}
 	}
+	keep := func(a Abstraction) (keep bool) {
+		keep = false
+		for _, k := range liveAbs {
+			keep = keep || AbstractionEquals(a, k)
+		}
+		return keep
+	}
+	check := func(obsoleteAbs []Abstraction) {
+		for _, abs := range obsoleteAbs {
+			if zfs.FilesystemVersionEqualIdentity(abs.GetFilesystemVersion(), toRecvd) {
+				panic(fmt.Sprintf("would destroy endpoint abstraction around the filesystem version we just received %s", abs))
+			}
+		}
+	}
+	destroyTypes := AbstractionTypeSet{
+		AbstractionLastReceivedHold: true,
+	}
+	abstractionsCacheSingleton.TryBatchDestroy(ctx, s.conf.JobID, lp.ToString(), destroyTypes, keep, check)

 	return &pdu.ReceiveRes{}, nil
 }
@@ -6,53 +6,54 @@ import (
 	"sync"

 	"github.com/prometheus/client_golang/prometheus"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"
 	"github.com/zrepl/zrepl/util/chainlock"
 )

-var sendAbstractionsCacheMetrics struct {
+var abstractionsCacheMetrics struct {
 	count prometheus.Gauge
 }

 func init() {
-	sendAbstractionsCacheMetrics.count = prometheus.NewGauge(prometheus.GaugeOpts{
+	abstractionsCacheMetrics.count = prometheus.NewGauge(prometheus.GaugeOpts{
 		Namespace: "zrepl",
 		Subsystem: "endpoint",
-		Name:      "send_abstractions_cache_entry_count",
-		Help:      "number of send abstractions tracked in the sendAbstractionsCache data structure",
+		Name:      "abstractions_cache_entry_count",
+		Help:      "number of abstractions tracked in the abstractionsCache data structure",
 	})
 }

-var sendAbstractionsCacheSingleton = newSendAbstractionsCache()
+var abstractionsCacheSingleton = newAbstractionsCache()

-func SendAbstractionsCacheInvalidate(fs string) {
-	sendAbstractionsCacheSingleton.InvalidateFSCache(fs)
+func AbstractionsCacheInvalidate(fs string) {
+	abstractionsCacheSingleton.InvalidateFSCache(fs)
 }

-type sendAbstractionsCacheDidLoadFSState int
+type abstractionsCacheDidLoadFSState int

 const (
-	sendAbstractionsCacheDidLoadFSStateNo sendAbstractionsCacheDidLoadFSState = iota // 0-value has meaning
-	sendAbstractionsCacheDidLoadFSStateInProgress
-	sendAbstractionsCacheDidLoadFSStateDone
+	abstractionsCacheDidLoadFSStateNo abstractionsCacheDidLoadFSState = iota // 0-value has meaning
+	abstractionsCacheDidLoadFSStateInProgress
+	abstractionsCacheDidLoadFSStateDone
 )

-type sendAbstractionsCache struct {
+type abstractionsCache struct {
 	mtx              chainlock.L
 	abstractions     []Abstraction
-	didLoadFS        map[string]sendAbstractionsCacheDidLoadFSState
+	didLoadFS        map[string]abstractionsCacheDidLoadFSState
 	didLoadFSChanged *sync.Cond
 }

-func newSendAbstractionsCache() *sendAbstractionsCache {
-	c := &sendAbstractionsCache{
-		didLoadFS: make(map[string]sendAbstractionsCacheDidLoadFSState),
+func newAbstractionsCache() *abstractionsCache {
+	c := &abstractionsCache{
+		didLoadFS: make(map[string]abstractionsCacheDidLoadFSState),
 	}
 	c.didLoadFSChanged = c.mtx.NewCond()
 	return c
 }

-func (s *sendAbstractionsCache) Put(a Abstraction) {
+func (s *abstractionsCache) Put(a Abstraction) {
 	defer s.mtx.Lock().Unlock()

 	var zeroJobId JobID
@@ -63,10 +64,10 @@ func (s *sendAbstractionsCache) Put(a Abstraction) {
 	}

 	s.abstractions = append(s.abstractions, a)
-	sendAbstractionsCacheMetrics.count.Set(float64(len(s.abstractions)))
+	abstractionsCacheMetrics.count.Set(float64(len(s.abstractions)))
 }

-func (s *sendAbstractionsCache) InvalidateFSCache(fs string) {
+func (s *abstractionsCache) InvalidateFSCache(fs string) {
 	// FIXME: O(n)
 	newAbs := make([]Abstraction, 0, len(s.abstractions))
 	for _, a := range s.abstractions {
@@ -75,9 +76,9 @@ func (s *sendAbstractionsCache) InvalidateFSCache(fs string) {
 		}
 	}
 	s.abstractions = newAbs
-	sendAbstractionsCacheMetrics.count.Set(float64(len(s.abstractions)))
+	abstractionsCacheMetrics.count.Set(float64(len(s.abstractions)))

-	s.didLoadFS[fs] = sendAbstractionsCacheDidLoadFSStateNo
+	s.didLoadFS[fs] = abstractionsCacheDidLoadFSStateNo
 	s.didLoadFSChanged.Broadcast()

 }
@@ -86,7 +87,7 @@ func (s *sendAbstractionsCache) InvalidateFSCache(fs string) {
 // - only fetches on-disk abstractions once, but every time from the in-memory store
 //
 // That means that for precise results, all abstractions created by the endpoint must be .Put into this cache.
-func (s *sendAbstractionsCache) GetAndDeleteByJobIDAndFS(ctx context.Context, jobID JobID, fs string, keep func(a Abstraction) bool) (ret []Abstraction) {
+func (s *abstractionsCache) GetAndDeleteByJobIDAndFS(ctx context.Context, jobID JobID, fs string, types AbstractionTypeSet, keep func(a Abstraction) bool) (ret []Abstraction) {
 	defer s.mtx.Lock().Unlock()
 	defer trace.WithSpanFromStackUpdateCtx(&ctx)()
 	var zeroJobId JobID
@@ -97,50 +98,50 @@ func (s *sendAbstractionsCache) GetAndDeleteByJobIDAndFS(ctx context.Context, jo
 		panic("must not pass zero-value fs")
 	}

-	s.tryLoadOnDiskSendAbstractions(ctx, fs)
+	s.tryLoadOnDiskAbstractions(ctx, fs)

 	// FIXME O(n)
 	var remaining []Abstraction
 	for _, a := range s.abstractions {
 		aJobId := *a.GetJobID()
 		aFS := a.GetFS()
-		if aJobId == jobID && aFS == fs && !keep(a) {
+		if aJobId == jobID && aFS == fs && types[a.GetType()] && !keep(a) {
 			ret = append(ret, a)
 		} else {
 			remaining = append(remaining, a)
 		}
 	}
 	s.abstractions = remaining
-	sendAbstractionsCacheMetrics.count.Set(float64(len(s.abstractions)))
+	abstractionsCacheMetrics.count.Set(float64(len(s.abstractions)))

 	return ret
 }

 // caller must hold s.mtx
-func (s *sendAbstractionsCache) tryLoadOnDiskSendAbstractions(ctx context.Context, fs string) {
-	for s.didLoadFS[fs] != sendAbstractionsCacheDidLoadFSStateDone {
-		if s.didLoadFS[fs] == sendAbstractionsCacheDidLoadFSStateInProgress {
+func (s *abstractionsCache) tryLoadOnDiskAbstractions(ctx context.Context, fs string) {
+	for s.didLoadFS[fs] != abstractionsCacheDidLoadFSStateDone {
+		if s.didLoadFS[fs] == abstractionsCacheDidLoadFSStateInProgress {
 			s.didLoadFSChanged.Wait()
 			continue
 		}
-		if s.didLoadFS[fs] != sendAbstractionsCacheDidLoadFSStateNo {
+		if s.didLoadFS[fs] != abstractionsCacheDidLoadFSStateNo {
 			panic(fmt.Sprintf("unreachable: %v", s.didLoadFS[fs]))
 		}

-		s.didLoadFS[fs] = sendAbstractionsCacheDidLoadFSStateInProgress
+		s.didLoadFS[fs] = abstractionsCacheDidLoadFSStateInProgress
 		defer s.didLoadFSChanged.Broadcast()

 		var onDiskAbs []Abstraction
 		var err error
 		s.mtx.DropWhile(func() {
-			onDiskAbs, err = s.tryLoadOnDiskSendAbstractionsImpl(ctx, fs) // no shadow
+			onDiskAbs, err = s.tryLoadOnDiskAbstractionsImpl(ctx, fs) // no shadow
 		})

 		if err != nil {
-			s.didLoadFS[fs] = sendAbstractionsCacheDidLoadFSStateNo
-			getLogger(ctx).WithField("fs", fs).WithError(err).Error("cannot list send step abstractions for filesystem")
+			s.didLoadFS[fs] = abstractionsCacheDidLoadFSStateNo
+			getLogger(ctx).WithField("fs", fs).WithError(err).Error("cannot list abstractions for filesystem")
 		} else {
-			s.didLoadFS[fs] = sendAbstractionsCacheDidLoadFSStateDone
+			s.didLoadFS[fs] = abstractionsCacheDidLoadFSStateDone
 			s.abstractions = append(s.abstractions, onDiskAbs...)
 			getLogger(ctx).WithField("fs", fs).WithField("abstractions", onDiskAbs).Debug("loaded step abstractions for filesystem")
 		}
@@ -149,7 +150,7 @@ func (s *sendAbstractionsCache) tryLoadOnDiskSendAbstractions(ctx context.Contex
 }

 // caller should _not hold s.mtx
-func (s *sendAbstractionsCache) tryLoadOnDiskSendAbstractionsImpl(ctx context.Context, fs string) ([]Abstraction, error) {
+func (s *abstractionsCache) tryLoadOnDiskAbstractionsImpl(ctx context.Context, fs string) ([]Abstraction, error) {
 	defer trace.WithSpanFromStackUpdateCtx(&ctx)()

 	q := ListZFSHoldsAndBookmarksQuery{
@@ -158,9 +159,10 @@ func (s *sendAbstractionsCache) tryLoadOnDiskSendAbstractionsImpl(ctx context.Co
 		},
 		JobID: nil,
 		What: AbstractionTypeSet{
-			AbstractionStepHold:                    true,
-			AbstractionStepBookmark:                true,
-			AbstractionReplicationCursorBookmarkV2: true,
+			AbstractionStepHold:                           true,
+			AbstractionTentativeReplicationCursorBookmark: true,
+			AbstractionReplicationCursorBookmarkV2:        true,
+			AbstractionLastReceivedHold:                   true,
 		},
 		Concurrency: 1,
 	}
@@ -175,12 +177,12 @@ func (s *sendAbstractionsCache) tryLoadOnDiskSendAbstractionsImpl(ctx context.Co
 	return abs, nil
 }

-func (s *sendAbstractionsCache) TryBatchDestroy(ctx context.Context, jobId JobID, fs string, keep func(a Abstraction) bool, check func(willDestroy []Abstraction)) {
+func (s *abstractionsCache) TryBatchDestroy(ctx context.Context, jobId JobID, fs string, types AbstractionTypeSet, keep func(a Abstraction) bool, check func(willDestroy []Abstraction)) {
 	// no s.mtx, we only use the public interface in this function

 	defer trace.WithSpanFromStackUpdateCtx(&ctx)()

-	obsoleteAbs := s.GetAndDeleteByJobIDAndFS(ctx, jobId, fs, keep)
+	obsoleteAbs := s.GetAndDeleteByJobIDAndFS(ctx, jobId, fs, types, keep)

 	if check != nil {
 		check(obsoleteAbs)
@@ -193,11 +195,11 @@ func (s *sendAbstractionsCache) TryBatchDestroy(ctx context.Context, jobId JobID
 			getLogger(ctx).
 				WithField("abstraction", res.Abstraction).
 				WithError(res.DestroyErr).
-				Error("cannot destroy stale send step abstraction")
+				Error("cannot destroy abstraction")
 		} else {
 			getLogger(ctx).
 				WithField("abstraction", res.Abstraction).
-				Info("destroyed stale send step abstraction")
+				Info("destroyed abstraction")
 		}
 	}
 	if hadErr {
@@ -0,0 +1,214 @@
+package endpoint
+
+import (
+	"context"
+	"fmt"
+
+	"github.com/pkg/errors"
+
+	"github.com/zrepl/zrepl/replication/logic/pdu"
+	"github.com/zrepl/zrepl/zfs"
+)
+
+type ReplicationGuaranteeOptions struct {
+	Initial     ReplicationGuaranteeKind
+	Incremental ReplicationGuaranteeKind
+}
+
+func replicationGuaranteeOptionsFromPDU(in *pdu.ReplicationConfigProtection) (o ReplicationGuaranteeOptions, _ error) {
+	if in == nil {
+		return o, errors.New("pdu.ReplicationConfigProtection must not be nil")
+	}
+	initial, err := replicationGuaranteeKindFromPDU(in.GetInitial())
+	if err != nil {
+		return o, errors.Wrap(err, "pdu.ReplicationConfigProtection: field Initial")
+	}
+	incremental, err := replicationGuaranteeKindFromPDU(in.GetIncremental())
+	if err != nil {
+		return o, errors.Wrap(err, "pdu.ReplicationConfigProtection: field Incremental")
+	}
+	o = ReplicationGuaranteeOptions{
+		Initial:     initial,
+		Incremental: incremental,
+	}
+	return o, nil
+}
+
+func replicationGuaranteeKindFromPDU(in pdu.ReplicationGuaranteeKind) (k ReplicationGuaranteeKind, _ error) {
+	switch in {
+	case pdu.ReplicationGuaranteeKind_GuaranteeNothing:
+		return ReplicationGuaranteeKindNone, nil
+	case pdu.ReplicationGuaranteeKind_GuaranteeIncrementalReplication:
+		return ReplicationGuaranteeKindIncremental, nil
+	case pdu.ReplicationGuaranteeKind_GuaranteeResumability:
+		return ReplicationGuaranteeKindResumability, nil
+
+	case pdu.ReplicationGuaranteeKind_GuaranteeInvalid:
+		fallthrough
+	default:
+		return k, errors.Errorf("%q", in.String())
+	}
+}
+
+func (o ReplicationGuaranteeOptions) Strategy(incremental bool) ReplicationGuaranteeStrategy {
+	g := o.Initial
+	if incremental {
+		g = o.Incremental
+	}
+	return ReplicationGuaranteeFromKind(g)
+}
+
+//go:generate enumer -type=ReplicationGuaranteeKind -json -transform=snake -trimprefix=ReplicationGuaranteeKind
+type ReplicationGuaranteeKind int
+
+const (
+	ReplicationGuaranteeKindResumability ReplicationGuaranteeKind = 1 << iota
+	ReplicationGuaranteeKindIncremental
+	ReplicationGuaranteeKindNone
+)
+
+type ReplicationGuaranteeStrategy interface {
+	Kind() ReplicationGuaranteeKind
+	SenderPreSend(ctx context.Context, jid JobID, sendArgs *zfs.ZFSSendArgsValidated) (keep []Abstraction, err error)
+	ReceiverPostRecv(ctx context.Context, jid JobID, fs string, toRecvd zfs.FilesystemVersion) (keep []Abstraction, err error)
+	SenderPostRecvConfirmed(ctx context.Context, jid JobID, fs string, to zfs.FilesystemVersion) (keep []Abstraction, err error)
+}
+
+func ReplicationGuaranteeFromKind(k ReplicationGuaranteeKind) ReplicationGuaranteeStrategy {
+	switch k {
+	case ReplicationGuaranteeKindNone:
+		return ReplicationGuaranteeNone{}
+	case ReplicationGuaranteeKindIncremental:
+		return ReplicationGuaranteeIncremental{}
+	case ReplicationGuaranteeKindResumability:
+		return ReplicationGuaranteeResumability{}
+	default:
+		panic(fmt.Sprintf("unreachable: %q %T", k, k))
+	}
+}
+
+type ReplicationGuaranteeNone struct{}
+
+func (g ReplicationGuaranteeNone) Kind() ReplicationGuaranteeKind {
+	return ReplicationGuaranteeKindNone
+}
+
+func (g ReplicationGuaranteeNone) SenderPreSend(ctx context.Context, jid JobID, sendArgs *zfs.ZFSSendArgsValidated) (keep []Abstraction, err error) {
+	return nil, nil
+}
+
+func (g ReplicationGuaranteeNone) ReceiverPostRecv(ctx context.Context, jid JobID, fs string, toRecvd zfs.FilesystemVersion) (keep []Abstraction, err error) {
+	return nil, nil
+}
+
+func (g ReplicationGuaranteeNone) SenderPostRecvConfirmed(ctx context.Context, jid JobID, fs string, to zfs.FilesystemVersion) (keep []Abstraction, err error) {
+	return nil, nil
+}
+
+type ReplicationGuaranteeIncremental struct{}
+
+func (g ReplicationGuaranteeIncremental) Kind() ReplicationGuaranteeKind {
+	return ReplicationGuaranteeKindIncremental
+}
+
+func (g ReplicationGuaranteeIncremental) SenderPreSend(ctx context.Context, jid JobID, sendArgs *zfs.ZFSSendArgsValidated) (keep []Abstraction, err error) {
+	if sendArgs.FromVersion != nil {
+		from, err := CreateTentativeReplicationCursor(ctx, sendArgs.FS, *sendArgs.FromVersion, jid)
+		if err != nil {
+			if err == zfs.ErrBookmarkCloningNotSupported {
+				getLogger(ctx).WithField("replication_guarantee", g).
+					WithField("bookmark", sendArgs.From.FullPath(sendArgs.FS)).
+					Info("bookmark cloning is not supported, speculating that `from` will not be destroyed until step is done")
+			} else {
+				return nil, err
+			}
+		}
+		keep = append(keep, from)
+	}
+	to, err := CreateTentativeReplicationCursor(ctx, sendArgs.FS, sendArgs.ToVersion, jid)
+	if err != nil {
+		return nil, err
+	}
+	keep = append(keep, to)
+
+	return keep, nil
+}
+
+func (g ReplicationGuaranteeIncremental) ReceiverPostRecv(ctx context.Context, jid JobID, fs string, toRecvd zfs.FilesystemVersion) (keep []Abstraction, err error) {
+	return receiverPostRecvCommon(ctx, jid, fs, toRecvd)
+}
+
+func (g ReplicationGuaranteeIncremental) SenderPostRecvConfirmed(ctx context.Context, jid JobID, fs string, to zfs.FilesystemVersion) (keep []Abstraction, err error) {
+	return senderPostRecvConfirmedCommon(ctx, jid, fs, to)
+}
+
+type ReplicationGuaranteeResumability struct{}
+
+func (g ReplicationGuaranteeResumability) Kind() ReplicationGuaranteeKind {
+	return ReplicationGuaranteeKindResumability
+}
+
+func (g ReplicationGuaranteeResumability) SenderPreSend(ctx context.Context, jid JobID, sendArgs *zfs.ZFSSendArgsValidated) (keep []Abstraction, err error) {
+	// try to hold the FromVersion
+	if sendArgs.FromVersion != nil {
+		if sendArgs.FromVersion.Type == zfs.Bookmark {
+			getLogger(ctx).WithField("replication_guarantee", g).WithField("fromVersion", sendArgs.FromVersion.FullPath(sendArgs.FS)).
+				Debug("cannot hold a bookmark, speculating that `from` will not be destroyed until step is done")
+		} else {
+			from, err := HoldStep(ctx, sendArgs.FS, *sendArgs.FromVersion, jid)
+			if err != nil {
+				return nil, err
+			}
+			keep = append(keep, from)
+		}
+		// fallthrough
+	}
+
+	to, err := HoldStep(ctx, sendArgs.FS, sendArgs.ToVersion, jid)
+	if err != nil {
+		return nil, err
+	}
+	keep = append(keep, to)
+
+	return keep, nil
+}
+
+func (g ReplicationGuaranteeResumability) ReceiverPostRecv(ctx context.Context, jid JobID, fs string, toRecvd zfs.FilesystemVersion) (keep []Abstraction, err error) {
+	return receiverPostRecvCommon(ctx, jid, fs, toRecvd)
+}
+
+func (g ReplicationGuaranteeResumability) SenderPostRecvConfirmed(ctx context.Context, jid JobID, fs string, to zfs.FilesystemVersion) (keep []Abstraction, err error) {
+	return senderPostRecvConfirmedCommon(ctx, jid, fs, to)
+}
+
+// helper function used by multiple strategies
+func senderPostRecvConfirmedCommon(ctx context.Context, jid JobID, fs string, to zfs.FilesystemVersion) (keep []Abstraction, err error) {
+
+	log := getLogger(ctx).WithField("toVersion", to.FullPath(fs))
+
+	toReplicationCursor, err := CreateReplicationCursor(ctx, fs, to, jid)
+	if err != nil {
+		if err == zfs.ErrBookmarkCloningNotSupported {
+			log.Debug("not setting replication cursor, bookmark cloning not supported")
+		} else {
+			msg := "cannot move replication cursor, keeping hold on `to` until successful"
+			log.WithError(err).Error(msg)
+			err = errors.Wrap(err, msg)
+			return nil, err
+		}
+	} else {
+		log.WithField("to_cursor", toReplicationCursor.String()).Info("successfully created `to` replication cursor")
+	}
+
+	return []Abstraction{toReplicationCursor}, nil
+}
+
+// helper function used by multiple strategies
+func receiverPostRecvCommon(ctx context.Context, jid JobID, fs string, toRecvd zfs.FilesystemVersion) (keep []Abstraction, err error) {
+	getLogger(ctx).Debug("create new last-received-hold")
+	lrh, err := CreateLastReceivedHold(ctx, fs, toRecvd, jid)
+	if err != nil {
+		return nil, err
+	}
+	return []Abstraction{lrh}, nil
+}
@@ -3,5 +3,5 @@ package endpoint
 import "github.com/prometheus/client_golang/prometheus"

 func RegisterMetrics(r prometheus.Registerer) {
-	r.MustRegister(sendAbstractionsCacheMetrics.count)
+	r.MustRegister(abstractionsCacheMetrics.count)
 }
@@ -23,19 +23,19 @@ type AbstractionType string
 // There are a lot of exhaustive switches on AbstractionType in the code base.
 // When adding a new abstraction type, make sure to search and update them!
 const (
-	AbstractionStepBookmark                AbstractionType = "step-bookmark"
-	AbstractionStepHold                    AbstractionType = "step-hold"
-	AbstractionLastReceivedHold            AbstractionType = "last-received-hold"
-	AbstractionReplicationCursorBookmarkV1 AbstractionType = "replication-cursor-bookmark-v1"
-	AbstractionReplicationCursorBookmarkV2 AbstractionType = "replication-cursor-bookmark-v2"
+	AbstractionStepHold                           AbstractionType = "step-hold"
+	AbstractionLastReceivedHold                   AbstractionType = "last-received-hold"
+	AbstractionTentativeReplicationCursorBookmark AbstractionType = "tentative-replication-cursor-bookmark-v2"
+	AbstractionReplicationCursorBookmarkV1        AbstractionType = "replication-cursor-bookmark-v1"
+	AbstractionReplicationCursorBookmarkV2        AbstractionType = "replication-cursor-bookmark-v2"
 )

 var AbstractionTypesAll = map[AbstractionType]bool{
-	AbstractionStepBookmark:                true,
-	AbstractionStepHold:                    true,
-	AbstractionLastReceivedHold:            true,
-	AbstractionReplicationCursorBookmarkV1: true,
-	AbstractionReplicationCursorBookmarkV2: true,
+	AbstractionStepHold:                           true,
+	AbstractionLastReceivedHold:                   true,
+	AbstractionTentativeReplicationCursorBookmark: true,
+	AbstractionReplicationCursorBookmarkV1:        true,
+	AbstractionReplicationCursorBookmarkV2:        true,
 }

 // Implementation Note:
@@ -80,12 +80,12 @@ func AbstractionEquals(a, b Abstraction) bool {

 func (t AbstractionType) Validate() error {
 	switch t {
-	case AbstractionStepBookmark:
-		return nil
 	case AbstractionStepHold:
 		return nil
 	case AbstractionLastReceivedHold:
 		return nil
+	case AbstractionTentativeReplicationCursorBookmark:
+		return nil
 	case AbstractionReplicationCursorBookmarkV1:
 		return nil
 	case AbstractionReplicationCursorBookmarkV2:
@@ -185,8 +185,8 @@ type BookmarkExtractor func(fs *zfs.DatasetPath, v zfs.FilesystemVersion) Abstra
 // returns nil if the abstraction type is not bookmark-based
 func (t AbstractionType) BookmarkExtractor() BookmarkExtractor {
 	switch t {
-	case AbstractionStepBookmark:
-		return StepBookmarkExtractor
+	case AbstractionTentativeReplicationCursorBookmark:
+		return TentativeReplicationCursorExtractor
 	case AbstractionReplicationCursorBookmarkV1:
 		return ReplicationCursorV1Extractor
 	case AbstractionReplicationCursorBookmarkV2:
@@ -205,7 +205,7 @@ type HoldExtractor = func(fs *zfs.DatasetPath, v zfs.FilesystemVersion, tag stri
 // returns nil if the abstraction type is not hold-based
 func (t AbstractionType) HoldExtractor() HoldExtractor {
 	switch t {
-	case AbstractionStepBookmark:
+	case AbstractionTentativeReplicationCursorBookmark:
 		return nil
 	case AbstractionReplicationCursorBookmarkV1:
 		return nil
@@ -220,6 +220,23 @@ func (t AbstractionType) HoldExtractor() HoldExtractor {
 	}
 }

+func (t AbstractionType) BookmarkNamer() func(fs string, guid uint64, jobId JobID) (string, error) {
+	switch t {
+	case AbstractionTentativeReplicationCursorBookmark:
+		return TentativeReplicationCursorBookmarkName
+	case AbstractionReplicationCursorBookmarkV1:
+		panic("shouldn't be creating new ones")
+	case AbstractionReplicationCursorBookmarkV2:
+		return ReplicationCursorBookmarkName
+	case AbstractionStepHold:
+		return nil
+	case AbstractionLastReceivedHold:
+		return nil
+	default:
+		panic(fmt.Sprintf("unimpl: %q", t))
+	}
+}
+
 type ListZFSHoldsAndBookmarksQuery struct {
 	FS ListZFSHoldsAndBookmarksQueryFilesystemFilter
 	// What abstraction types should match (any contained in the set)
@@ -697,11 +714,9 @@ func ListStale(ctx context.Context, q ListZFSHoldsAndBookmarksQuery) (*Staleness
 		return nil, &ListStaleQueryError{errors.New("ListStale cannot have Until != nil set on query")}
 	}

-	// if asking for step holds, must also as for step bookmarks (same kind of abstraction)
-	// as well as replication cursor bookmarks (for firstNotStale)
+	// if asking for step holds must also ask for replication cursor bookmarks (for firstNotStale)
 	ifAnyThenAll := AbstractionTypeSet{
 		AbstractionStepHold:                    true,
-		AbstractionStepBookmark:                true,
 		AbstractionReplicationCursorBookmarkV2: true,
 	}
 	if q.What.ContainsAnyOf(ifAnyThenAll) && !q.What.ContainsAll(ifAnyThenAll) {
@@ -730,7 +745,7 @@ type fsAjobAtype struct {
 }

 // For step holds and bookmarks, only those older than the most recent replication cursor
-// of their (filesystem,job) is considered because younger ones cannot be stale by definition
+// of their (filesystem,job) are considered because younger ones cannot be stale by definition
 // (if we destroy them, we might actually lose the hold on the `To` for an ongoing incremental replication)
 //
 // For replication cursors and last-received-holds, only the most recent one is kept.
@@ -772,8 +787,6 @@ func listStaleFiltering(abs []Abstraction, sinceBound *CreateTXGRangeBound) *Sta
 			}

 		// stepFirstNotStaleCandidate.step
-		case AbstractionStepBookmark:
-			fallthrough
 		case AbstractionStepHold:
 			if c.step == nil || (*c.step).GetCreateTXG() < a.GetCreateTXG() {
 				a := a
@@ -797,7 +810,7 @@ func listStaleFiltering(abs []Abstraction, sinceBound *CreateTXGRangeBound) *Sta
 	for k := range by {
 		l := by[k]

-		if k.Type == AbstractionStepHold || k.Type == AbstractionStepBookmark {
+		if k.Type == AbstractionStepHold {
 			// all older than the most recent cursor are stale, others are always live

 			// if we don't have a replication cursor yet, use untilBound = nil
@@ -0,0 +1,92 @@
+package endpoint
+
+import (
+	"context"
+	"fmt"
+	"regexp"
+
+	"github.com/pkg/errors"
+
+	"github.com/zrepl/zrepl/zfs"
+)
+
+const (
+	LastReceivedHoldTagNamePrefix = "zrepl_last_received_J_"
+)
+
+var lastReceivedHoldTagRE = regexp.MustCompile("^zrepl_last_received_J_(.+)$")
+
+var _ HoldExtractor = LastReceivedHoldExtractor
+
+func LastReceivedHoldExtractor(fs *zfs.DatasetPath, v zfs.FilesystemVersion, holdTag string) Abstraction {
+	var err error
+
+	if v.Type != zfs.Snapshot {
+		panic("impl error")
+	}
+
+	jobID, err := ParseLastReceivedHoldTag(holdTag)
+	if err == nil {
+		return &holdBasedAbstraction{
+			Type:              AbstractionLastReceivedHold,
+			FS:                fs.ToString(),
+			FilesystemVersion: v,
+			Tag:               holdTag,
+			JobID:             jobID,
+		}
+	}
+	return nil
+}
+
+// err != nil always means that the bookmark is not a step bookmark
+func ParseLastReceivedHoldTag(tag string) (JobID, error) {
+	match := lastReceivedHoldTagRE.FindStringSubmatch(tag)
+	if match == nil {
+		return JobID{}, errors.Errorf("parse last-received-hold tag: does not match regex %s", lastReceivedHoldTagRE.String())
+	}
+	jobId, err := MakeJobID(match[1])
+	if err != nil {
+		return JobID{}, errors.Wrap(err, "parse last-received-hold tag: invalid job id field")
+	}
+	return jobId, nil
+}
+
+func LastReceivedHoldTag(jobID JobID) (string, error) {
+	return lastReceivedHoldImpl(jobID.String())
+}
+
+func lastReceivedHoldImpl(jobid string) (string, error) {
+	tag := fmt.Sprintf("%s%s", LastReceivedHoldTagNamePrefix, jobid)
+	if err := zfs.ValidHoldTag(tag); err != nil {
+		return "", err
+	}
+	return tag, nil
+}
+
+func CreateLastReceivedHold(ctx context.Context, fs string, to zfs.FilesystemVersion, jobID JobID) (Abstraction, error) {
+
+	if !to.IsSnapshot() {
+		return nil, errors.Errorf("last-received-hold: target must be a snapshot: %s", to.FullPath(fs))
+	}
+
+	tag, err := LastReceivedHoldTag(jobID)
+	if err != nil {
+		return nil, errors.Wrap(err, "last-received-hold: hold tag")
+	}
+
+	// we never want to be without a hold
+	// => hold new one before releasing old hold
+
+	err = zfs.ZFSHold(ctx, fs, to, tag)
+	if err != nil {
+		return nil, errors.Wrap(err, "last-received-hold: hold newly received")
+	}
+
+	return &holdBasedAbstraction{
+		Type:              AbstractionLastReceivedHold,
+		FS:                fs,
+		FilesystemVersion: to,
+		JobID:             jobID,
+		Tag:               tag,
+	}, nil
+}
@@ -4,12 +4,10 @@ import (
 	"context"
 	"encoding/json"
 	"fmt"
-	"regexp"
 	"sort"

 	"github.com/pkg/errors"

-	"github.com/zrepl/zrepl/util/errorarray"
 	"github.com/zrepl/zrepl/zfs"
 )

@@ -53,6 +51,28 @@ func ParseReplicationCursorBookmarkName(fullname string) (uint64, JobID, error)
 	return guid, jobID, err
 }

+const tentativeReplicationCursorBookmarkNamePrefix = "zrepl_CURSORTENTATIVE_"
+
+// v must be validated by caller
+func TentativeReplicationCursorBookmarkName(fs string, guid uint64, id JobID) (string, error) {
+	return tentativeReplicationCursorBookmarkNameImpl(fs, guid, id.String())
+}
+
+func tentativeReplicationCursorBookmarkNameImpl(fs string, guid uint64, jobid string) (string, error) {
+	return makeJobAndGuidBookmarkName(tentativeReplicationCursorBookmarkNamePrefix, fs, guid, jobid)
+}
+
+// name is the full bookmark name, including dataset path
+//
+// err != nil always means that the bookmark is not a step bookmark
+func ParseTentativeReplicationCursorBookmarkName(fullname string) (guid uint64, jobID JobID, err error) {
+	guid, jobID, err = parseJobAndGuidBookmarkName(fullname, tentativeReplicationCursorBookmarkNamePrefix)
+	if err != nil {
+		err = errors.Wrap(err, "parse step bookmark name") // no shadow!
+	}
+	return guid, jobID, err
+}
+
 // may return nil for both values, indicating there is no cursor
 func GetMostRecentReplicationCursorOfJob(ctx context.Context, fs string, jobID JobID) (*zfs.FilesystemVersion, error) {
 	fsp, err := zfs.NewDatasetPath(fs)
@@ -122,23 +142,23 @@ func GetReplicationCursors(ctx context.Context, dp *zfs.DatasetPath, jobID JobID
 //
 // returns ErrBookmarkCloningNotSupported if version is a bookmark and bookmarking bookmarks is not supported by ZFS
 func CreateReplicationCursor(ctx context.Context, fs string, target zfs.FilesystemVersion, jobID JobID) (a Abstraction, err error) {
+	return createBookmarkAbstraction(ctx, AbstractionReplicationCursorBookmarkV2, fs, target, jobID)
+}

-	bookmarkname, err := ReplicationCursorBookmarkName(fs, target.GetGuid(), jobID)
+func CreateTentativeReplicationCursor(ctx context.Context, fs string, target zfs.FilesystemVersion, jobID JobID) (a Abstraction, err error) {
+	return createBookmarkAbstraction(ctx, AbstractionTentativeReplicationCursorBookmark, fs, target, jobID)
+}
+
+func createBookmarkAbstraction(ctx context.Context, abstractionType AbstractionType, fs string, target zfs.FilesystemVersion, jobID JobID) (a Abstraction, err error) {
+
+	bookmarkNamer := abstractionType.BookmarkNamer()
+	if bookmarkNamer == nil {
+		panic(abstractionType)
+	}
+
+	bookmarkname, err := bookmarkNamer(fs, target.GetGuid(), jobID)
 	if err != nil {
-		return nil, errors.Wrap(err, "determine replication cursor name")
-	}
-
-	if target.IsBookmark() && target.GetName() == bookmarkname {
-		return &bookmarkBasedAbstraction{
-			Type:              AbstractionReplicationCursorBookmarkV2,
-			FS:                fs,
-			FilesystemVersion: target,
-			JobID:             jobID,
-		}, nil
-	}
-
-	if !target.IsSnapshot() {
-		return nil, zfs.ErrBookmarkCloningNotSupported
+		return nil, errors.Wrapf(err, "determine %s name", abstractionType)
 	}

 	// idempotently create bookmark (guid is encoded in it)
@@ -152,125 +172,13 @@ func CreateReplicationCursor(ctx context.Context, fs string, target zfs.Filesyst
 	}

 	return &bookmarkBasedAbstraction{
-		Type:              AbstractionReplicationCursorBookmarkV2,
+		Type:              abstractionType,
 		FS:                fs,
 		FilesystemVersion: cursorBookmark,
 		JobID:             jobID,
 	}, nil
 }

-const (
-	ReplicationCursorBookmarkNamePrefix = "zrepl_last_received_J_"
-)
-
-var lastReceivedHoldTagRE = regexp.MustCompile("^zrepl_last_received_J_(.+)$")
-
-// err != nil always means that the bookmark is not a step bookmark
-func ParseLastReceivedHoldTag(tag string) (JobID, error) {
-	match := lastReceivedHoldTagRE.FindStringSubmatch(tag)
-	if match == nil {
-		return JobID{}, errors.Errorf("parse last-received-hold tag: does not match regex %s", lastReceivedHoldTagRE.String())
-	}
-	jobId, err := MakeJobID(match[1])
-	if err != nil {
-		return JobID{}, errors.Wrap(err, "parse last-received-hold tag: invalid job id field")
-	}
-	return jobId, nil
-}
-
-func LastReceivedHoldTag(jobID JobID) (string, error) {
-	return lastReceivedHoldImpl(jobID.String())
-}
-
-func lastReceivedHoldImpl(jobid string) (string, error) {
-	tag := fmt.Sprintf("%s%s", ReplicationCursorBookmarkNamePrefix, jobid)
-	if err := zfs.ValidHoldTag(tag); err != nil {
-		return "", err
-	}
-	return tag, nil
-}
-
-func CreateLastReceivedHold(ctx context.Context, fs string, to zfs.FilesystemVersion, jobID JobID) (Abstraction, error) {
-
-	if !to.IsSnapshot() {
-		return nil, errors.Errorf("last-received-hold: target must be a snapshot: %s", to.FullPath(fs))
-	}
-
-	tag, err := LastReceivedHoldTag(jobID)
-	if err != nil {
-		return nil, errors.Wrap(err, "last-received-hold: hold tag")
-	}
-
-	// we never want to be without a hold
-	// => hold new one before releasing old hold
-
-	err = zfs.ZFSHold(ctx, fs, to, tag)
-	if err != nil {
-		return nil, errors.Wrap(err, "last-received-hold: hold newly received")
-	}
-
-	return &holdBasedAbstraction{
-		Type:              AbstractionLastReceivedHold,
-		FS:                fs,
-		FilesystemVersion: to,
-		JobID:             jobID,
-		Tag:               tag,
-	}, nil
-}
-
-func MoveLastReceivedHold(ctx context.Context, fs string, to zfs.FilesystemVersion, jobID JobID) error {
-
-	_, err := CreateLastReceivedHold(ctx, fs, to, jobID)
-	if err != nil {
-		return err
-	}
-
-	q := ListZFSHoldsAndBookmarksQuery{
-		What: AbstractionTypeSet{
-			AbstractionLastReceivedHold: true,
-		},
-		FS: ListZFSHoldsAndBookmarksQueryFilesystemFilter{
-			FS: &fs,
-		},
-		JobID: &jobID,
-		CreateTXG: CreateTXGRange{
-			Since: nil,
-			Until: &CreateTXGRangeBound{
-				CreateTXG: to.GetCreateTXG(),
-				Inclusive: &zfs.NilBool{B: false},
-			},
-		},
-		Concurrency: 1,
-	}
-	abs, absErrs, err := ListAbstractions(ctx, q)
-	if err != nil {
-		return errors.Wrap(err, "last-received-hold: list")
-	}
-	if len(absErrs) > 0 {
-		return errors.Wrap(ListAbstractionsErrors(absErrs), "last-received-hold: list")
-	}
-
-	getLogger(ctx).WithField("last-received-holds", fmt.Sprintf("%s", abs)).Debug("releasing last-received-holds")
-
-	var errs []error
-	for res := range BatchDestroy(ctx, abs) {
-		log := getLogger(ctx).
-			WithField("last-received-hold", res.Abstraction)
-		if res.DestroyErr != nil {
-			errs = append(errs, res.DestroyErr)
-			log.WithError(err).
-				Error("cannot release last-received-hold")
-		} else {
-			log.Info("released last-received-hold")
-		}
-	}
-	if len(errs) == 0 {
-		return nil
-	} else {
-		return errorarray.Wrap(errs, "last-received-hold: release")
-	}
-}
-
 func ReplicationCursorV2Extractor(fs *zfs.DatasetPath, v zfs.FilesystemVersion) (_ Abstraction) {
 	if v.Type != zfs.Bookmark {
 		panic("impl error")
@@ -308,24 +216,28 @@ func ReplicationCursorV1Extractor(fs *zfs.DatasetPath, v zfs.FilesystemVersion)
 	return nil
 }

-var _ HoldExtractor = LastReceivedHoldExtractor
+var _ BookmarkExtractor = TentativeReplicationCursorExtractor

-func LastReceivedHoldExtractor(fs *zfs.DatasetPath, v zfs.FilesystemVersion, holdTag string) Abstraction {
-	var err error
-
-	if v.Type != zfs.Snapshot {
+func TentativeReplicationCursorExtractor(fs *zfs.DatasetPath, v zfs.FilesystemVersion) (_ Abstraction) {
+	if v.Type != zfs.Bookmark {
 		panic("impl error")
 	}

-	jobID, err := ParseLastReceivedHoldTag(holdTag)
+	fullname := v.ToAbsPath(fs)
+
+	guid, jobid, err := ParseTentativeReplicationCursorBookmarkName(fullname)
+	if guid != v.Guid {
+		// TODO log this possibly tinkered-with bookmark
+		return nil
+	}
 	if err == nil {
-		return &holdBasedAbstraction{
-			Type:              AbstractionLastReceivedHold,
+		bm := &bookmarkBasedAbstraction{
+			Type:              AbstractionTentativeReplicationCursorBookmark,
 			FS:                fs.ToString(),
 			FilesystemVersion: v,
-			Tag:               holdTag,
-			JobID:             jobID,
+			JobID:             jobid,
 		}
+		return bm
 	}
 	return nil
 }
@@ -1,159 +0,0 @@
-package endpoint
-
-import (
-	"context"
-	"fmt"
-	"regexp"
-
-	"github.com/pkg/errors"
-
-	"github.com/zrepl/zrepl/zfs"
-)
-
-var stepHoldTagRE = regexp.MustCompile("^zrepl_STEP_J_(.+)")
-
-func StepHoldTag(jobid JobID) (string, error) {
-	return stepHoldTagImpl(jobid.String())
-}
-
-func stepHoldTagImpl(jobid string) (string, error) {
-	t := fmt.Sprintf("zrepl_STEP_J_%s", jobid)
-	if err := zfs.ValidHoldTag(t); err != nil {
-		return "", err
-	}
-	return t, nil
-}
-
-// err != nil always means that the bookmark is not a step bookmark
-func ParseStepHoldTag(tag string) (JobID, error) {
-	match := stepHoldTagRE.FindStringSubmatch(tag)
-	if match == nil {
-		return JobID{}, fmt.Errorf("parse hold tag: match regex %q", stepHoldTagRE)
-	}
-	jobID, err := MakeJobID(match[1])
-	if err != nil {
-		return JobID{}, errors.Wrap(err, "parse hold tag: invalid job id field")
-	}
-	return jobID, nil
-}
-
-const stepBookmarkNamePrefix = "zrepl_STEP"
-
-// v must be validated by caller
-func StepBookmarkName(fs string, guid uint64, id JobID) (string, error) {
-	return stepBookmarkNameImpl(fs, guid, id.String())
-}
-
-func stepBookmarkNameImpl(fs string, guid uint64, jobid string) (string, error) {
-	return makeJobAndGuidBookmarkName(stepBookmarkNamePrefix, fs, guid, jobid)
-}
-
-// name is the full bookmark name, including dataset path
-//
-// err != nil always means that the bookmark is not a step bookmark
-func ParseStepBookmarkName(fullname string) (guid uint64, jobID JobID, err error) {
-	guid, jobID, err = parseJobAndGuidBookmarkName(fullname, stepBookmarkNamePrefix)
-	if err != nil {
-		err = errors.Wrap(err, "parse step bookmark name") // no shadow!
-	}
-	return guid, jobID, err
-}
-
-// idempotently hold / step-bookmark `version`
-//
-// returns ErrBookmarkCloningNotSupported if version is a bookmark and bookmarking bookmarks is not supported by ZFS
-func HoldStep(ctx context.Context, fs string, v zfs.FilesystemVersion, jobID JobID) (Abstraction, error) {
-	if v.IsSnapshot() {
-
-		tag, err := StepHoldTag(jobID)
-		if err != nil {
-			return nil, errors.Wrap(err, "step hold tag")
-		}
-
-		if err := zfs.ZFSHold(ctx, fs, v, tag); err != nil {
-			return nil, errors.Wrap(err, "step hold: zfs")
-		}
-
-		return &holdBasedAbstraction{
-			Type:              AbstractionStepHold,
-			FS:                fs,
-			Tag:               tag,
-			JobID:             jobID,
-			FilesystemVersion: v,
-		}, nil
-	}
-
-	if !v.IsBookmark() {
-		panic(fmt.Sprintf("version must bei either snapshot or bookmark, got %#v", v))
-	}
-
-	bmname, err := StepBookmarkName(fs, v.Guid, jobID)
-	if err != nil {
-		return nil, errors.Wrap(err, "create step bookmark: determine bookmark name")
-	}
-	// idempotently create bookmark
-	stepBookmark, err := zfs.ZFSBookmark(ctx, fs, v, bmname)
-	if err != nil {
-		if err == zfs.ErrBookmarkCloningNotSupported {
-			// TODO we could actually try to find a local snapshot that has the requested GUID
-			// 		however, the replication algorithm prefers snapshots anyways, so this quest
-			// 		is most likely not going to be successful. Also, there's the possibility that
-			//      the caller might want to filter what snapshots are eligibile, and this would
-			//      complicate things even further.
-			return nil, err // TODO go1.13 use wrapping
-		}
-		return nil, errors.Wrap(err, "create step bookmark: zfs")
-	}
-	return &bookmarkBasedAbstraction{
-		Type:              AbstractionStepBookmark,
-		FS:                fs,
-		FilesystemVersion: stepBookmark,
-		JobID:             jobID,
-	}, nil
-}
-
-var _ BookmarkExtractor = StepBookmarkExtractor
-
-func StepBookmarkExtractor(fs *zfs.DatasetPath, v zfs.FilesystemVersion) (_ Abstraction) {
-	if v.Type != zfs.Bookmark {
-		panic("impl error")
-	}
-
-	fullname := v.ToAbsPath(fs)
-
-	guid, jobid, err := ParseStepBookmarkName(fullname)
-	if guid != v.Guid {
-		// TODO log this possibly tinkered-with bookmark
-		return nil
-	}
-	if err == nil {
-		bm := &bookmarkBasedAbstraction{
-			Type:              AbstractionStepBookmark,
-			FS:                fs.ToString(),
-			FilesystemVersion: v,
-			JobID:             jobid,
-		}
-		return bm
-	}
-	return nil
-}
-
-var _ HoldExtractor = StepHoldExtractor
-
-func StepHoldExtractor(fs *zfs.DatasetPath, v zfs.FilesystemVersion, holdTag string) Abstraction {
-	if v.Type != zfs.Snapshot {
-		panic("impl error")
-	}
-
-	jobID, err := ParseStepHoldTag(holdTag)
-	if err == nil {
-		return &holdBasedAbstraction{
-			Type:              AbstractionStepHold,
-			FS:                fs.ToString(),
-			Tag:               holdTag,
-			FilesystemVersion: v,
-			JobID:             jobID,
-		}
-	}
-	return nil
-}
@@ -0,0 +1,83 @@
+package endpoint
+
+import (
+	"context"
+	"fmt"
+	"regexp"
+
+	"github.com/pkg/errors"
+
+	"github.com/zrepl/zrepl/zfs"
+)
+
+var stepHoldTagRE = regexp.MustCompile("^zrepl_STEP_J_(.+)")
+
+func StepHoldTag(jobid JobID) (string, error) {
+	return stepHoldTagImpl(jobid.String())
+}
+
+func stepHoldTagImpl(jobid string) (string, error) {
+	t := fmt.Sprintf("zrepl_STEP_J_%s", jobid)
+	if err := zfs.ValidHoldTag(t); err != nil {
+		return "", err
+	}
+	return t, nil
+}
+
+// err != nil always means that the bookmark is not a step bookmark
+func ParseStepHoldTag(tag string) (JobID, error) {
+	match := stepHoldTagRE.FindStringSubmatch(tag)
+	if match == nil {
+		return JobID{}, fmt.Errorf("parse hold tag: match regex %q", stepHoldTagRE)
+	}
+	jobID, err := MakeJobID(match[1])
+	if err != nil {
+		return JobID{}, errors.Wrap(err, "parse hold tag: invalid job id field")
+	}
+	return jobID, nil
+}
+
+// idempotently hold `version`
+func HoldStep(ctx context.Context, fs string, v zfs.FilesystemVersion, jobID JobID) (Abstraction, error) {
+	if !v.IsSnapshot() {
+		panic(fmt.Sprintf("version must be a snapshot got %#v", v))
+	}
+
+	tag, err := StepHoldTag(jobID)
+	if err != nil {
+		return nil, errors.Wrap(err, "step hold tag")
+	}
+
+	if err := zfs.ZFSHold(ctx, fs, v, tag); err != nil {
+		return nil, errors.Wrap(err, "step hold: zfs")
+	}
+
+	return &holdBasedAbstraction{
+		Type:              AbstractionStepHold,
+		FS:                fs,
+		Tag:               tag,
+		JobID:             jobID,
+		FilesystemVersion: v,
+	}, nil
+
+}
+
+var _ HoldExtractor = StepHoldExtractor
+
+func StepHoldExtractor(fs *zfs.DatasetPath, v zfs.FilesystemVersion, holdTag string) Abstraction {
+	if v.Type != zfs.Snapshot {
+		panic("impl error")
+	}
+
+	jobID, err := ParseStepHoldTag(holdTag)
+	if err == nil {
+		return &holdBasedAbstraction{
+			Type:              AbstractionStepHold,
+			FS:                fs.ToString(),
+			Tag:               holdTag,
+			FilesystemVersion: v,
+			JobID:             jobID,
+		}
+	}
+	return nil
+}
@@ -24,9 +24,9 @@ func MakeJobID(s string) (JobID, error) {
 		return JobID{}, errors.Wrap(err, "must be usable as a dataset path component")
 	}

-	if _, err := stepBookmarkNameImpl("pool/ds", 0xface601d, s); err != nil {
+	if _, err := tentativeReplicationCursorBookmarkNameImpl("pool/ds", 0xface601d, s); err != nil {
 		// note that this might still fail due to total maximum name length, but we can't enforce that
-		return JobID{}, errors.Wrap(err, "must be usable for a step bookmark")
+		return JobID{}, errors.Wrap(err, "must be usable for a tentative replication cursor bookmark")
 	}

 	if _, err := stepHoldTagImpl(s); err != nil {
@@ -0,0 +1,80 @@
+// Code generated by "enumer -type=ReplicationGuaranteeKind -json -transform=snake -trimprefix=ReplicationGuaranteeKind"; DO NOT EDIT.
+
+//
+package endpoint
+
+import (
+	"encoding/json"
+	"fmt"
+)
+
+const (
+	_ReplicationGuaranteeKindName_0 = "resumabilityincremental"
+	_ReplicationGuaranteeKindName_1 = "none"
+)
+
+var (
+	_ReplicationGuaranteeKindIndex_0 = [...]uint8{0, 12, 23}
+	_ReplicationGuaranteeKindIndex_1 = [...]uint8{0, 4}
+)
+
+func (i ReplicationGuaranteeKind) String() string {
+	switch {
+	case 1 <= i && i <= 2:
+		i -= 1
+		return _ReplicationGuaranteeKindName_0[_ReplicationGuaranteeKindIndex_0[i]:_ReplicationGuaranteeKindIndex_0[i+1]]
+	case i == 4:
+		return _ReplicationGuaranteeKindName_1
+	default:
+		return fmt.Sprintf("ReplicationGuaranteeKind(%d)", i)
+	}
+}
+
+var _ReplicationGuaranteeKindValues = []ReplicationGuaranteeKind{1, 2, 4}
+
+var _ReplicationGuaranteeKindNameToValueMap = map[string]ReplicationGuaranteeKind{
+	_ReplicationGuaranteeKindName_0[0:12]:  1,
+	_ReplicationGuaranteeKindName_0[12:23]: 2,
+	_ReplicationGuaranteeKindName_1[0:4]:   4,
+}
+
+// ReplicationGuaranteeKindString retrieves an enum value from the enum constants string name.
+// Throws an error if the param is not part of the enum.
+func ReplicationGuaranteeKindString(s string) (ReplicationGuaranteeKind, error) {
+	if val, ok := _ReplicationGuaranteeKindNameToValueMap[s]; ok {
+		return val, nil
+	}
+	return 0, fmt.Errorf("%s does not belong to ReplicationGuaranteeKind values", s)
+}
+
+// ReplicationGuaranteeKindValues returns all values of the enum
+func ReplicationGuaranteeKindValues() []ReplicationGuaranteeKind {
+	return _ReplicationGuaranteeKindValues
+}
+
+// IsAReplicationGuaranteeKind returns "true" if the value is listed in the enum definition. "false" otherwise
+func (i ReplicationGuaranteeKind) IsAReplicationGuaranteeKind() bool {
+	for _, v := range _ReplicationGuaranteeKindValues {
+		if i == v {
+			return true
+		}
+	}
+	return false
+}
+
+// MarshalJSON implements the json.Marshaler interface for ReplicationGuaranteeKind
+func (i ReplicationGuaranteeKind) MarshalJSON() ([]byte, error) {
+	return json.Marshal(i.String())
+}
+
+// UnmarshalJSON implements the json.Unmarshaler interface for ReplicationGuaranteeKind
+func (i *ReplicationGuaranteeKind) UnmarshalJSON(data []byte) error {
+	var s string
+	if err := json.Unmarshal(data, &s); err != nil {
+		return fmt.Errorf("ReplicationGuaranteeKind should be a string, got %s", data)
+	}
+
+	var err error
+	*i, err = ReplicationGuaranteeKindString(s)
+	return err
+}
@@ -10,6 +10,7 @@ require (
 	github.com/go-sql-driver/mysql v1.4.1-0.20190907122137-b2c03bcae3d4
 	github.com/golang/protobuf v1.3.2
 	github.com/google/uuid v1.1.1
+	github.com/hashicorp/yamux v0.0.0-20200609203250-aecfd211c9ce
 	github.com/jinzhu/copier v0.0.0-20170922082739-db4671f3a9b8
 	github.com/kr/pretty v0.1.0
 	github.com/lib/pq v1.2.0
@@ -33,6 +34,7 @@ require (
 	github.com/yudai/gojsondiff v0.0.0-20170107030110-7b1b7adf999d
 	github.com/yudai/golcs v0.0.0-20170316035057-ecda9a501e82 // go1.12 thinks it needs this
 	github.com/zrepl/yaml-config v0.0.0-20191220194647-cbb6b0cf4bdd
+	golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2
 	golang.org/x/net v0.0.0-20190613194153-d28f0bde5980
 	golang.org/x/sync v0.0.0-20190423024810-112230192c58
 	golang.org/x/sys v0.0.0-20191026070338-33540a1f6037
@@ -40,3 +42,5 @@ require (
 	gonum.org/v1/gonum v0.7.0 // indirect
 	google.golang.org/grpc v1.17.0
 )
+
+replace github.com/problame/go-netssh => /home/cs/zrepl/go-netssh
@@ -120,6 +120,8 @@ github.com/google/uuid v1.1.1/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+
 github.com/gostaticanalysis/analysisutil v0.0.0-20190318220348-4088753ea4d3/go.mod h1:eEOZF4jCKGi+aprrirO9e7WKB3beBRtWgqGunKl6pKE=
 github.com/hashicorp/hcl v0.0.0-20180404174102-ef8a98b0bbce/go.mod h1:oZtUIOe8dh44I2q6ScRibXws4Ajl+d+nod3AaR9vL5w=
 github.com/hashicorp/hcl v1.0.0/go.mod h1:E5yfLk+7swimpb2L/Alb/PJmXilQ/rhwaUYs4T20WEQ=
+github.com/hashicorp/yamux v0.0.0-20200609203250-aecfd211c9ce h1:7UnVY3T/ZnHUrfviiAgIUjg2PXxsQfs5bphsG8F7Keo=
+github.com/hashicorp/yamux v0.0.0-20200609203250-aecfd211c9ce/go.mod h1:+NfK9FKeTrX5uv1uIXGdwYDTeHna2qgaIlx54MXqjAM=
 github.com/hpcloud/tail v1.0.0 h1:nfCOvKYfkgYP8hkirhJocXT2+zOD8yUNjXaWfTlyFKI=
 github.com/hpcloud/tail v1.0.0/go.mod h1:ab1qPbhIpdTxEkNHXyeSf5vhxWSCs/tWer42PpOxQnU=
 github.com/inconshreveable/mousetrap v1.0.0 h1:Z8tu5sraLXCXIcARxBp/8cbvlwVa7Z1NHg9XEKhtSvM=
@@ -300,6 +302,7 @@ github.com/zrepl/yaml-config v0.0.0-20191220194647-cbb6b0cf4bdd/go.mod h1:JmNwis
 github.com/zrepl/zrepl v0.2.0/go.mod h1:M3Zv2IGSO8iYpUjsZD6ayZ2LHy7zyMfzet9XatKOrZ8=
 golang.org/x/crypto v0.0.0-20180904163835-0709b304e793/go.mod h1:6SG95UA2DQfeDnfUPMdvaQW0Q7yPrPDi9nlGo2tz2b4=
 golang.org/x/crypto v0.0.0-20181203042331-505ab145d0a9/go.mod h1:6SG95UA2DQfeDnfUPMdvaQW0Q7yPrPDi9nlGo2tz2b4=
+golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2 h1:VklqNMn3ovrHsnt90PveolxSbWFaJdECFbxSq0Mqo2M=
 golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
 golang.org/x/crypto v0.0.0-20190313024323-a1f597ede03a/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
 golang.org/x/crypto v0.0.0-20190325154230-a5d413f7728c/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
@@ -1,4 +1,4 @@
-#!/bin/bash
+#!/usr/bin/env bash
 set -e


@@ -14,23 +14,31 @@ step() {
    echo "${bold}$1${normal}"
 }

-if ! type go >/dev/null; then
-    step "go binary not installed or not in \$PATH" 1>&2
-    exit 1
-fi
-
-if [ -z "$GOPATH" ]; then
-    step "Make sure you have your GOPATH configured correctly" 1>&2
-    exit 1
-fi
-
-CHECKOUTPATH="${GOPATH}/src/github.com/zrepl/zrepl"
-
 godep() {
    step "install build dependencies (versions pinned in build/go.mod and build/tools.go)"
+
+    if ! type go >/dev/null; then
+        step "go binary not installed or not in \$PATH" 1>&2
+        exit 1
+    fi
+
+    if [ -z "$GOPATH" ]; then
+        step "Your GOPATH is not configured correctly" 1>&2
+        exit 1
+    fi
+
+    if ! (echo "$PATH" | grep "${GOPATH}/bin" > /dev/null); then
+        step "GOPATH/bin is not in your PATH (it should be towards the start of it)"
+        exit 1
+    fi
+
    pushd "$(dirname "${BASH_SOURCE[0]}")"/build
    set -x
    export GO111MODULE=on # otherwise, a checkout of this repo in GOPATH will disable modules on Go 1.12 and earlier
+    source <(go env)
+    export GOOS="$GOHOSTOS"
+    export GOARCH="$GOHOSTARCH"
+    # TODO GOARM=$GOHOSTARM?
    go build -v -mod=readonly -o "$GOPATH/bin/stringer"      golang.org/x/tools/cmd/stringer
    go build -v -mod=readonly -o "$GOPATH/bin/protoc-gen-go" github.com/golang/protobuf/protoc-gen-go
    go build -v -mod=readonly -o "$GOPATH/bin/enumer"        github.com/alvaroloes/enumer
@@ -51,8 +59,9 @@ docdep() {
        exit 1
    fi
    step "Installing doc build dependencies"
-    local reqpath="${CHECKOUTPATH}/docs/requirements.txt"
-    if [ ! -z "$ZREPL_LAZY_DOCS_REQPATH" ]; then
+    # shellcheck disable=SC2155
+    local reqpath="$(dirname "${BASH_SOURCE[0]}")/docs/requirements.txt"
+    if [ -n "$ZREPL_LAZY_DOCS_REQPATH" ]; then
        reqpath="$ZREPL_LAZY_DOCS_REQPATH"
    fi
    pip3 install -r "$reqpath"
@@ -63,9 +72,15 @@ release() {
    make release
 }

+# shellcheck disable=SC2198
+if [ -z "$@" ]; then
+    step "No command specified, exiting"
+    exit 1
+fi
+
 for cmd in "$@"; do
    case "$cmd" in
-        godep|docdep|release_bins|docs)
+        godep|docdep|release|docs)
            eval $cmd
            continue
            ;;
@@ -82,7 +82,7 @@ func (l Level) String() string {
 	case Error:
 		return "error"
 	default:
-		return string(l)
+		return fmt.Sprintf("unknown level %d", l)
 	}
 }

@@ -0,0 +1,11 @@
+FROM debian:latest
+
+RUN apt-get update && apt-get install -y \
+	build-essential \
+    devscripts \
+    dh-exec
+
+RUN mkdir -p /build/src && chmod -R 0777 /build
+
+WORKDIR /build/src
+
@@ -0,0 +1,5 @@
+zrepl (VERSION) unstable;
+
+  Check https://zrepl.github.io/changelog.html.
+
+ -- zrepl build process <me@cschwarz.com>  DATE_DASH_R_OUTPUT
@@ -0,0 +1 @@
+11
@@ -0,0 +1,8 @@
+Source: zrepl
+Maintainer: Christian Schwarz <me@cschwarz.com>
+Build-Depends: dh-exec
+
+Package: zrepl
+Architecture: arm64 amd64 armhf i386
+Description: ZFS replication
+  zrepl is a one-stop solution for ZFS filesystem replication
@@ -0,0 +1,50 @@
+#!/usr/bin/make -f
+
+%:
+	dh $@
+
+override_dh_strip:
+
+override_dh_auto_build:
+
+override_dh_auto_clean:
+
+	rm -rf debian/renamedir
+
+override_dh_auto_install:
+
+	# install the zrepl.service for
+	# dh_install_systemd and dh_systemd_{start,enable}
+	install dist/systemd/zrepl.service debian/zrepl.service
+	sed -i 's#ExecStart=/usr/local/bin/zrepl #ExecStart=/usr/bin/zrepl #' debian/zrepl.service
+	sed -i 's#ExecStartPre=/usr/local/bin/zrepl #ExecStartPre=/usr/bin/zrepl #' debian/zrepl.service
+
+	mkdir -p debian/renamedir
+
+	# install binary
+	stat artifacts/$(ZREPL_DPKG_ZREPL_BINARY_FILENAME)
+	cp --preserve=all artifacts/$(ZREPL_DPKG_ZREPL_BINARY_FILENAME) debian/renamedir/zrepl
+	dh_install debian/renamedir/zrepl usr/bin
+
+	# install bash completion
+	dh_install artifacts/bash_completion etc/bash_completion.d/zrepl
+	dh_install artifacts/_zrepl.zsh_completion usr/share/zsh/vendor-completions
+
+	# install docs
+	dh_install artifacts/docs/html usr/share/doc/zrepl/docs/
+
+	# install examples
+	dh_install config/samples/* usr/share/doc/zrepl/examples
+
+	# install default config
+	mkdir -p debian/tmp/etc/zrepl
+	chmod 0700 debian/tmp/etc/zrepl
+	sed 's#USR_SHARE_ZREPL#/usr/share/doc/zrepl#' packaging/systemd-default-zrepl.yml > debian/renamedir/zrepl.yml
+	dh_install debian/renamedir/zrepl.yml etc/zrepl
+
+	# save git revision of this packaging repo
+	echo $$(git rev-parse HEAD) > debian/packaging-repo-git-revision
+	dh_install debian/packaging-repo-git-revision usr/share/doc/zrepl
+
+override_dh_auto_test:
+	# don't run tests at this point
@@ -0,0 +1,7 @@
+FROM fedora:latest
+
+RUN dnf install -y git make bash rpm-build 'dnf-command(builddep)'
+ADD packaging/rpm/zrepl.spec /tmp/zrepl.spec
+RUN dnf builddep -y /tmp/zrepl.spec
+RUN mkdir -p /build/src && chmod -R 0777 /build
+WORKDIR /build/src
@@ -0,0 +1,78 @@
+# This Spec file packages pre-built artifacts in the artifacts directory
+# into an RPM. This means that rpmbuild won't actually build any new artifacts
+#
+# Read the Makefile first, then come back here.
+
+# Global meta data
+Version:          SUBSTITUTED_BY_MAKEFILE
+%global common_description %{expand:
+zrepl is a one-stop, integrated solution for ZFS replication.}
+
+# use gzip to be compatible with centos7
+%define _source_payload w9.gzdio
+%define _binary_payload w9.gzdio
+
+# don't strip pre-built binaries
+%define __strip /usr/bin/true
+
+Name:             zrepl
+Release:          1
+Summary:          One-stop, integrated solution for ZFS replication
+License:          MIT
+URL:              https://zrepl.github.io/
+# Source: we use rpmbuild --build-in-tree => no source
+BuildRequires:    systemd
+Requires(post):   systemd
+Requires(preun):  systemd
+Requires(postun): systemd
+
+%description
+%{common_description}
+
+%prep
+# we use rpmbuild --build-in-tree => no prep or setup
+
+%build
+# we don't actually need to build anything here, that has already been done by the Makefile
+
+# Correct the path in the systemd unit file
+sed s:/usr/local/bin/:%{_bindir}/:g dist/systemd/zrepl.service > artifacts/rpmbuild/zrepl.service
+
+# Generate the default configuration file
+sed 's#USR_SHARE_ZREPL#%{_datadir}/doc/zrepl#' packaging/systemd-default-zrepl.yml > artifacts/rpmbuild/zrepl.yml
+
+%install
+install -Dm 0755 artifacts/%{_zrepl_binary_filename}    %{buildroot}%{_bindir}/zrepl
+install -Dm 0644 artifacts/rpmbuild/zrepl.service       %{buildroot}%{_unitdir}/zrepl.service
+install -Dm 0644 artifacts/_zrepl.zsh_completion        %{buildroot}%{_datadir}/zsh/site-functions/_zrepl
+install -Dm 0644 artifacts/bash_completion              %{buildroot}%{_datadir}/bash-completion/completions/zrepl
+install -Dm 0644 artifacts/rpmbuild/zrepl.yml           %{buildroot}%{_sysconfdir}/zrepl/zrepl.yml
+install -d                                              %{buildroot}%{_datadir}/doc/zrepl
+cp -a   artifacts/docs/html                             %{buildroot}%{_datadir}/doc/zrepl/html
+cp -a   config/samples                                  %{buildroot}%{_datadir}/doc/zrepl/examples
+
+%post
+%systemd_post zrepl.service
+
+
+%preun
+%systemd_preun zrepl.service
+
+
+%postun
+%systemd_postun_with_restart zrepl.service
+
+
+%files
+%defattr(-,root,root)
+%license LICENSE
+%{_bindir}/zrepl
+%config %{_unitdir}/zrepl.service
+%dir %{_sysconfdir}/zrepl
+%config %{_sysconfdir}/zrepl/zrepl.yml
+%{_datadir}/zsh/site-functions/_zrepl
+%{_datadir}/bash-completion/completions/zrepl
+%{_datadir}/doc/zrepl
+
+%changelog
+# TODO: auto-fill changelog from git? -> need same solution for debian
@@ -0,0 +1,13 @@
+global:
+  logging:
+    # use syslog instead of stdout because it makes journald happy
+    - type: syslog
+      format: human
+      level: warn
+
+jobs:
+#  - name: foo
+#    type: bar
+
+# see USR_SHARE_ZREPL/examples
+# or https://zrepl.github.io/configuration/overview.html
@@ -11,6 +11,7 @@ import (

 	"github.com/fatih/color"
 	"github.com/pkg/errors"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"

 	"github.com/zrepl/zrepl/config"
@@ -16,15 +16,21 @@ var Cases = []Case{BatchDestroy,
 	ListFilesystemsNoFilter,
 	ReceiveForceIntoEncryptedErr,
 	ReceiveForceRollbackWorksUnencrypted,
+	ReplicationFailingInitialParentProhibitsChildReplication,
 	ReplicationIncrementalCleansUpStaleAbstractionsWithCacheOnSecondReplication,
 	ReplicationIncrementalCleansUpStaleAbstractionsWithoutCacheOnSecondReplication,
 	ReplicationIncrementalDestroysStepHoldsIffIncrementalStepHoldsAreDisabledButStepHoldsExist,
 	ReplicationIncrementalIsPossibleIfCommonSnapshotIsDestroyed,
-	ReplicationIsResumableFullSend__DisableIncrementalStepHolds_False,
-	ReplicationIsResumableFullSend__DisableIncrementalStepHolds_True,
+	ReplicationIsResumableFullSend__both_GuaranteeResumability,
+	ReplicationIsResumableFullSend__initial_GuaranteeIncrementalReplication_incremental_GuaranteeIncrementalReplication,
+	ReplicationIsResumableFullSend__initial_GuaranteeResumability_incremental_GuaranteeIncrementalReplication,
+	ReplicationReceiverErrorWhileStillSending,
+	ReplicationStepCompletedLostBehavior__GuaranteeIncrementalReplication,
+	ReplicationStepCompletedLostBehavior__GuaranteeResumability,
 	ResumableRecvAndTokenHandling,
 	ResumeTokenParsing,
-	SendArgsValidationEncryptedSendOfUnencryptedDatasetForbidden,
+	SendArgsValidationEncryptedSendOfUnencryptedDatasetForbidden__EncryptionSupported_false,
+	SendArgsValidationEncryptedSendOfUnencryptedDatasetForbidden__EncryptionSupported_true,
 	SendArgsValidationResumeTokenDifferentFilesystemForbidden,
 	SendArgsValidationResumeTokenEncryptionMismatchForbidden,
 	UndestroyableSnapshotParsing,
@@ -4,6 +4,7 @@ import (
 	"path"

 	"github.com/stretchr/testify/require"
+
 	"github.com/zrepl/zrepl/platformtest"
 	"github.com/zrepl/zrepl/zfs"
 )
@@ -4,6 +4,7 @@ import (
 	"fmt"

 	"github.com/stretchr/testify/assert"
+
 	"github.com/zrepl/zrepl/platformtest"
 	"github.com/zrepl/zrepl/zfs"
 )
@@ -4,11 +4,20 @@ import (
 	"fmt"

 	"github.com/stretchr/testify/require"
+
 	"github.com/zrepl/zrepl/platformtest"
 	"github.com/zrepl/zrepl/zfs"
 )

 func ReceiveForceIntoEncryptedErr(ctx *platformtest.Context) {
+
+	supported, err := zfs.EncryptionCLISupported(ctx)
+	require.NoError(ctx, err, "encryption feature test failed")
+	if !supported {
+		ctx.SkipNow()
+		return
+	}
+
 	platformtest.Run(ctx, platformtest.PanicErr, ctx.RootDataset, `
 		DESTROYROOT
 		CREATEROOT
@@ -32,6 +41,7 @@ func ReceiveForceIntoEncryptedErr(ctx *platformtest.Context) {

 	sendStream, err := zfs.ZFSSend(ctx, sendArgs)
 	require.NoError(ctx, err)
+	defer sendStream.Close()

 	recvOpts := zfs.RecvOptions{
 		RollbackAndForceRecv: true,
@@ -5,6 +5,7 @@ import (

 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
+
 	"github.com/zrepl/zrepl/platformtest"
 	"github.com/zrepl/zrepl/zfs"
 )
@@ -4,11 +4,13 @@ import (
 	"context"
 	"fmt"
 	"io"
+	"os"
 	"path"
 	"sort"

 	"github.com/kr/pretty"
 	"github.com/stretchr/testify/require"
+
 	"github.com/zrepl/zrepl/daemon/filters"
 	"github.com/zrepl/zrepl/endpoint"
 	"github.com/zrepl/zrepl/platformtest"
@@ -27,11 +29,13 @@ import (
 // of a new sender and receiver instance and one blocking invocation
 // of the replication engine without encryption
 type replicationInvocation struct {
-	sjid, rjid                  endpoint.JobID
-	sfs                         string
-	rfsRoot                     string
-	interceptSender             func(e *endpoint.Sender) logic.Sender
-	disableIncrementalStepHolds bool
+	sjid, rjid        endpoint.JobID
+	sfs               string
+	sfilter           *filters.DatasetMapFilter
+	rfsRoot           string
+	interceptSender   func(e *endpoint.Sender) logic.Sender
+	interceptReceiver func(e *endpoint.Receiver) logic.Receiver
+	guarantee         pdu.ReplicationConfigProtection
 }

 func (i replicationInvocation) Do(ctx *platformtest.Context) *report.Report {
@@ -39,24 +43,33 @@ func (i replicationInvocation) Do(ctx *platformtest.Context) *report.Report {
 	if i.interceptSender == nil {
 		i.interceptSender = func(e *endpoint.Sender) logic.Sender { return e }
 	}
+	if i.interceptReceiver == nil {
+		i.interceptReceiver = func(e *endpoint.Receiver) logic.Receiver { return e }
+	}

-	sfilter := filters.NewDatasetMapFilter(1, true)
-	err := sfilter.Add(i.sfs, "ok")
-	require.NoError(ctx, err)
+	if i.sfs != "" && i.sfilter != nil || i.sfs == "" && i.sfilter == nil {
+		panic("either sfs or sfilter must be set")
+	}
+	if i.sfilter == nil {
+		i.sfilter = filters.NewDatasetMapFilter(1, true)
+		err := i.sfilter.Add(i.sfs, "ok")
+		require.NoError(ctx, err)
+	}
 	sender := i.interceptSender(endpoint.NewSender(endpoint.SenderConfig{
-		FSF:                         sfilter.AsFilter(),
-		Encrypt:                     &zfs.NilBool{B: false},
-		DisableIncrementalStepHolds: i.disableIncrementalStepHolds,
-		JobID:                       i.sjid,
+		FSF:     i.sfilter.AsFilter(),
+		Encrypt: &zfs.NilBool{B: false},
+		JobID:   i.sjid,
 	}))
-	receiver := endpoint.NewReceiver(endpoint.ReceiverConfig{
+	receiver := i.interceptReceiver(endpoint.NewReceiver(endpoint.ReceiverConfig{
 		JobID:                      i.rjid,
 		AppendClientIdentity:       false,
 		RootWithoutClientComponent: mustDatasetPath(i.rfsRoot),
-		UpdateLastReceivedHold:     true,
-	})
+	}))
 	plannerPolicy := logic.PlannerPolicy{
 		EncryptedSend: logic.TriFromBool(false),
+		ReplicationConfig: pdu.ReplicationConfig{
+			Protection: &i.guarantee,
+		},
 	}

 	report, wait := replication.Do(
@@ -89,11 +102,11 @@ func ReplicationIncrementalIsPossibleIfCommonSnapshotIsDestroyed(ctx *platformte
 	snap1 := fsversion(ctx, sfs, "@1")

 	rep := replicationInvocation{
-		sjid:                        sjid,
-		rjid:                        rjid,
-		sfs:                         sfs,
-		rfsRoot:                     rfsRoot,
-		disableIncrementalStepHolds: false,
+		sjid:      sjid,
+		rjid:      rjid,
+		sfs:       sfs,
+		rfsRoot:   rfsRoot,
+		guarantee: *pdu.ReplicationConfigProtectionWithKind(pdu.ReplicationGuaranteeKind_GuaranteeResumability),
 	}
 	rfs := rep.ReceiveSideFilesystem()

@@ -153,11 +166,11 @@ func implReplicationIncrementalCleansUpStaleAbstractions(ctx *platformtest.Conte
 	rfsRoot := ctx.RootDataset + "/receiver"

 	rep := replicationInvocation{
-		sjid:                        sjid,
-		rjid:                        rjid,
-		sfs:                         sfs,
-		rfsRoot:                     rfsRoot,
-		disableIncrementalStepHolds: false,
+		sjid:      sjid,
+		rjid:      rjid,
+		sfs:       sfs,
+		rfsRoot:   rfsRoot,
+		guarantee: *pdu.ReplicationConfigProtectionWithKind(pdu.ReplicationGuaranteeKind_GuaranteeResumability),
 	}
 	rfs := rep.ReceiveSideFilesystem()

@@ -207,7 +220,7 @@ func implReplicationIncrementalCleansUpStaleAbstractions(ctx *platformtest.Conte
 	snap5 := fsversion(ctx, sfs, "@5")

 	if invalidateCacheBeforeSecondReplication {
-		endpoint.SendAbstractionsCacheInvalidate(sfs)
+		endpoint.AbstractionsCacheInvalidate(sfs)
 	}

 	// do another replication
@@ -327,15 +340,59 @@ func (s *PartialSender) Send(ctx context.Context, r *pdu.SendReq) (r1 *pdu.SendR
 	return r1, r2, r3
 }

-func ReplicationIsResumableFullSend__DisableIncrementalStepHolds_False(ctx *platformtest.Context) {
-	implReplicationIsResumableFullSend(ctx, false)
+func ReplicationIsResumableFullSend__both_GuaranteeResumability(ctx *platformtest.Context) {
+
+	setup := replicationIsResumableFullSendSetup{
+		protection: pdu.ReplicationConfigProtection{
+			Initial:     pdu.ReplicationGuaranteeKind_GuaranteeResumability,
+			Incremental: pdu.ReplicationGuaranteeKind_GuaranteeResumability,
+		},
+		expectDatasetIsBusyErrorWhenDestroySnapshotWhilePartiallyReplicated: true,
+		expectAllThreeSnapshotsToThreeBePresentAfterLoop:                    true,
+		expectNoSnapshotsOnReceiverAfterLoop:                                false,
+	}
+
+	implReplicationIsResumableFullSend(ctx, setup)
 }

-func ReplicationIsResumableFullSend__DisableIncrementalStepHolds_True(ctx *platformtest.Context) {
-	implReplicationIsResumableFullSend(ctx, true)
+func ReplicationIsResumableFullSend__initial_GuaranteeResumability_incremental_GuaranteeIncrementalReplication(ctx *platformtest.Context) {
+
+	setup := replicationIsResumableFullSendSetup{
+		protection: pdu.ReplicationConfigProtection{
+			Initial:     pdu.ReplicationGuaranteeKind_GuaranteeResumability,
+			Incremental: pdu.ReplicationGuaranteeKind_GuaranteeIncrementalReplication,
+		},
+		expectDatasetIsBusyErrorWhenDestroySnapshotWhilePartiallyReplicated: true,
+		expectAllThreeSnapshotsToThreeBePresentAfterLoop:                    true,
+		expectNoSnapshotsOnReceiverAfterLoop:                                false,
+	}
+
+	implReplicationIsResumableFullSend(ctx, setup)
 }

-func implReplicationIsResumableFullSend(ctx *platformtest.Context, disableIncrementalStepHolds bool) {
+func ReplicationIsResumableFullSend__initial_GuaranteeIncrementalReplication_incremental_GuaranteeIncrementalReplication(ctx *platformtest.Context) {
+
+	setup := replicationIsResumableFullSendSetup{
+		protection: pdu.ReplicationConfigProtection{
+			Initial:     pdu.ReplicationGuaranteeKind_GuaranteeIncrementalReplication,
+			Incremental: pdu.ReplicationGuaranteeKind_GuaranteeIncrementalReplication,
+		},
+		expectDatasetIsBusyErrorWhenDestroySnapshotWhilePartiallyReplicated: false,
+		expectAllThreeSnapshotsToThreeBePresentAfterLoop:                    false,
+		expectNoSnapshotsOnReceiverAfterLoop:                                true,
+	}
+
+	implReplicationIsResumableFullSend(ctx, setup)
+}
+
+type replicationIsResumableFullSendSetup struct {
+	protection                                                          pdu.ReplicationConfigProtection
+	expectDatasetIsBusyErrorWhenDestroySnapshotWhilePartiallyReplicated bool
+	expectAllThreeSnapshotsToThreeBePresentAfterLoop                    bool
+	expectNoSnapshotsOnReceiverAfterLoop                                bool
+}
+
+func implReplicationIsResumableFullSend(ctx *platformtest.Context, setup replicationIsResumableFullSendSetup) {

 	platformtest.Run(ctx, platformtest.PanicErr, ctx.RootDataset, `
 		CREATEROOT
@@ -366,8 +423,9 @@ func implReplicationIsResumableFullSend(ctx *platformtest.Context, disableIncrem
 		interceptSender: func(e *endpoint.Sender) logic.Sender {
 			return &PartialSender{Sender: e, failAfterByteCount: 1 << 20}
 		},
-		disableIncrementalStepHolds: disableIncrementalStepHolds,
+		guarantee: setup.protection,
 	}
+
 	rfs := rep.ReceiveSideFilesystem()

 	for i := 2; i < 10; i++ {
@@ -381,8 +439,11 @@ func implReplicationIsResumableFullSend(ctx *platformtest.Context, disableIncrem
 			// and we wrote dummy data 1<<22 bytes, thus at least
 			// for the first 4 times this should not be possible
 			// due to step holds
-			require.Error(ctx, err)
-			require.Contains(ctx, err.Error(), "dataset is busy")
+			if setup.expectDatasetIsBusyErrorWhenDestroySnapshotWhilePartiallyReplicated {
+				ctx.Logf("i=%v", i)
+				require.Error(ctx, err)
+				require.Contains(ctx, err.Error(), "dataset is busy")
+			}
 		}

 		// and create some additional snapshots that could
@@ -401,11 +462,19 @@ func implReplicationIsResumableFullSend(ctx *platformtest.Context, disableIncrem
 		}
 	}

-	// make sure all the filesystem versions we created
-	// were replicated by the replication loop
-	_ = fsversion(ctx, rfs, "@1")
-	_ = fsversion(ctx, rfs, "@2")
-	_ = fsversion(ctx, rfs, "@3")
+	if setup.expectAllThreeSnapshotsToThreeBePresentAfterLoop {
+		// make sure all the filesystem versions we created
+		// were replicated by the replication loop
+		_ = fsversion(ctx, rfs, "@1")
+		_ = fsversion(ctx, rfs, "@2")
+		_ = fsversion(ctx, rfs, "@3")
+	}
+
+	if setup.expectNoSnapshotsOnReceiverAfterLoop {
+		versions, err := zfs.ZFSListFilesystemVersions(ctx, mustDatasetPath(rfs), zfs.ListFilesystemVersionsOptions{})
+		require.NoError(ctx, err)
+		require.Empty(ctx, versions)
+	}

 }

@@ -428,11 +497,11 @@ func ReplicationIncrementalDestroysStepHoldsIffIncrementalStepHoldsAreDisabledBu
 	{
 		mustSnapshot(ctx, sfs+"@1")
 		rep := replicationInvocation{
-			sjid:                        sjid,
-			rjid:                        rjid,
-			sfs:                         sfs,
-			rfsRoot:                     rfsRoot,
-			disableIncrementalStepHolds: false,
+			sjid:      sjid,
+			rjid:      rjid,
+			sfs:       sfs,
+			rfsRoot:   rfsRoot,
+			guarantee: *pdu.ReplicationConfigProtectionWithKind(pdu.ReplicationGuaranteeKind_GuaranteeResumability),
 		}
 		rfs := rep.ReceiveSideFilesystem()
 		report := rep.Do(ctx)
@@ -455,11 +524,11 @@ func ReplicationIncrementalDestroysStepHoldsIffIncrementalStepHoldsAreDisabledBu
 	// to effect a step-holds situation
 	{
 		rep := replicationInvocation{
-			sjid:                        sjid,
-			rjid:                        rjid,
-			sfs:                         sfs,
-			rfsRoot:                     rfsRoot,
-			disableIncrementalStepHolds: false, // !
+			sjid:      sjid,
+			rjid:      rjid,
+			sfs:       sfs,
+			rfsRoot:   rfsRoot,
+			guarantee: *pdu.ReplicationConfigProtectionWithKind(pdu.ReplicationGuaranteeKind_GuaranteeResumability), // !
 			interceptSender: func(e *endpoint.Sender) logic.Sender {
 				return &PartialSender{Sender: e, failAfterByteCount: 1 << 20}
 			},
@@ -495,17 +564,17 @@ func ReplicationIncrementalDestroysStepHoldsIffIncrementalStepHoldsAreDisabledBu
 	// end of test setup
 	//

-	// retry replication with incremental step holds disabled
+	// retry replication with incremental step holds disabled (set to bookmarks-only in this case)
 	// - replication should not fail due to holds-related stuff
 	// - replication should fail intermittently due to partial sender being fully read
 	// - the partial sender is 1/4th the length of the stream, thus expect
 	//   successful replication after 5 more attempts
 	rep := replicationInvocation{
-		sjid:                        sjid,
-		rjid:                        rjid,
-		sfs:                         sfs,
-		rfsRoot:                     rfsRoot,
-		disableIncrementalStepHolds: true, // !
+		sjid:      sjid,
+		rjid:      rjid,
+		sfs:       sfs,
+		rfsRoot:   rfsRoot,
+		guarantee: *pdu.ReplicationConfigProtectionWithKind(pdu.ReplicationGuaranteeKind_GuaranteeIncrementalReplication), // !
 		interceptSender: func(e *endpoint.Sender) logic.Sender {
 			return &PartialSender{Sender: e, failAfterByteCount: 1 << 20}
 		},
@@ -551,3 +620,298 @@ func ReplicationIncrementalDestroysStepHoldsIffIncrementalStepHoldsAreDisabledBu
 	require.Len(ctx, abs, 1)
 	require.True(ctx, zfs.FilesystemVersionEqualIdentity(abs[0].GetFilesystemVersion(), snap2sfs))
 }
+
+func ReplicationStepCompletedLostBehavior__GuaranteeResumability(ctx *platformtest.Context) {
+	scenario := replicationStepCompletedLostBehavior_impl(ctx, pdu.ReplicationGuaranteeKind_GuaranteeResumability)
+
+	require.Error(ctx, scenario.deleteSfs1Err, "protected by holds")
+	require.Contains(ctx, scenario.deleteSfs1Err.Error(), "dataset is busy")
+
+	require.Error(ctx, scenario.deleteSfs2Err, "protected by holds")
+	require.Contains(ctx, scenario.deleteSfs2Err.Error(), "dataset is busy")
+
+	require.Nil(ctx, scenario.finalReport.Error())
+	_ = fsversion(ctx, scenario.rfs, "@3") // @3 ade it to the other side
+}
+
+func ReplicationStepCompletedLostBehavior__GuaranteeIncrementalReplication(ctx *platformtest.Context) {
+	scenario := replicationStepCompletedLostBehavior_impl(ctx, pdu.ReplicationGuaranteeKind_GuaranteeIncrementalReplication)
+
+	require.NoError(ctx, scenario.deleteSfs1Err, "not protected by holds")
+	require.NoError(ctx, scenario.deleteSfs2Err, "not protected by holds")
+
+	// step bookmarks should protect against loss of StepCompleted message
+	require.Nil(ctx, scenario.finalReport.Error())
+	_ = fsversion(ctx, scenario.rfs, "@3") // @3 ade it to the other side
+}
+
+type FailSendCompletedSender struct {
+	*endpoint.Sender
+}
+
+var _ logic.Sender = (*FailSendCompletedSender)(nil)
+
+func (p *FailSendCompletedSender) SendCompleted(ctx context.Context, r *pdu.SendCompletedReq) (*pdu.SendCompletedRes, error) {
+	return nil, fmt.Errorf("[mock] SendCompleted not delivered to actual endpoint")
+}
+
+type replicationStepCompletedLost_scenario struct {
+	rfs                          string
+	deleteSfs1Err, deleteSfs2Err error
+	finalReport                  *report.FilesystemReport
+}
+
+func replicationStepCompletedLostBehavior_impl(ctx *platformtest.Context, guaranteeKind pdu.ReplicationGuaranteeKind) *replicationStepCompletedLost_scenario {
+
+	platformtest.Run(ctx, platformtest.PanicErr, ctx.RootDataset, `
+		CREATEROOT
+		+  "sender"
+		+  "receiver"
+		R  zfs create -p "${ROOTDS}/receiver/${ROOTDS}"
+	`)
+
+	sjid := endpoint.MustMakeJobID("sender-job")
+	rjid := endpoint.MustMakeJobID("receiver-job")
+
+	sfs := ctx.RootDataset + "/sender"
+	rfsRoot := ctx.RootDataset + "/receiver"
+
+	// fully replicate snapshots @1
+	{
+		mustSnapshot(ctx, sfs+"@1")
+		rep := replicationInvocation{
+			sjid:      sjid,
+			rjid:      rjid,
+			sfs:       sfs,
+			rfsRoot:   rfsRoot,
+			guarantee: *pdu.ReplicationConfigProtectionWithKind(guaranteeKind),
+		}
+		rfs := rep.ReceiveSideFilesystem()
+		report := rep.Do(ctx)
+		ctx.Logf("\n%s", pretty.Sprint(report))
+		// assert this worked (not the main subject of the test)
+		_ = fsversion(ctx, rfs, "@1")
+	}
+
+	// create a second snapshot @2
+	mustSnapshot(ctx, sfs+"@2")
+
+	// fake loss of stepcompleted message
+	rep := replicationInvocation{
+		sjid:      sjid,
+		rjid:      rjid,
+		sfs:       sfs,
+		rfsRoot:   rfsRoot,
+		guarantee: *pdu.ReplicationConfigProtectionWithKind(guaranteeKind),
+		interceptSender: func(e *endpoint.Sender) logic.Sender {
+			return &FailSendCompletedSender{e}
+		},
+	}
+	rfs := rep.ReceiveSideFilesystem()
+	report := rep.Do(ctx)
+	ctx.Logf("\n%s", pretty.Sprint(report))
+
+	// assert the replication worked
+	_ = fsversion(ctx, rfs, "@2")
+	// and that we hold it using a last-received-hold
+	abs, absErrs, err := endpoint.ListAbstractions(ctx, endpoint.ListZFSHoldsAndBookmarksQuery{
+		FS: endpoint.ListZFSHoldsAndBookmarksQueryFilesystemFilter{
+			FS: &rfs,
+		},
+		Concurrency: 1,
+		JobID:       &rjid,
+		What:        endpoint.AbstractionTypeSet{endpoint.AbstractionLastReceivedHold: true},
+	})
+	require.NoError(ctx, err)
+	require.Empty(ctx, absErrs)
+	require.Len(ctx, abs, 1)
+	require.True(ctx, zfs.FilesystemVersionEqualIdentity(abs[0].GetFilesystemVersion(), fsversion(ctx, rfs, "@2")))
+
+	// now try to delete @2 on the sender, this should work because don't have step holds on it
+	deleteSfs2Err := zfs.ZFSDestroy(ctx, sfs+"@2")
+	// defer check to caller
+
+	// and create a new snapshot on the sender
+	mustSnapshot(ctx, sfs+"@3")
+
+	// now we have: sender @1, @3
+	//              recver @1, @2
+
+	// delete @1 on both sides to demonstrate that, if we didn't have bookmarks, we would be out of sync
+	deleteSfs1Err := zfs.ZFSDestroy(ctx, sfs+"@1")
+	// defer check to caller
+	err = zfs.ZFSDestroy(ctx, rfs+"@1")
+	require.NoError(ctx, err)
+
+	// attempt replication and return the filesystem report report
+	{
+		rep := replicationInvocation{
+			sjid:      sjid,
+			rjid:      rjid,
+			sfs:       sfs,
+			rfsRoot:   rfsRoot,
+			guarantee: *pdu.ReplicationConfigProtectionWithKind(guaranteeKind),
+		}
+		report := rep.Do(ctx)
+		ctx.Logf("expecting failure:\n%s", pretty.Sprint(report))
+		require.Len(ctx, report.Attempts, 1)
+		require.Len(ctx, report.Attempts[0].Filesystems, 1)
+		return &replicationStepCompletedLost_scenario{
+			rfs:           rfs,
+			deleteSfs1Err: deleteSfs1Err,
+			deleteSfs2Err: deleteSfs2Err,
+			finalReport:   report.Attempts[0].Filesystems[0],
+		}
+	}
+
+}
+
+type ErroringReceiver struct {
+	recvErr error
+	*endpoint.Receiver
+}
+
+func (r *ErroringReceiver) Receive(ctx context.Context, req *pdu.ReceiveReq, stream io.ReadCloser) (*pdu.ReceiveRes, error) {
+	return nil, r.recvErr
+}
+
+type NeverEndingSender struct {
+	*endpoint.Sender
+}
+
+func (s *NeverEndingSender) Send(ctx context.Context, req *pdu.SendReq) (r *pdu.SendRes, stream io.ReadCloser, _ error) {
+	stream = nil
+	r = &pdu.SendRes{
+		UsedResumeToken: false,
+		ExpectedSize:    1 << 30,
+	}
+	if req.DryRun {
+		return r, stream, nil
+	}
+	dz, err := os.Open("/dev/zero")
+	if err != nil {
+		panic(err)
+	}
+	return r, dz, nil
+}
+
+func ReplicationReceiverErrorWhileStillSending(ctx *platformtest.Context) {
+
+	platformtest.Run(ctx, platformtest.PanicErr, ctx.RootDataset, `
+		CREATEROOT
+		+  "sender"
+		+  "sender@1"
+		+  "receiver"
+		R  zfs create -p "${ROOTDS}/receiver/${ROOTDS}"
+	`)
+
+	sjid := endpoint.MustMakeJobID("sender-job")
+	rjid := endpoint.MustMakeJobID("receiver-job")
+
+	sfs := ctx.RootDataset + "/sender"
+	rfsRoot := ctx.RootDataset + "/receiver"
+
+	mockRecvErr := fmt.Errorf("YiezahK3thie8ahKiel5sah2uugei2ize1yi8feivuu7musoat")
+
+	rep := replicationInvocation{
+		sjid:      sjid,
+		rjid:      rjid,
+		sfs:       sfs,
+		rfsRoot:   rfsRoot,
+		guarantee: *pdu.ReplicationConfigProtectionWithKind(pdu.ReplicationGuaranteeKind_GuaranteeNothing),
+		interceptReceiver: func(r *endpoint.Receiver) logic.Receiver {
+			return &ErroringReceiver{recvErr: mockRecvErr, Receiver: r}
+		},
+		interceptSender: func(s *endpoint.Sender) logic.Sender {
+			return &NeverEndingSender{s}
+		},
+	}
+
+	// first replication
+	report := rep.Do(ctx)
+	ctx.Logf("\n%s", pretty.Sprint(report))
+
+	require.Len(ctx, report.Attempts, 1)
+	attempt := report.Attempts[0]
+	require.Nil(ctx, attempt.PlanError)
+	require.Len(ctx, attempt.Filesystems, 1)
+	afs := attempt.Filesystems[0]
+	require.Nil(ctx, afs.PlanError)
+	require.Len(ctx, afs.Steps, 1)
+	require.Nil(ctx, afs.PlanError)
+	require.NotNil(ctx, afs.StepError)
+	require.Contains(ctx, afs.StepError.Err, mockRecvErr.Error())
+}
+
+func ReplicationFailingInitialParentProhibitsChildReplication(ctx *platformtest.Context) {
+
+	platformtest.Run(ctx, platformtest.PanicErr, ctx.RootDataset, `
+		CREATEROOT
+		+  "sender"
+		+  "sender/a"
+		+  "sender/a/child"
+		+  "sender/aa"
+		+  "receiver"
+		R  zfs create -p "${ROOTDS}/receiver/${ROOTDS}"
+		R  zfs snapshot -r ${ROOTDS}/sender@initial
+	`)
+
+	sjid := endpoint.MustMakeJobID("sender-job")
+	rjid := endpoint.MustMakeJobID("receiver-job")
+
+	fsA := ctx.RootDataset + "/sender/a"
+	fsAChild := ctx.RootDataset + "/sender/a/child"
+	fsAA := ctx.RootDataset + "/sender/aa"
+
+	sfilter := filters.NewDatasetMapFilter(3, true)
+	mustAddToSFilter := func(fs string) {
+		err := sfilter.Add(fs, "ok")
+		require.NoError(ctx, err)
+	}
+	mustAddToSFilter(fsA)
+	mustAddToSFilter(fsAChild)
+	mustAddToSFilter(fsAA)
+	rfsRoot := ctx.RootDataset + "/receiver"
+
+	mockRecvErr := fmt.Errorf("yifae4ohPhaquaes0hohghiep9oufie4roo7quoWooluaj2ee8")
+
+	rep := replicationInvocation{
+		sjid:      sjid,
+		rjid:      rjid,
+		sfilter:   sfilter,
+		rfsRoot:   rfsRoot,
+		guarantee: *pdu.ReplicationConfigProtectionWithKind(pdu.ReplicationGuaranteeKind_GuaranteeNothing),
+		interceptReceiver: func(r *endpoint.Receiver) logic.Receiver {
+			return &ErroringReceiver{recvErr: mockRecvErr, Receiver: r}
+		},
+	}
+
+	r := rep.Do(ctx)
+	ctx.Logf("\n%s", pretty.Sprint(r))
+
+	require.Len(ctx, r.Attempts, 1)
+	attempt := r.Attempts[0]
+	require.Nil(ctx, attempt.PlanError)
+	require.Len(ctx, attempt.Filesystems, 3)
+
+	fsByName := make(map[string]*report.FilesystemReport, len(attempt.Filesystems))
+	for _, fs := range attempt.Filesystems {
+		fsByName[fs.Info.Name] = fs
+	}
+
+	require.Contains(ctx, fsByName, fsA)
+	require.Contains(ctx, fsByName, fsAA)
+	require.Contains(ctx, fsByName, fsAA)
+
+	checkFS := func(fs string, expectErrMsg string) {
+		rep := fsByName[fs]
+		require.Len(ctx, rep.Steps, 1)
+		require.Nil(ctx, rep.PlanError)
+		require.NotNil(ctx, rep.StepError)
+		require.Contains(ctx, rep.StepError.Err, expectErrMsg)
+	}
+
+	checkFS(fsA, mockRecvErr.Error())
+	checkFS(fsAChild, "parent(s) failed during initial replication")
+	checkFS(fsAA, mockRecvErr.Error()) // fsAA is not treated as a child of fsA
+}
@@ -22,6 +22,7 @@ func CreateReplicationCursor(ctx *platformtest.Context) {
 		R  zfs bookmark "${ROOTDS}/foo bar@2 with space" "${ROOTDS}/foo bar#2 with space"
 		+  "foo bar@3 with space"
 		R  zfs bookmark "${ROOTDS}/foo bar@3 with space" "${ROOTDS}/foo bar#3 with space"
+		-  "foo bar@3 with space"
 	`)

 	jobid := endpoint.MustMakeJobID("zreplplatformtest")
@@ -42,6 +43,7 @@ func CreateReplicationCursor(ctx *platformtest.Context) {

 	snap := fsversion(ctx, fs, "@1 with space")
 	book := fsversion(ctx, fs, "#1 with space")
+	book3 := fsversion(ctx, fs, "#3 with space")

 	// create first cursor
 	cursorOfSnap, err := endpoint.CreateReplicationCursor(ctx, fs, snap, jobid)
@@ -49,8 +51,11 @@ func CreateReplicationCursor(ctx *platformtest.Context) {
 	// check CreateReplicationCursor is idempotent (for snapshot target)
 	cursorOfSnapIdemp, err := endpoint.CreateReplicationCursor(ctx, fs, snap, jobid)
 	checkCreateCursor(err, cursorOfSnap, snap)
+	// check CreateReplicationCursor is idempotent (for bookmark target of snapshot)
+	cursorOfBook, err := endpoint.CreateReplicationCursor(ctx, fs, book, jobid)
+	checkCreateCursor(err, cursorOfBook, snap)
 	// ... for target = non-cursor bookmark
-	_, err = endpoint.CreateReplicationCursor(ctx, fs, book, jobid)
+	_, err = endpoint.CreateReplicationCursor(ctx, fs, book3, jobid)
 	assert.Equal(ctx, zfs.ErrBookmarkCloningNotSupported, err)
 	// ... for target = replication cursor bookmark to be created
 	cursorOfCursor, err := endpoint.CreateReplicationCursor(ctx, fs, cursorOfSnapIdemp.GetFilesystemVersion(), jobid)
@@ -9,11 +9,22 @@ import (
 	"github.com/zrepl/zrepl/zfs"
 )

-func SendArgsValidationEncryptedSendOfUnencryptedDatasetForbidden(ctx *platformtest.Context) {
+func SendArgsValidationEncryptedSendOfUnencryptedDatasetForbidden__EncryptionSupported_true(ctx *platformtest.Context) {
+	sendArgsValidationEncryptedSendOfUnencryptedDatasetForbidden_impl(ctx, true)
+}
+
+func SendArgsValidationEncryptedSendOfUnencryptedDatasetForbidden__EncryptionSupported_false(ctx *platformtest.Context) {
+	sendArgsValidationEncryptedSendOfUnencryptedDatasetForbidden_impl(ctx, false)
+}
+
+func sendArgsValidationEncryptedSendOfUnencryptedDatasetForbidden_impl(ctx *platformtest.Context, testForEncryptionSupported bool) {

 	supported, err := zfs.EncryptionCLISupported(ctx)
 	check(err)
-	expectNotSupportedErr := !supported
+	if supported != testForEncryptionSupported {
+		ctx.SkipNow()
+	}
+	noEncryptionCLISupport := !supported

 	platformtest.Run(ctx, platformtest.PanicErr, ctx.RootDataset, `
 	DESTROYROOT
@@ -44,9 +55,11 @@ func SendArgsValidationEncryptedSendOfUnencryptedDatasetForbidden(ctx *platformt
 		// fallthrough
 	}

-	if expectNotSupportedErr {
+	if noEncryptionCLISupport {
 		require.Error(ctx, err)
-		require.Equal(ctx, zfs.ErrEncryptedSendNotSupported, err)
+		saverr, ok := err.(*zfs.ZFSSendArgsValidationError)
+		require.True(ctx, ok, "%T", err)
+		require.Equal(ctx, zfs.ZFSSendArgsEncryptedSendRequestedButFSUnencrypted, saverr.What)
 		return
 	}
 	require.Error(ctx, err)
@@ -138,13 +151,7 @@ func SendArgsValidationResumeTokenEncryptionMismatchForbidden(ctx *platformtest.
 }

 func SendArgsValidationResumeTokenDifferentFilesystemForbidden(ctx *platformtest.Context) {
-
-	supported, err := zfs.EncryptionCLISupported(ctx)
-	check(err)
-	if !supported {
-		ctx.SkipNow()
-	}
-	supported, err = zfs.ResumeSendSupported(ctx)
+	supported, err := zfs.ResumeSendSupported(ctx)
 	check(err)
 	if !supported {
 		ctx.SkipNow()
@@ -3,7 +3,6 @@ package pruning
 import (
 	"fmt"
 	"regexp"
-	"sort"
 	"time"

 	"github.com/pkg/errors"
@@ -30,75 +29,88 @@ func NewKeepGrid(in *config.PruneGrid) (p *KeepGrid, err error) {
 		return nil, errors.Wrap(err, "Regex is invalid")
 	}

+	return newKeepGrid(re, in.Grid)
+}
+
+func MustNewKeepGrid(regex, gridspec string) *KeepGrid {
+
+	ris, err := config.ParseRetentionIntervalSpec(gridspec)
+	if err != nil {
+		panic(err)
+	}
+
+	re := regexp.MustCompile(regex)
+
+	grid, err := newKeepGrid(re, ris)
+	if err != nil {
+		panic(err)
+	}
+	return grid
+}
+
+func newKeepGrid(re *regexp.Regexp, configIntervals []config.RetentionInterval) (*KeepGrid, error) {
+	if re == nil {
+		panic("re must not be nil")
+	}
+
+	if len(configIntervals) == 0 {
+		return nil, errors.New("retention grid must specify at least one interval")
+	}
+
+	intervals := make([]retentiongrid.Interval, len(configIntervals))
+	for i := range configIntervals {
+		intervals[i] = &configIntervals[i]
+	}
+
 	// Assert intervals are of increasing length (not necessarily required, but indicates config mistake)
 	lastDuration := time.Duration(0)
-	for i := range in.Grid {
+	for i := range intervals {

-		if in.Grid[i].Length() < lastDuration {
+		if intervals[i].Length() < lastDuration {
 			// If all intervals before were keep=all, this is ok
 			allPrevKeepCountAll := true
 			for j := i - 1; allPrevKeepCountAll && j >= 0; j-- {
-				allPrevKeepCountAll = in.Grid[j].KeepCount() == config.RetentionGridKeepCountAll
+				allPrevKeepCountAll = intervals[j].KeepCount() == config.RetentionGridKeepCountAll
 			}
 			if allPrevKeepCountAll {
 				goto isMonotonicIncrease
 			}
-			err = errors.New("retention grid interval length must be monotonically increasing")
-			return
+			return nil, errors.New("retention grid interval length must be monotonically increasing")
 		}
 	isMonotonicIncrease:
-		lastDuration = in.Grid[i].Length()
-
-	}
-
-	retentionIntervals := make([]retentiongrid.Interval, len(in.Grid))
-	for i := range in.Grid {
-		retentionIntervals[i] = &in.Grid[i]
+		lastDuration = intervals[i].Length()
 	}

 	return &KeepGrid{
-		retentiongrid.NewGrid(retentionIntervals),
-		re,
+		retentionGrid: retentiongrid.NewGrid(intervals),
+		re:            re,
 	}, nil
 }

-type retentionGridAdaptor struct {
-	Snapshot
-}
-
-func (a retentionGridAdaptor) LessThan(b retentiongrid.Entry) bool {
-	return a.Date().Before(b.Date())
-}
-
 // Prune filters snapshots with the retention grid.
 func (p *KeepGrid) KeepRule(snaps []Snapshot) (destroyList []Snapshot) {

-	snaps = filterSnapList(snaps, func(snapshot Snapshot) bool {
+	matching, notMatching := partitionSnapList(snaps, func(snapshot Snapshot) bool {
 		return p.re.MatchString(snapshot.Name())
 	})
-	if len(snaps) == 0 {
-		return nil
-	}

-	// Build adaptors for retention grid
-	adaptors := make([]retentiongrid.Entry, 0)
-	for i := range snaps {
-		adaptors = append(adaptors, retentionGridAdaptor{snaps[i]})
-	}
+	// snaps that don't match the regex are not kept by this rule
+	destroyList = append(destroyList, notMatching...)

-	// determine 'now' edge
-	sort.SliceStable(adaptors, func(i, j int) bool {
-		return adaptors[i].LessThan(adaptors[j])
-	})
-	now := adaptors[len(adaptors)-1].Date()
+	if len(matching) == 0 {
+		return destroyList
+	}

 	// Evaluate retention grid
-	_, removea := p.retentionGrid.FitEntries(now, adaptors)
+	entrySlice := make([]retentiongrid.Entry, 0)
+	for i := range matching {
+		entrySlice = append(entrySlice, matching[i])
+	}
+	_, gridDestroyList := p.retentionGrid.FitEntries(entrySlice)

 	// Revert adaptors
-	destroyList = make([]Snapshot, len(removea))
-	for i := range removea {
-		destroyList[i] = removea[i].(retentionGridAdaptor).Snapshot
+	for i := range gridDestroyList {
+		destroyList = append(destroyList, gridDestroyList[i].(Snapshot))
 	}
 	return destroyList
 }
@@ -10,6 +10,17 @@ func filterSnapList(snaps []Snapshot, predicate func(Snapshot) bool) []Snapshot
 	return r
 }

+func partitionSnapList(snaps []Snapshot, predicate func(Snapshot) bool) (sTrue, sFalse []Snapshot) {
+	for i := range snaps {
+		if predicate(snaps[i]) {
+			sTrue = append(sTrue, snaps[i])
+		} else {
+			sFalse = append(sFalse, snaps[i])
+		}
+	}
+	return
+}
+
 func shallowCopySnapList(snaps []Snapshot) []Snapshot {
 	c := make([]Snapshot, len(snaps))
 	copy(c, snaps)
@@ -1,20 +1,35 @@
 package pruning

 import (
+	"regexp"
 	"sort"
+	"strings"

 	"github.com/pkg/errors"
 )

 type KeepLastN struct {
-	n int
+	n  int
+	re *regexp.Regexp
 }

-func NewKeepLastN(n int) (*KeepLastN, error) {
+func MustKeepLastN(n int, regex string) *KeepLastN {
+	k, err := NewKeepLastN(n, regex)
+	if err != nil {
+		panic(err)
+	}
+	return k
+}
+
+func NewKeepLastN(n int, regex string) (*KeepLastN, error) {
 	if n <= 0 {
 		return nil, errors.Errorf("must specify positive number as 'keep last count', got %d", n)
 	}
-	return &KeepLastN{n}, nil
+	re, err := regexp.Compile(regex)
+	if err != nil {
+		return nil, errors.Errorf("invalid regex %q: %s", regex, err)
+	}
+	return &KeepLastN{n, re}, nil
 }

 func (k KeepLastN) KeepRule(snaps []Snapshot) (destroyList []Snapshot) {
@@ -23,11 +38,25 @@ func (k KeepLastN) KeepRule(snaps []Snapshot) (destroyList []Snapshot) {
 		return []Snapshot{}
 	}

-	res := shallowCopySnapList(snaps)
-
-	sort.Slice(res, func(i, j int) bool {
-		return res[i].Date().After(res[j].Date())
+	matching, notMatching := partitionSnapList(snaps, func(snapshot Snapshot) bool {
+		return k.re.MatchString(snapshot.Name())
 	})
+	// snaps that don't match the regex are not kept by this rule
+	destroyList = append(destroyList, notMatching...)

-	return res[k.n:]
+	if len(matching) == 0 {
+		return destroyList
+	}
+
+	sort.Slice(matching, func(i, j int) bool {
+		// by date (youngest first)
+		id, jd := matching[i].Date(), matching[j].Date()
+		if !id.Equal(jd) {
+			return id.After(jd)
+		}
+		// then lexicographically descending (e.g. b, a)
+		return strings.Compare(matching[i].Name(), matching[j].Name()) == 1
+	})
+	destroyList = append(destroyList, matching[k.n:]...)
+	return destroyList
 }
@@ -5,6 +5,7 @@ import (
 	"time"

 	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
 )

 func TestKeepLastN(t *testing.T) {
@@ -28,7 +29,7 @@ func TestKeepLastN(t *testing.T) {
 		"keep2": {
 			inputs: inputs["s1"],
 			rules: []KeepRule{
-				KeepLastN{2},
+				MustKeepLastN(2, ""),
 			},
 			expDestroy: map[string]bool{
 				"1": true, "2": true, "3": true,
@@ -37,37 +38,66 @@ func TestKeepLastN(t *testing.T) {
 		"keep1OfTwoWithSameTime": { // Keep one of two with same time
 			inputs: inputs["s1"],
 			rules: []KeepRule{
-				KeepLastN{1},
-			},
-			expDestroyAlternatives: []map[string]bool{
-				{"1": true, "2": true, "3": true, "4": true},
-				{"1": true, "2": true, "3": true, "5": true},
+				MustKeepLastN(1, ""),
 			},
+			expDestroy: map[string]bool{"1": true, "2": true, "3": true, "4": true},
 		},
 		"keepMany": {
 			inputs: inputs["s1"],
 			rules: []KeepRule{
-				KeepLastN{100},
+				MustKeepLastN(100, ""),
 			},
 			expDestroy: map[string]bool{},
 		},
-		"empty": {
+		"empty_input": {
 			inputs: inputs["s2"],
 			rules: []KeepRule{
-				KeepLastN{100},
+				MustKeepLastN(100, ""),
 			},
 			expDestroy: map[string]bool{},
 		},
+		"empty_regex": {
+			inputs: inputs["s1"],
+			rules: []KeepRule{
+				MustKeepLastN(4, ""),
+			},
+			expDestroy: map[string]bool{
+				"1": true,
+			},
+		},
+		"multiple_regexes": {
+			inputs: []Snapshot{
+				stubSnap{"a1", false, o(10)},
+				stubSnap{"b1", false, o(11)},
+				stubSnap{"a2", false, o(20)},
+				stubSnap{"b2", false, o(21)},
+				stubSnap{"a3", false, o(30)},
+				stubSnap{"b3", false, o(31)},
+			},
+			rules: []KeepRule{
+				MustKeepLastN(2, "^a"),
+				MustKeepLastN(2, "^b"),
+			},
+			expDestroy: map[string]bool{
+				"a1": true,
+				"b1": true,
+			},
+		},
 	}

 	testTable(tcs, t)

 	t.Run("mustBePositive", func(t *testing.T) {
 		var err error
-		_, err = NewKeepLastN(0)
+		_, err = NewKeepLastN(0, "foo")
 		assert.Error(t, err)
-		_, err = NewKeepLastN(-5)
+		_, err = NewKeepLastN(-5, "foo")
 		assert.Error(t, err)
 	})

+	t.Run("emptyRegexAllowed", func(t *testing.T) {
+		_, err := NewKeepLastN(23, "")
+		require.NoError(t, err)
+	})
+
 }
@@ -60,7 +60,7 @@ func RuleFromConfig(in config.PruningEnum) (KeepRule, error) {
 	case *config.PruneKeepNotReplicated:
 		return NewKeepNotReplicated(), nil
 	case *config.PruneKeepLastN:
-		return NewKeepLastN(v.Count)
+		return NewKeepLastN(v.Count, v.Regex)
 	case *config.PruneKeepRegex:
 		return NewKeepRegex(v.Regex, v.Negate)
 	case *config.PruneGrid:
@@ -3,6 +3,9 @@ package pruning
 import (
 	"testing"
 	"time"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
 )

 type stubSnap struct {
@@ -18,10 +21,9 @@ func (s stubSnap) Replicated() bool { return s.replicated }
 func (s stubSnap) Date() time.Time { return s.date }

 type testCase struct {
-	inputs                 []Snapshot
-	rules                  []KeepRule
-	expDestroy, effDestroy map[string]bool
-	expDestroyAlternatives []map[string]bool
+	inputs     []Snapshot
+	rules      []KeepRule
+	expDestroy map[string]bool
 }

 type snapshotList []Snapshot
@@ -44,41 +46,20 @@ func (l snapshotList) NameList() []string {
 }

 func testTable(tcs map[string]testCase, t *testing.T) {
-	mapEqual := func(a, b map[string]bool) bool {
-		if len(a) != len(b) {
-			return false
-		}
-		for k, v := range a {
-			if w, ok := b[k]; !ok || v != w {
-				return false
-			}
-		}
-		return true
-	}
-
 	for name := range tcs {
 		t.Run(name, func(t *testing.T) {
 			tc := tcs[name]
-			remove := PruneSnapshots(tc.inputs, tc.rules)
-			tc.effDestroy = make(map[string]bool)
-			for _, s := range remove {
-				tc.effDestroy[s.Name()] = true
+			destroyList := PruneSnapshots(tc.inputs, tc.rules)
+			destroySet := make(map[string]bool, len(destroyList))
+			for _, s := range destroyList {
+				destroySet[s.Name()] = true
 			}
-			if tc.expDestroyAlternatives == nil {
-				if tc.expDestroy == nil {
-					panic("must specify either expDestroyAlternatives or expDestroy")
-				}
-				tc.expDestroyAlternatives = []map[string]bool{tc.expDestroy}
-			}
-			var okAlt map[string]bool = nil
-			for _, alt := range tc.expDestroyAlternatives {
-				t.Logf("testing possible result: %v", alt)
-				if mapEqual(alt, tc.effDestroy) {
-					okAlt = alt
-				}
-			}
-			if okAlt == nil {
-				t.Errorf("no alternatives matched result: %v", tc.effDestroy)
+			t.Logf("destroySet:\n%#v", destroySet)
+			t.Logf("expected:\n%#v", tc.expDestroy)
+
+			require.Equal(t, len(tc.expDestroy), len(destroySet))
+			for name := range destroySet {
+				assert.True(t, tc.expDestroy[name], "%q", name)
 			}
 		})
 	}
@@ -94,6 +75,10 @@ func TestPruneSnapshots(t *testing.T) {
 		},
 	}

+	reltime := func(secs int64) time.Time {
+		return time.Unix(secs, 0)
+	}
+
 	tcs := map[string]testCase{
 		"simple": {
 			inputs: inputs["s1"],
@@ -140,6 +125,24 @@ func TestPruneSnapshots(t *testing.T) {
 			},
 			expDestroy: map[string]bool{},
 		},
+		"multiple_grids_with_disjoint_regexes": {
+			inputs: []Snapshot{
+				stubSnap{"p1_a", false, reltime(4)},
+				stubSnap{"p2_a", false, reltime(5)},
+				stubSnap{"p1_b", false, reltime(14)},
+				stubSnap{"p2_b", false, reltime(15)},
+				stubSnap{"p1_c", false, reltime(29)},
+				stubSnap{"p2_c", false, reltime(30)},
+			},
+			rules: []KeepRule{
+				MustNewKeepGrid("^p1_", `1x10s | 1x10s`),
+				MustNewKeepGrid("^p2_", `1x10s | 1x10s`),
+			},
+			expDestroy: map[string]bool{
+				"p1_a": true,
+				"p2_a": true,
+			},
+		},
 	}

 	testTable(tcs, t)
@@ -16,85 +16,127 @@ type Grid struct {
 	intervals []Interval
 }

-//A point inside the grid, i.e. a thing the grid can decide to remove
 type Entry interface {
 	Date() time.Time
-	LessThan(b Entry) bool
-}
-
-func dateInInterval(date, startDateInterval time.Time, i Interval) bool {
-	return date.After(startDateInterval) && date.Before(startDateInterval.Add(i.Length()))
 }

 func NewGrid(l []Interval) *Grid {
+	if len(l) == 0 {
+		panic("must specify at least one interval")
+	}
 	// TODO Maybe check for ascending interval lengths here, although the algorithm
 	// 		itself doesn't care about that.
 	return &Grid{l}
 }

-// Partition a list of RetentionGridEntries into the Grid,
-// relative to a given start date `now`.
-//
-// The `keepCount` oldest entries per `retentiongrid.Interval` are kept (`keep`),
-// the others are removed (`remove`).
-//
-// Entries that are younger than `now` are always kept.
-// Those that are older than the earliest beginning of an interval are removed.
-func (g Grid) FitEntries(now time.Time, entries []Entry) (keep, remove []Entry) {
+func (g Grid) FitEntries(entries []Entry) (keep, remove []Entry) {

-	type bucket struct {
-		entries []Entry
+	if len(entries) == 0 {
+		return
 	}
+
+	// determine 'now' based on youngest snapshot
+	// => sort youngest-to-oldest
+	sort.SliceStable(entries, func(i, j int) bool {
+		return entries[i].Date().After(entries[j].Date())
+	})
+	now := entries[0].Date()
+
+	return g.fitEntriesWithNow(now, entries)
+}
+
+type bucket struct {
+	keepCount     int
+	youngerThan   time.Time
+	olderThanOrEq time.Time
+	entries       []Entry
+}
+
+func makeBucketFromInterval(olderThanOrEq time.Time, i Interval) bucket {
+	var b bucket
+	kc := i.KeepCount()
+	if kc == 0 {
+		panic("keep count 0 is not allowed")
+	}
+	if (kc < 0) && kc != RetentionGridKeepCountAll {
+		panic("negative keep counts are not allowed")
+	}
+	b.keepCount = kc
+	b.olderThanOrEq = olderThanOrEq
+	b.youngerThan = b.olderThanOrEq.Add(-i.Length())
+	return b
+}
+
+func (b *bucket) Contains(e Entry) bool {
+	d := e.Date()
+	olderThan := d.Before(b.olderThanOrEq)
+	eq := d.Equal(b.olderThanOrEq)
+	youngerThan := d.After(b.youngerThan)
+	return (olderThan || eq) && youngerThan
+}
+
+func (b *bucket) AddIfContains(e Entry) (added bool) {
+	added = b.Contains(e)
+	if added {
+		b.entries = append(b.entries, e)
+	}
+	return
+}
+
+func (b *bucket) RemoveYoungerSnapsExceedingKeepCount() (removed []Entry) {
+
+	if b.keepCount == RetentionGridKeepCountAll {
+		return nil
+	}
+
+	removeCount := len(b.entries) - b.keepCount
+	if removeCount <= 0 {
+		return nil
+	}
+
+	// sort youngest-to-oldest
+	sort.SliceStable(b.entries, func(i, j int) bool {
+		return b.entries[i].Date().After(b.entries[j].Date())
+	})
+
+	return b.entries[:removeCount]
+}
+
+func (g Grid) fitEntriesWithNow(now time.Time, entries []Entry) (keep, remove []Entry) {
+
 	buckets := make([]bucket, len(g.intervals))

+	buckets[0] = makeBucketFromInterval(now, g.intervals[0])
+	for i := 1; i < len(g.intervals); i++ {
+		buckets[i] = makeBucketFromInterval(buckets[i-1].youngerThan, g.intervals[i])
+	}
+
 	keep = make([]Entry, 0)
 	remove = make([]Entry, 0)

-	oldestIntervalStart := now
-	for i := range g.intervals {
-		oldestIntervalStart = oldestIntervalStart.Add(-g.intervals[i].Length())
-	}
-
+assignEntriesToBuckets:
 	for ei := 0; ei < len(entries); ei++ {
 		e := entries[ei]
-
-		date := e.Date()
-
-		if date == now || date.After(now) {
+		// unconditionally keep entries that are in the future
+		if now.Before(e.Date()) {
 			keep = append(keep, e)
-			continue
-		} else if date.Before(oldestIntervalStart) {
-			remove = append(remove, e)
-			continue
+			continue assignEntriesToBuckets
 		}
-
-		iStartTime := now
-		for i := 0; i < len(g.intervals); i++ {
-			iStartTime = iStartTime.Add(-g.intervals[i].Length())
-			if date == iStartTime || dateInInterval(date, iStartTime, g.intervals[i]) {
-				buckets[i].entries = append(buckets[i].entries, e)
+		// add to matching bucket, if any
+		for bi := range buckets {
+			if buckets[bi].AddIfContains(e) {
+				continue assignEntriesToBuckets
 			}
 		}
+		// unconditionally remove entries older than the oldest bucket
+		remove = append(remove, e)
 	}

-	for bi, b := range buckets {
-
-		interval := g.intervals[bi]
-
-		sort.SliceStable(b.entries, func(i, j int) bool {
-			return b.entries[i].LessThan((b.entries[j]))
-		})
-
-		i := 0
-		for ; (interval.KeepCount() == RetentionGridKeepCountAll || i < interval.KeepCount()) && i < len(b.entries); i++ {
-			keep = append(keep, b.entries[i])
-		}
-		for ; i < len(b.entries); i++ {
-			remove = append(remove, b.entries[i])
-		}
-
+	// now apply the `KeepCount` per bucket
+	for _, b := range buckets {
+		destroy := b.RemoveYoungerSnapsExceedingKeepCount()
+		remove = append(remove, destroy...)
+		keep = append(keep, b.entries...)
 	}
-
 	return
-
 }
@@ -10,25 +10,18 @@ import (
 	"github.com/stretchr/testify/assert"
 )

-type retentionIntervalStub struct {
+type testInterval struct {
 	length    time.Duration
 	keepCount int
 }

-func (i *retentionIntervalStub) Length() time.Duration {
-	return i.length
-}
-
-func (i *retentionIntervalStub) KeepCount() int {
-	return i.keepCount
-}
+func (i *testInterval) Length() time.Duration { return i.length }
+func (i *testInterval) KeepCount() int        { return i.keepCount }

 func gridFromString(gs string) (g *Grid) {
-	intervals := strings.Split(gs, "|")
-	g = &Grid{
-		intervals: make([]Interval, len(intervals)),
-	}
-	for idx, i := range intervals {
+	sintervals := strings.Split(gs, "|")
+	intervals := make([]Interval, len(sintervals))
+	for idx, i := range sintervals {
 		comps := strings.SplitN(i, ",", 2)
 		var durationStr, numSnapsStr string
 		durationStr = comps[0]
@@ -39,7 +32,7 @@ func gridFromString(gs string) (g *Grid) {
 		}

 		var err error
-		var interval retentionIntervalStub
+		var interval testInterval

 		if interval.keepCount, err = strconv.Atoi(numSnapsStr); err != nil {
 			panic(err)
@@ -48,44 +41,38 @@ func gridFromString(gs string) (g *Grid) {
 			panic(err)
 		}

-		g.intervals[idx] = &interval
+		intervals[idx] = &interval
 	}
-	return
+	return NewGrid(intervals)
 }

-type dummySnap struct {
+type testSnap struct {
 	Name       string
 	ShouldKeep bool
 	date       time.Time
 }

-func (ds dummySnap) Date() time.Time {
-	return ds.date
-}
-
-func (ds dummySnap) LessThan(b Entry) bool {
-	return ds.date.Before(b.(dummySnap).date) // don't have a txg here
-}
+func (ds testSnap) Date() time.Time { return ds.date }

 func validateRetentionGridFitEntries(t *testing.T, now time.Time, input, keep, remove []Entry) {

-	snapDescr := func(d dummySnap) string {
+	snapDescr := func(d testSnap) string {
 		return fmt.Sprintf("%s@%s", d.Name, d.date.Sub(now))
 	}

 	t.Logf("keep list:\n")
 	for k := range keep {
-		t.Logf("\t%s\n", snapDescr(keep[k].(dummySnap)))
+		t.Logf("\t%s\n", snapDescr(keep[k].(testSnap)))
 	}
 	t.Logf("remove list:\n")
 	for k := range remove {
-		t.Logf("\t%s\n", snapDescr(remove[k].(dummySnap)))
+		t.Logf("\t%s\n", snapDescr(remove[k].(testSnap)))
 	}

 	t.Logf("\n\n")

 	for _, s := range input {
-		d := s.(dummySnap)
+		d := s.(testSnap)
 		descr := snapDescr(d)
 		t.Logf("testing %s\n", descr)
 		if d.ShouldKeep {
@@ -97,21 +84,18 @@ func validateRetentionGridFitEntries(t *testing.T, now time.Time, input, keep, r

 	t.Logf("resulting list:\n")
 	for k := range keep {
-		t.Logf("\t%s\n", snapDescr(keep[k].(dummySnap)))
+		t.Logf("\t%s\n", snapDescr(keep[k].(testSnap)))
 	}
 }

-func TestRetentionGridFitEntriesEmptyInput(t *testing.T) {
+func TestEmptyInput(t *testing.T) {
 	g := gridFromString("10m|10m|10m|1h")
-	keep, remove := g.FitEntries(time.Now(), []Entry{})
+	keep, remove := g.FitEntries([]Entry{})
 	assert.Empty(t, keep)
 	assert.Empty(t, remove)
 }

-func TestRetentionGridFitEntriesIntervalBoundariesAndAlignment(t *testing.T) {
-
-	// Intervals are (duration], i.e. 10min is in the first interval, not in the second
-
+func TestIntervalBoundariesAndAlignment(t *testing.T) {
 	g := gridFromString("10m|10m|10m")

 	t.Logf("%#v\n", g)
@@ -119,20 +103,51 @@ func TestRetentionGridFitEntriesIntervalBoundariesAndAlignment(t *testing.T) {
 	now := time.Unix(0, 0)

 	snaps := []Entry{
-		dummySnap{"0", true, now.Add(1 * time.Minute)},    // before now
-		dummySnap{"1", true, now},                         // before now
-		dummySnap{"2", true, now.Add(-10 * time.Minute)},  // 1st interval
-		dummySnap{"3", true, now.Add(-20 * time.Minute)},  // 2nd interval
-		dummySnap{"4", true, now.Add(-30 * time.Minute)},  // 3rd interval
-		dummySnap{"5", false, now.Add(-40 * time.Minute)}, // after last interval
+		testSnap{"0", true, now.Add(1 * time.Minute)},    // before now => keep unconditionally
+		testSnap{"1", true, now},                         // 1st interval left edge => inclusive
+		testSnap{"2", true, now.Add(-10 * time.Minute)},  // 2nd interval left edge => inclusive
+		testSnap{"3", true, now.Add(-20 * time.Minute)},  // 3rd interval left edge => inclusuive
+		testSnap{"4", false, now.Add(-30 * time.Minute)}, // 3rd interval right edge => excludive
+		testSnap{"5", false, now.Add(-40 * time.Minute)}, // after last interval => remove unconditionally
 	}

-	keep, remove := g.FitEntries(now, snaps)
+	keep, remove := g.fitEntriesWithNow(now, snaps)
 	validateRetentionGridFitEntries(t, now, snaps, keep, remove)
-
 }

-func TestRetentionGridFitEntries(t *testing.T) {
+func TestKeepsOldestSnapsInABucket(t *testing.T) {
+	g := gridFromString("1m,2")
+
+	relt := func(secs int64) time.Time { return time.Unix(secs, 0) }
+
+	snaps := []Entry{
+		testSnap{"1", true, relt(1)},
+		testSnap{"2", true, relt(2)},
+		testSnap{"3", false, relt(3)},
+		testSnap{"4", false, relt(4)},
+		testSnap{"5", false, relt(5)},
+	}
+
+	now := relt(6)
+	keep, remove := g.FitEntries(snaps)
+	validateRetentionGridFitEntries(t, now, snaps, keep, remove)
+}
+
+func TestRespectsKeepCountAll(t *testing.T) {
+	g := gridFromString("1m,-1|1m,1")
+	relt := func(secs int64) time.Time { return time.Unix(secs, 0) }
+	snaps := []Entry{
+		testSnap{"a", true, relt(0)},
+		testSnap{"b", true, relt(-1)},
+		testSnap{"c", true, relt(-2)},
+		testSnap{"d", false, relt(-60)},
+		testSnap{"e", true, relt(-61)},
+	}
+	keep, remove := g.FitEntries(snaps)
+	validateRetentionGridFitEntries(t, relt(61), snaps, keep, remove)
+}
+
+func TestComplex(t *testing.T) {

 	g := gridFromString("10m,-1|10m|10m,2|1h")

@@ -141,20 +156,29 @@ func TestRetentionGridFitEntries(t *testing.T) {
 	now := time.Unix(0, 0)

 	snaps := []Entry{
-		dummySnap{"1", true, now.Add(3 * time.Minute)},   // pre-now must always be kept
-		dummySnap{"b1", true, now.Add(-6 * time.Minute)}, // 1st interval allows unlimited entries
-		dummySnap{"b3", true, now.Add(-8 * time.Minute)}, // 1st interval allows unlimited entries
-		dummySnap{"b2", true, now.Add(-9 * time.Minute)}, // 1st interval allows unlimited entries
-		dummySnap{"a", false, now.Add(-11 * time.Minute)},
-		dummySnap{"c", true, now.Add(-19 * time.Minute)}, // 2nd interval allows 1 entry
-		dummySnap{"foo", false, now.Add(-25 * time.Minute)},
-		dummySnap{"bar", true, now.Add(-26 * time.Minute)}, // 3rd interval allows 2 entries
-		dummySnap{"border", true, now.Add(-30 * time.Minute)},
-		dummySnap{"d", true, now.Add(-1*time.Hour - 15*time.Minute)},
-		dummySnap{"e", false, now.Add(-1*time.Hour - 31*time.Minute)}, // before earliest interval must always be deleted
-		dummySnap{"f", false, now.Add(-2 * time.Hour)},
+		// pre-now must always be kept
+		testSnap{"1", true, now.Add(3 * time.Minute)},
+		// 1st interval allows unlimited entries
+		testSnap{"b1", true, now.Add(-6 * time.Minute)},
+		testSnap{"b3", true, now.Add(-8 * time.Minute)},
+		testSnap{"b2", true, now.Add(-9 * time.Minute)},
+		// 2nd interval allows 1 entry
+		testSnap{"a", false, now.Add(-11 * time.Minute)},
+		testSnap{"c", true, now.Add(-19 * time.Minute)},
+		// 3rd interval allows 2 entries
+		testSnap{"foo", true, now.Add(-25 * time.Minute)},
+		testSnap{"bar", true, now.Add(-26 * time.Minute)},
+		// this is at the left edge of the 4th interval
+		testSnap{"border", false, now.Add(-30 * time.Minute)},
+		// right in the 4th interval
+		testSnap{"d", true, now.Add(-1*time.Hour - 15*time.Minute)},
+		// on the right edge of 4th interval => not in it => delete
+		testSnap{"q", false, now.Add(-1*time.Hour - 30*time.Minute)},
+		// older then 4th interval => always delete
+		testSnap{"e", false, now.Add(-1*time.Hour - 31*time.Minute)},
+		testSnap{"f", false, now.Add(-2 * time.Hour)},
 	}
-	keep, remove := g.FitEntries(now, snaps)
+	keep, remove := g.fitEntriesWithNow(now, snaps)

 	validateRetentionGridFitEntries(t, now, snaps, keep, remove)

@@ -2,7 +2,6 @@ package driver

 import (
 	"context"
-	"errors"
 	"fmt"
 	"net"
 	"sort"
@@ -10,10 +9,14 @@ import (
 	"sync"
 	"time"

-	"github.com/zrepl/zrepl/daemon/logging/trace"
+	"github.com/kr/pretty"
+	"github.com/pkg/errors"
 	"google.golang.org/grpc/codes"
 	"google.golang.org/grpc/status"

+	"github.com/zrepl/zrepl/daemon/logging/trace"
+	"github.com/zrepl/zrepl/zfs"
+
 	"github.com/zrepl/zrepl/replication/report"
 	"github.com/zrepl/zrepl/util/chainlock"
 	"github.com/zrepl/zrepl/util/envconst"
@@ -227,7 +230,10 @@ func Do(ctx context.Context, planner Planner) (ReportFunc, WaitFunc) {
 			errRep := cur.errorReport()

 			if rep.State == report.AttemptDone {
-				log.Debug("attempt completed successfully")
+				if len(rep.Filesystems) == 0 {
+					log.Warn("no filesystems were considered for replication")
+				}
+				log.Debug("attempt completed")
 				break
 			}

@@ -235,7 +241,7 @@ func Do(ctx context.Context, planner Planner) (ReportFunc, WaitFunc) {
 			log.WithField("most_recent_err", mostRecentErr).WithField("most_recent_err_class", mostRecentErrClass).Debug("most recent error used for re-connect decision")
 			if mostRecentErr == nil {
 				// inconsistent reporting, let's bail out
-				log.Warn("attempt does not report done but error report does not report errors, aborting run")
+				log.WithField("attempt_state", rep.State).Warn("attempt does not report done but error report does not report errors, aborting run")
 				break
 			}
 			log.WithError(mostRecentErr.Err).Error("most recent error in this attempt")
@@ -309,6 +315,9 @@ func (a *attempt) doGlobalPlanning(ctx context.Context, prev *attempt) map[*fs]*
 		return nil
 	}

+	// a.fss != nil indicates that there was no planning error (see doc comment)
+	a.fss = make([]*fs, 0)
+
 	for _, pfs := range pfss {
 		fs := &fs{
 			fs: pfs,
@@ -374,11 +383,28 @@ func (a *attempt) doGlobalPlanning(ctx context.Context, prev *attempt) map[*fs]*
 	// invariant: prevs contains an entry for each unambiguous correspondence

 	// build up parent-child relationship (FIXME (O(n^2), but who's going to have that many filesystems...))
+	mustDatasetPathOrPlanFail := func(fs string) *zfs.DatasetPath {
+		dp, err := zfs.NewDatasetPath(fs)
+		if err != nil {
+			now := time.Now()
+			a.planErr = newTimedError(errors.Wrapf(err, "%q", fs), now)
+			a.fss = nil
+			a.finishedAt = now
+			return nil
+		}
+		return dp
+	}
 	for _, f1 := range a.fss {
-		fs1 := f1.fs.ReportInfo().Name
+		fs1 := mustDatasetPathOrPlanFail(f1.fs.ReportInfo().Name)
+		if fs1 == nil {
+			return nil
+		}
 		for _, f2 := range a.fss {
-			fs2 := f2.fs.ReportInfo().Name
-			if strings.HasPrefix(fs1, fs2) && fs1 != fs2 {
+			fs2 := mustDatasetPathOrPlanFail(f2.fs.ReportInfo().Name)
+			if fs2 == nil {
+				return nil
+			}
+			if fs1.HasPrefix(fs2) && !fs1.Equal(fs2) {
 				f1.initialRepOrd.parents = append(f1.initialRepOrd.parents, f2)
 				f2.initialRepOrd.children = append(f2.initialRepOrd.children, f1)
 			}
@@ -657,10 +683,12 @@ func (a *attempt) report() *report.AttemptReport {
 		r.Filesystems[i] = a.fss[i].report()
 	}

-	state := report.AttemptPlanning
-	if a.planErr != nil {
+	var state report.AttemptState
+	if a.planErr == nil && a.fss == nil {
+		state = report.AttemptPlanning
+	} else if a.planErr != nil && a.fss == nil {
 		state = report.AttemptPlanningError
-	} else if a.fss != nil {
+	} else if a.planErr == nil && a.fss != nil {
 		if a.finishedAt.IsZero() {
 			state = report.AttemptFanOutFSs
 		} else {
@@ -673,6 +701,8 @@ func (a *attempt) report() *report.AttemptReport {
 				state = report.AttemptFanOutError
 			}
 		}
+	} else {
+		panic(fmt.Sprintf("attempt.planErr and attempt.fss must not both be != nil:\n%s\n%s", pretty.Sprint(a.planErr), pretty.Sprint(a.fss)))
 	}
 	r.State = state

@@ -10,6 +10,7 @@ import (
 	"time"

 	"github.com/stretchr/testify/require"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"

 	"github.com/zrepl/zrepl/replication/report"
@@ -12,6 +12,7 @@ import (

 	"github.com/montanaflynn/stats"
 	"github.com/stretchr/testify/assert"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"
 )

@@ -77,56 +77,48 @@ func IncrementalPath(receiver, sender []*FilesystemVersion) (incPath []*Filesyst
 	receiver = SortVersionListByCreateTXGThenBookmarkLTSnapshot(receiver)
 	sender = SortVersionListByCreateTXGThenBookmarkLTSnapshot(sender)

-	// Find most recent common ancestor by name, preferring snapshots over bookmarks
+	var mrcaCandidate struct {
+		found bool
+		guid  uint64
+		r, s  int
+	}

-	mrcaRcv := len(receiver) - 1
-	mrcaSnd := len(sender) - 1
-
-	for mrcaRcv >= 0 && mrcaSnd >= 0 {
-		if receiver[mrcaRcv].Guid == sender[mrcaSnd].Guid {
-			// Since we arrive from the end of the array, and because we defined bookmark < snapshot,
-			// this condition will match snapshot first, which is what we want because it gives us
-			// size estimation
-			break
-		}
-		receiverCreation, err := receiver[mrcaRcv].CreationAsTime()
-		if err != nil {
-			panic(err) // FIXME move this to a sorting phase before
-		}
-		senderCreation, err := sender[mrcaSnd].CreationAsTime()
-		if err != nil {
-			panic(err) // FIXME move this to the sorting phase before
-		}
-
-		if receiverCreation.Before(senderCreation) {
-			mrcaSnd--
-		} else {
-			mrcaRcv--
+findCandidate:
+	for r := len(receiver) - 1; r >= 0; r-- {
+		for s := len(sender) - 1; s >= 0; s-- {
+			if sender[s].GetGuid() == receiver[r].GetGuid() {
+				mrcaCandidate.guid = sender[s].GetGuid()
+				mrcaCandidate.s = s
+				mrcaCandidate.r = r
+				mrcaCandidate.found = true
+				break findCandidate
+			}
 		}
 	}

-	if mrcaRcv == -1 || mrcaSnd == -1 {
+	// handle failure cases
+	if !mrcaCandidate.found {
 		return nil, &ConflictNoCommonAncestor{
 			SortedSenderVersions:   sender,
 			SortedReceiverVersions: receiver,
 		}
-	}
-
-	if mrcaRcv != len(receiver)-1 {
+	} else if mrcaCandidate.r != len(receiver)-1 {
 		return nil, &ConflictDiverged{
 			SortedSenderVersions:   sender,
 			SortedReceiverVersions: receiver,
-			CommonAncestor:         sender[mrcaSnd],
-			SenderOnly:             sender[mrcaSnd+1:],
-			ReceiverOnly:           receiver[mrcaRcv+1:],
+			CommonAncestor:         sender[mrcaCandidate.s],
+			SenderOnly:             sender[mrcaCandidate.s+1:],
+			ReceiverOnly:           receiver[mrcaCandidate.r+1:],
 		}
 	}

+	// incPath is possible
+
 	// incPath must not contain bookmarks except initial one,
 	incPath = make([]*FilesystemVersion, 0, len(sender))
-	incPath = append(incPath, sender[mrcaSnd])
+	incPath = append(incPath, sender[mrcaCandidate.s])
 	// it's ok if incPath[0] is a bookmark, but not the subsequent ones in the incPath
-	for i := mrcaSnd + 1; i < len(sender); i++ {
+	for i := mrcaCandidate.s + 1; i < len(sender); i++ {
 		if sender[i].Type == FilesystemVersion_Snapshot && incPath[len(incPath)-1].Guid != sender[i].Guid {
 			incPath = append(incPath, sender[i])
 		}
@@ -46,7 +46,36 @@ func (x Tri) String() string {
 	return proto.EnumName(Tri_name, int32(x))
 }
 func (Tri) EnumDescriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{0}
+	return fileDescriptor_pdu_616c27178643eca4, []int{0}
+}
+
+type ReplicationGuaranteeKind int32
+
+const (
+	ReplicationGuaranteeKind_GuaranteeInvalid                ReplicationGuaranteeKind = 0
+	ReplicationGuaranteeKind_GuaranteeResumability           ReplicationGuaranteeKind = 1
+	ReplicationGuaranteeKind_GuaranteeIncrementalReplication ReplicationGuaranteeKind = 2
+	ReplicationGuaranteeKind_GuaranteeNothing                ReplicationGuaranteeKind = 3
+)
+
+var ReplicationGuaranteeKind_name = map[int32]string{
+	0: "GuaranteeInvalid",
+	1: "GuaranteeResumability",
+	2: "GuaranteeIncrementalReplication",
+	3: "GuaranteeNothing",
+}
+var ReplicationGuaranteeKind_value = map[string]int32{
+	"GuaranteeInvalid":                0,
+	"GuaranteeResumability":           1,
+	"GuaranteeIncrementalReplication": 2,
+	"GuaranteeNothing":                3,
+}
+
+func (x ReplicationGuaranteeKind) String() string {
+	return proto.EnumName(ReplicationGuaranteeKind_name, int32(x))
+}
+func (ReplicationGuaranteeKind) EnumDescriptor() ([]byte, []int) {
+	return fileDescriptor_pdu_616c27178643eca4, []int{1}
 }

 type FilesystemVersion_VersionType int32
@@ -69,7 +98,7 @@ func (x FilesystemVersion_VersionType) String() string {
 	return proto.EnumName(FilesystemVersion_VersionType_name, int32(x))
 }
 func (FilesystemVersion_VersionType) EnumDescriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{5, 0}
+	return fileDescriptor_pdu_616c27178643eca4, []int{5, 0}
 }

 type ListFilesystemReq struct {
@@ -82,7 +111,7 @@ func (m *ListFilesystemReq) Reset()         { *m = ListFilesystemReq{} }
 func (m *ListFilesystemReq) String() string { return proto.CompactTextString(m) }
 func (*ListFilesystemReq) ProtoMessage()    {}
 func (*ListFilesystemReq) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{0}
+	return fileDescriptor_pdu_616c27178643eca4, []int{0}
 }
 func (m *ListFilesystemReq) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_ListFilesystemReq.Unmarshal(m, b)
@@ -113,7 +142,7 @@ func (m *ListFilesystemRes) Reset()         { *m = ListFilesystemRes{} }
 func (m *ListFilesystemRes) String() string { return proto.CompactTextString(m) }
 func (*ListFilesystemRes) ProtoMessage()    {}
 func (*ListFilesystemRes) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{1}
+	return fileDescriptor_pdu_616c27178643eca4, []int{1}
 }
 func (m *ListFilesystemRes) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_ListFilesystemRes.Unmarshal(m, b)
@@ -154,7 +183,7 @@ func (m *Filesystem) Reset()         { *m = Filesystem{} }
 func (m *Filesystem) String() string { return proto.CompactTextString(m) }
 func (*Filesystem) ProtoMessage()    {}
 func (*Filesystem) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{2}
+	return fileDescriptor_pdu_616c27178643eca4, []int{2}
 }
 func (m *Filesystem) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_Filesystem.Unmarshal(m, b)
@@ -213,7 +242,7 @@ func (m *ListFilesystemVersionsReq) Reset()         { *m = ListFilesystemVersion
 func (m *ListFilesystemVersionsReq) String() string { return proto.CompactTextString(m) }
 func (*ListFilesystemVersionsReq) ProtoMessage()    {}
 func (*ListFilesystemVersionsReq) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{3}
+	return fileDescriptor_pdu_616c27178643eca4, []int{3}
 }
 func (m *ListFilesystemVersionsReq) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_ListFilesystemVersionsReq.Unmarshal(m, b)
@@ -251,7 +280,7 @@ func (m *ListFilesystemVersionsRes) Reset()         { *m = ListFilesystemVersion
 func (m *ListFilesystemVersionsRes) String() string { return proto.CompactTextString(m) }
 func (*ListFilesystemVersionsRes) ProtoMessage()    {}
 func (*ListFilesystemVersionsRes) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{4}
+	return fileDescriptor_pdu_616c27178643eca4, []int{4}
 }
 func (m *ListFilesystemVersionsRes) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_ListFilesystemVersionsRes.Unmarshal(m, b)
@@ -293,7 +322,7 @@ func (m *FilesystemVersion) Reset()         { *m = FilesystemVersion{} }
 func (m *FilesystemVersion) String() string { return proto.CompactTextString(m) }
 func (*FilesystemVersion) ProtoMessage()    {}
 func (*FilesystemVersion) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{5}
+	return fileDescriptor_pdu_616c27178643eca4, []int{5}
 }
 func (m *FilesystemVersion) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_FilesystemVersion.Unmarshal(m, b)
@@ -359,19 +388,20 @@ type SendReq struct {
 	// SHOULD clear the resume token on their side and use From and To instead If
 	// ResumeToken is not empty, the GUIDs of From and To MUST correspond to those
 	// encoded in the ResumeToken. Otherwise, the Sender MUST return an error.
-	ResumeToken          string   `protobuf:"bytes,4,opt,name=ResumeToken,proto3" json:"ResumeToken,omitempty"`
-	Encrypted            Tri      `protobuf:"varint,5,opt,name=Encrypted,proto3,enum=Tri" json:"Encrypted,omitempty"`
-	DryRun               bool     `protobuf:"varint,6,opt,name=DryRun,proto3" json:"DryRun,omitempty"`
-	XXX_NoUnkeyedLiteral struct{} `json:"-"`
-	XXX_unrecognized     []byte   `json:"-"`
-	XXX_sizecache        int32    `json:"-"`
+	ResumeToken          string             `protobuf:"bytes,4,opt,name=ResumeToken,proto3" json:"ResumeToken,omitempty"`
+	Encrypted            Tri                `protobuf:"varint,5,opt,name=Encrypted,proto3,enum=Tri" json:"Encrypted,omitempty"`
+	DryRun               bool               `protobuf:"varint,6,opt,name=DryRun,proto3" json:"DryRun,omitempty"`
+	ReplicationConfig    *ReplicationConfig `protobuf:"bytes,7,opt,name=ReplicationConfig,proto3" json:"ReplicationConfig,omitempty"`
+	XXX_NoUnkeyedLiteral struct{}           `json:"-"`
+	XXX_unrecognized     []byte             `json:"-"`
+	XXX_sizecache        int32              `json:"-"`
 }

 func (m *SendReq) Reset()         { *m = SendReq{} }
 func (m *SendReq) String() string { return proto.CompactTextString(m) }
 func (*SendReq) ProtoMessage()    {}
 func (*SendReq) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{6}
+	return fileDescriptor_pdu_616c27178643eca4, []int{6}
 }
 func (m *SendReq) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_SendReq.Unmarshal(m, b)
@@ -433,6 +463,97 @@ func (m *SendReq) GetDryRun() bool {
 	return false
 }

+func (m *SendReq) GetReplicationConfig() *ReplicationConfig {
+	if m != nil {
+		return m.ReplicationConfig
+	}
+	return nil
+}
+
+type ReplicationConfig struct {
+	Protection           *ReplicationConfigProtection `protobuf:"bytes,1,opt,name=protection,proto3" json:"protection,omitempty"`
+	XXX_NoUnkeyedLiteral struct{}                     `json:"-"`
+	XXX_unrecognized     []byte                       `json:"-"`
+	XXX_sizecache        int32                        `json:"-"`
+}
+
+func (m *ReplicationConfig) Reset()         { *m = ReplicationConfig{} }
+func (m *ReplicationConfig) String() string { return proto.CompactTextString(m) }
+func (*ReplicationConfig) ProtoMessage()    {}
+func (*ReplicationConfig) Descriptor() ([]byte, []int) {
+	return fileDescriptor_pdu_616c27178643eca4, []int{7}
+}
+func (m *ReplicationConfig) XXX_Unmarshal(b []byte) error {
+	return xxx_messageInfo_ReplicationConfig.Unmarshal(m, b)
+}
+func (m *ReplicationConfig) XXX_Marshal(b []byte, deterministic bool) ([]byte, error) {
+	return xxx_messageInfo_ReplicationConfig.Marshal(b, m, deterministic)
+}
+func (dst *ReplicationConfig) XXX_Merge(src proto.Message) {
+	xxx_messageInfo_ReplicationConfig.Merge(dst, src)
+}
+func (m *ReplicationConfig) XXX_Size() int {
+	return xxx_messageInfo_ReplicationConfig.Size(m)
+}
+func (m *ReplicationConfig) XXX_DiscardUnknown() {
+	xxx_messageInfo_ReplicationConfig.DiscardUnknown(m)
+}
+
+var xxx_messageInfo_ReplicationConfig proto.InternalMessageInfo
+
+func (m *ReplicationConfig) GetProtection() *ReplicationConfigProtection {
+	if m != nil {
+		return m.Protection
+	}
+	return nil
+}
+
+type ReplicationConfigProtection struct {
+	Initial              ReplicationGuaranteeKind `protobuf:"varint,1,opt,name=Initial,proto3,enum=ReplicationGuaranteeKind" json:"Initial,omitempty"`
+	Incremental          ReplicationGuaranteeKind `protobuf:"varint,2,opt,name=Incremental,proto3,enum=ReplicationGuaranteeKind" json:"Incremental,omitempty"`
+	XXX_NoUnkeyedLiteral struct{}                 `json:"-"`
+	XXX_unrecognized     []byte                   `json:"-"`
+	XXX_sizecache        int32                    `json:"-"`
+}
+
+func (m *ReplicationConfigProtection) Reset()         { *m = ReplicationConfigProtection{} }
+func (m *ReplicationConfigProtection) String() string { return proto.CompactTextString(m) }
+func (*ReplicationConfigProtection) ProtoMessage()    {}
+func (*ReplicationConfigProtection) Descriptor() ([]byte, []int) {
+	return fileDescriptor_pdu_616c27178643eca4, []int{8}
+}
+func (m *ReplicationConfigProtection) XXX_Unmarshal(b []byte) error {
+	return xxx_messageInfo_ReplicationConfigProtection.Unmarshal(m, b)
+}
+func (m *ReplicationConfigProtection) XXX_Marshal(b []byte, deterministic bool) ([]byte, error) {
+	return xxx_messageInfo_ReplicationConfigProtection.Marshal(b, m, deterministic)
+}
+func (dst *ReplicationConfigProtection) XXX_Merge(src proto.Message) {
+	xxx_messageInfo_ReplicationConfigProtection.Merge(dst, src)
+}
+func (m *ReplicationConfigProtection) XXX_Size() int {
+	return xxx_messageInfo_ReplicationConfigProtection.Size(m)
+}
+func (m *ReplicationConfigProtection) XXX_DiscardUnknown() {
+	xxx_messageInfo_ReplicationConfigProtection.DiscardUnknown(m)
+}
+
+var xxx_messageInfo_ReplicationConfigProtection proto.InternalMessageInfo
+
+func (m *ReplicationConfigProtection) GetInitial() ReplicationGuaranteeKind {
+	if m != nil {
+		return m.Initial
+	}
+	return ReplicationGuaranteeKind_GuaranteeInvalid
+}
+
+func (m *ReplicationConfigProtection) GetIncremental() ReplicationGuaranteeKind {
+	if m != nil {
+		return m.Incremental
+	}
+	return ReplicationGuaranteeKind_GuaranteeInvalid
+}
+
 type Property struct {
 	Name                 string   `protobuf:"bytes,1,opt,name=Name,proto3" json:"Name,omitempty"`
 	Value                string   `protobuf:"bytes,2,opt,name=Value,proto3" json:"Value,omitempty"`
@@ -445,7 +566,7 @@ func (m *Property) Reset()         { *m = Property{} }
 func (m *Property) String() string { return proto.CompactTextString(m) }
 func (*Property) ProtoMessage()    {}
 func (*Property) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{7}
+	return fileDescriptor_pdu_616c27178643eca4, []int{9}
 }
 func (m *Property) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_Property.Unmarshal(m, b)
@@ -496,7 +617,7 @@ func (m *SendRes) Reset()         { *m = SendRes{} }
 func (m *SendRes) String() string { return proto.CompactTextString(m) }
 func (*SendRes) ProtoMessage()    {}
 func (*SendRes) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{8}
+	return fileDescriptor_pdu_616c27178643eca4, []int{10}
 }
 func (m *SendRes) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_SendRes.Unmarshal(m, b)
@@ -548,7 +669,7 @@ func (m *SendCompletedReq) Reset()         { *m = SendCompletedReq{} }
 func (m *SendCompletedReq) String() string { return proto.CompactTextString(m) }
 func (*SendCompletedReq) ProtoMessage()    {}
 func (*SendCompletedReq) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{9}
+	return fileDescriptor_pdu_616c27178643eca4, []int{11}
 }
 func (m *SendCompletedReq) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_SendCompletedReq.Unmarshal(m, b)
@@ -585,7 +706,7 @@ func (m *SendCompletedRes) Reset()         { *m = SendCompletedRes{} }
 func (m *SendCompletedRes) String() string { return proto.CompactTextString(m) }
 func (*SendCompletedRes) ProtoMessage()    {}
 func (*SendCompletedRes) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{10}
+	return fileDescriptor_pdu_616c27178643eca4, []int{12}
 }
 func (m *SendCompletedRes) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_SendCompletedRes.Unmarshal(m, b)
@@ -610,17 +731,18 @@ type ReceiveReq struct {
 	To         *FilesystemVersion `protobuf:"bytes,2,opt,name=To,proto3" json:"To,omitempty"`
 	// If true, the receiver should clear the resume token before performing the
 	// zfs recv of the stream in the request
-	ClearResumeToken     bool     `protobuf:"varint,3,opt,name=ClearResumeToken,proto3" json:"ClearResumeToken,omitempty"`
-	XXX_NoUnkeyedLiteral struct{} `json:"-"`
-	XXX_unrecognized     []byte   `json:"-"`
-	XXX_sizecache        int32    `json:"-"`
+	ClearResumeToken     bool               `protobuf:"varint,3,opt,name=ClearResumeToken,proto3" json:"ClearResumeToken,omitempty"`
+	ReplicationConfig    *ReplicationConfig `protobuf:"bytes,4,opt,name=ReplicationConfig,proto3" json:"ReplicationConfig,omitempty"`
+	XXX_NoUnkeyedLiteral struct{}           `json:"-"`
+	XXX_unrecognized     []byte             `json:"-"`
+	XXX_sizecache        int32              `json:"-"`
 }

 func (m *ReceiveReq) Reset()         { *m = ReceiveReq{} }
 func (m *ReceiveReq) String() string { return proto.CompactTextString(m) }
 func (*ReceiveReq) ProtoMessage()    {}
 func (*ReceiveReq) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{11}
+	return fileDescriptor_pdu_616c27178643eca4, []int{13}
 }
 func (m *ReceiveReq) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_ReceiveReq.Unmarshal(m, b)
@@ -661,6 +783,13 @@ func (m *ReceiveReq) GetClearResumeToken() bool {
 	return false
 }

+func (m *ReceiveReq) GetReplicationConfig() *ReplicationConfig {
+	if m != nil {
+		return m.ReplicationConfig
+	}
+	return nil
+}
+
 type ReceiveRes struct {
 	XXX_NoUnkeyedLiteral struct{} `json:"-"`
 	XXX_unrecognized     []byte   `json:"-"`
@@ -671,7 +800,7 @@ func (m *ReceiveRes) Reset()         { *m = ReceiveRes{} }
 func (m *ReceiveRes) String() string { return proto.CompactTextString(m) }
 func (*ReceiveRes) ProtoMessage()    {}
 func (*ReceiveRes) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{12}
+	return fileDescriptor_pdu_616c27178643eca4, []int{14}
 }
 func (m *ReceiveRes) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_ReceiveRes.Unmarshal(m, b)
@@ -704,7 +833,7 @@ func (m *DestroySnapshotsReq) Reset()         { *m = DestroySnapshotsReq{} }
 func (m *DestroySnapshotsReq) String() string { return proto.CompactTextString(m) }
 func (*DestroySnapshotsReq) ProtoMessage()    {}
 func (*DestroySnapshotsReq) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{13}
+	return fileDescriptor_pdu_616c27178643eca4, []int{15}
 }
 func (m *DestroySnapshotsReq) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_DestroySnapshotsReq.Unmarshal(m, b)
@@ -750,7 +879,7 @@ func (m *DestroySnapshotRes) Reset()         { *m = DestroySnapshotRes{} }
 func (m *DestroySnapshotRes) String() string { return proto.CompactTextString(m) }
 func (*DestroySnapshotRes) ProtoMessage()    {}
 func (*DestroySnapshotRes) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{14}
+	return fileDescriptor_pdu_616c27178643eca4, []int{16}
 }
 func (m *DestroySnapshotRes) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_DestroySnapshotRes.Unmarshal(m, b)
@@ -795,7 +924,7 @@ func (m *DestroySnapshotsRes) Reset()         { *m = DestroySnapshotsRes{} }
 func (m *DestroySnapshotsRes) String() string { return proto.CompactTextString(m) }
 func (*DestroySnapshotsRes) ProtoMessage()    {}
 func (*DestroySnapshotsRes) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{15}
+	return fileDescriptor_pdu_616c27178643eca4, []int{17}
 }
 func (m *DestroySnapshotsRes) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_DestroySnapshotsRes.Unmarshal(m, b)
@@ -833,7 +962,7 @@ func (m *ReplicationCursorReq) Reset()         { *m = ReplicationCursorReq{} }
 func (m *ReplicationCursorReq) String() string { return proto.CompactTextString(m) }
 func (*ReplicationCursorReq) ProtoMessage()    {}
 func (*ReplicationCursorReq) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{16}
+	return fileDescriptor_pdu_616c27178643eca4, []int{18}
 }
 func (m *ReplicationCursorReq) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_ReplicationCursorReq.Unmarshal(m, b)
@@ -874,7 +1003,7 @@ func (m *ReplicationCursorRes) Reset()         { *m = ReplicationCursorRes{} }
 func (m *ReplicationCursorRes) String() string { return proto.CompactTextString(m) }
 func (*ReplicationCursorRes) ProtoMessage()    {}
 func (*ReplicationCursorRes) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{17}
+	return fileDescriptor_pdu_616c27178643eca4, []int{19}
 }
 func (m *ReplicationCursorRes) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_ReplicationCursorRes.Unmarshal(m, b)
@@ -1010,7 +1139,7 @@ func (m *PingReq) Reset()         { *m = PingReq{} }
 func (m *PingReq) String() string { return proto.CompactTextString(m) }
 func (*PingReq) ProtoMessage()    {}
 func (*PingReq) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{18}
+	return fileDescriptor_pdu_616c27178643eca4, []int{20}
 }
 func (m *PingReq) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_PingReq.Unmarshal(m, b)
@@ -1049,7 +1178,7 @@ func (m *PingRes) Reset()         { *m = PingRes{} }
 func (m *PingRes) String() string { return proto.CompactTextString(m) }
 func (*PingRes) ProtoMessage()    {}
 func (*PingRes) Descriptor() ([]byte, []int) {
-	return fileDescriptor_pdu_483c6918b7b3d747, []int{19}
+	return fileDescriptor_pdu_616c27178643eca4, []int{21}
 }
 func (m *PingRes) XXX_Unmarshal(b []byte) error {
 	return xxx_messageInfo_PingRes.Unmarshal(m, b)
@@ -1084,6 +1213,8 @@ func init() {
 	proto.RegisterType((*ListFilesystemVersionsRes)(nil), "ListFilesystemVersionsRes")
 	proto.RegisterType((*FilesystemVersion)(nil), "FilesystemVersion")
 	proto.RegisterType((*SendReq)(nil), "SendReq")
+	proto.RegisterType((*ReplicationConfig)(nil), "ReplicationConfig")
+	proto.RegisterType((*ReplicationConfigProtection)(nil), "ReplicationConfigProtection")
 	proto.RegisterType((*Property)(nil), "Property")
 	proto.RegisterType((*SendRes)(nil), "SendRes")
 	proto.RegisterType((*SendCompletedReq)(nil), "SendCompletedReq")
@@ -1098,6 +1229,7 @@ func init() {
 	proto.RegisterType((*PingReq)(nil), "PingReq")
 	proto.RegisterType((*PingRes)(nil), "PingRes")
 	proto.RegisterEnum("Tri", Tri_name, Tri_value)
+	proto.RegisterEnum("ReplicationGuaranteeKind", ReplicationGuaranteeKind_name, ReplicationGuaranteeKind_value)
 	proto.RegisterEnum("FilesystemVersion_VersionType", FilesystemVersion_VersionType_name, FilesystemVersion_VersionType_value)
 }

@@ -1338,61 +1470,71 @@ var _Replication_serviceDesc = grpc.ServiceDesc{
 	Metadata: "pdu.proto",
 }

-func init() { proto.RegisterFile("pdu.proto", fileDescriptor_pdu_483c6918b7b3d747) }
+func init() { proto.RegisterFile("pdu.proto", fileDescriptor_pdu_616c27178643eca4) }

-var fileDescriptor_pdu_483c6918b7b3d747 = []byte{
-	// 833 bytes of a gzipped FileDescriptorProto
-	0x1f, 0x8b, 0x08, 0x00, 0x00, 0x00, 0x00, 0x00, 0x02, 0xff, 0x8c, 0x56, 0x5f, 0x6f, 0xe3, 0x44,
-	0x10, 0xaf, 0x13, 0xa7, 0x75, 0x26, 0x3d, 0x2e, 0x9d, 0x96, 0x93, 0xb1, 0xe0, 0x54, 0x2d, 0x08,
-	0xe5, 0x2a, 0x61, 0xa1, 0xf2, 0x47, 0x42, 0x48, 0x27, 0xd1, 0xb4, 0xbd, 0x3b, 0x01, 0x47, 0xb4,
-	0x35, 0x27, 0x74, 0x6f, 0x26, 0x19, 0xb5, 0x56, 0x1d, 0xaf, 0xbb, 0xe3, 0xa0, 0x0b, 0xe2, 0x89,
-	0x47, 0xbe, 0x1e, 0x7c, 0x10, 0x3e, 0x02, 0xf2, 0xc6, 0x4e, 0x9c, 0xd8, 0x41, 0x79, 0xca, 0xce,
-	0x6f, 0x66, 0x77, 0x67, 0x7f, 0xf3, 0x9b, 0x71, 0xa0, 0x9b, 0x4e, 0x66, 0x7e, 0xaa, 0x55, 0xa6,
-	0xc4, 0x31, 0x1c, 0xfd, 0x10, 0x71, 0x76, 0x1d, 0xc5, 0xc4, 0x73, 0xce, 0x68, 0x2a, 0xe9, 0x41,
-	0x5c, 0xd4, 0x41, 0xc6, 0xcf, 0xa0, 0xb7, 0x02, 0xd8, 0xb5, 0x4e, 0xdb, 0x83, 0xde, 0x79, 0xcf,
-	0xaf, 0x04, 0x55, 0xfd, 0xe2, 0x2f, 0x0b, 0x60, 0x65, 0x23, 0x82, 0x3d, 0x0a, 0xb3, 0x3b, 0xd7,
-	0x3a, 0xb5, 0x06, 0x5d, 0x69, 0xd6, 0x78, 0x0a, 0x3d, 0x49, 0x3c, 0x9b, 0x52, 0xa0, 0xee, 0x29,
-	0x71, 0x5b, 0xc6, 0x55, 0x85, 0xf0, 0x13, 0x78, 0xf4, 0x8a, 0x47, 0x71, 0x38, 0xa6, 0x3b, 0x15,
-	0x4f, 0x48, 0xbb, 0xed, 0x53, 0x6b, 0xe0, 0xc8, 0x75, 0x30, 0x3f, 0xe7, 0x15, 0x5f, 0x25, 0x63,
-	0x3d, 0x4f, 0x33, 0x9a, 0xb8, 0xb6, 0x89, 0xa9, 0x42, 0xe2, 0x5b, 0xf8, 0x60, 0xfd, 0x41, 0x6f,
-	0x48, 0x73, 0xa4, 0x12, 0x96, 0xf4, 0x80, 0x4f, 0xab, 0x89, 0x16, 0x09, 0x56, 0x10, 0xf1, 0xfd,
-	0xf6, 0xcd, 0x8c, 0x3e, 0x38, 0xa5, 0x59, 0x50, 0x82, 0x7e, 0x2d, 0x52, 0x2e, 0x63, 0xc4, 0x3f,
-	0x16, 0x1c, 0xd5, 0xfc, 0x78, 0x0e, 0x76, 0x30, 0x4f, 0xc9, 0x5c, 0xfe, 0xde, 0xf9, 0xd3, 0xfa,
-	0x09, 0x7e, 0xf1, 0x9b, 0x47, 0x49, 0x13, 0x9b, 0x33, 0xfa, 0x3a, 0x9c, 0x52, 0x41, 0x9b, 0x59,
-	0xe7, 0xd8, 0x8b, 0x59, 0x34, 0x31, 0x34, 0xd9, 0xd2, 0xac, 0xf1, 0x43, 0xe8, 0x0e, 0x35, 0x85,
-	0x19, 0x05, 0xbf, 0xbc, 0x30, 0xdc, 0xd8, 0x72, 0x05, 0xa0, 0x07, 0x8e, 0x31, 0x22, 0x95, 0xb8,
-	0x1d, 0x73, 0xd2, 0xd2, 0x16, 0xcf, 0xa0, 0x57, 0xb9, 0x16, 0x0f, 0xc1, 0xb9, 0x49, 0xc2, 0x94,
-	0xef, 0x54, 0xd6, 0xdf, 0xcb, 0xad, 0x0b, 0xa5, 0xee, 0xa7, 0xa1, 0xbe, 0xef, 0x5b, 0xe2, 0x6f,
-	0x0b, 0x0e, 0x6e, 0x28, 0x99, 0xec, 0xc0, 0x27, 0x7e, 0x0a, 0xf6, 0xb5, 0x56, 0x53, 0x93, 0x78,
-	0x33, 0x5d, 0xc6, 0x8f, 0x02, 0x5a, 0x81, 0x32, 0x4f, 0x69, 0x8e, 0x6a, 0x05, 0x6a, 0x53, 0x42,
-	0x76, 0x5d, 0x42, 0x02, 0xba, 0x2b, 0x69, 0x74, 0x0c, 0xbf, 0xb6, 0x1f, 0xe8, 0x48, 0xae, 0x60,
-	0x7c, 0x02, 0xfb, 0x97, 0x7a, 0x2e, 0x67, 0x89, 0xbb, 0x6f, 0xb4, 0x53, 0x58, 0xe2, 0x4b, 0x70,
-	0x46, 0x5a, 0xa5, 0xa4, 0xb3, 0xf9, 0x92, 0x6e, 0xab, 0x42, 0xf7, 0x09, 0x74, 0xde, 0x84, 0xf1,
-	0xac, 0xac, 0xc1, 0xc2, 0x10, 0x7f, 0x2e, 0xb9, 0x60, 0x1c, 0xc0, 0xe3, 0x9f, 0x99, 0x26, 0x9b,
-	0x32, 0x77, 0xe4, 0x26, 0x8c, 0x02, 0x0e, 0xaf, 0xde, 0xa5, 0x34, 0xce, 0x68, 0x72, 0x13, 0xfd,
-	0x4e, 0xe6, 0xdd, 0x6d, 0xb9, 0x86, 0xe1, 0x33, 0x80, 0x22, 0x9f, 0x88, 0xd8, 0xb5, 0x8d, 0xdc,
-	0xba, 0x7e, 0x99, 0xa2, 0xac, 0x38, 0xc5, 0x73, 0xe8, 0xe7, 0x39, 0x0c, 0xd5, 0x34, 0x8d, 0x29,
-	0x23, 0x53, 0x98, 0x33, 0xe8, 0xfd, 0xa4, 0xa3, 0xdb, 0x28, 0x09, 0x63, 0x49, 0x0f, 0x05, 0xff,
-	0x8e, 0x5f, 0xd4, 0x4d, 0x56, 0x9d, 0x02, 0x6b, 0xfb, 0x59, 0xfc, 0x01, 0x20, 0x69, 0x4c, 0xd1,
-	0x6f, 0xb4, 0x4b, 0x99, 0x17, 0xe5, 0x6b, 0xfd, 0x6f, 0xf9, 0xce, 0xa0, 0x3f, 0x8c, 0x29, 0xd4,
-	0x55, 0x7e, 0x16, 0x2d, 0x5e, 0xc3, 0xc5, 0x61, 0xe5, 0x76, 0x16, 0xb7, 0x70, 0x7c, 0x49, 0x9c,
-	0x69, 0x35, 0x2f, 0x35, 0xb9, 0x4b, 0x2f, 0xe3, 0xe7, 0xd0, 0x5d, 0xc6, 0xbb, 0xad, 0xad, 0xfd,
-	0xba, 0x0a, 0x12, 0x6f, 0x01, 0x37, 0x2e, 0x2a, 0xda, 0xbe, 0x34, 0xcd, 0x2d, 0x5b, 0xda, 0xbe,
-	0x8c, 0xc9, 0x95, 0x72, 0xa5, 0xb5, 0xd2, 0xa5, 0x52, 0x8c, 0x21, 0x2e, 0x9b, 0x1e, 0x91, 0x4f,
-	0xda, 0x83, 0xfc, 0xe1, 0x71, 0x56, 0x8e, 0x94, 0x63, 0xbf, 0x9e, 0x82, 0x2c, 0x63, 0xc4, 0xd7,
-	0x70, 0x22, 0x29, 0x8d, 0xa3, 0xb1, 0xe9, 0xda, 0xe1, 0x4c, 0xb3, 0xd2, 0xbb, 0xcc, 0xb5, 0xa0,
-	0x71, 0x1f, 0xe3, 0x49, 0x31, 0x44, 0xf2, 0x1d, 0xf6, 0xcb, 0xbd, 0xe5, 0x18, 0x71, 0x5e, 0xab,
-	0x8c, 0xde, 0x45, 0x9c, 0x2d, 0x24, 0xfc, 0x72, 0x4f, 0x2e, 0x91, 0x0b, 0x07, 0xf6, 0x17, 0xe9,
-	0x88, 0x8f, 0xe1, 0x60, 0x14, 0x25, 0xb7, 0x79, 0x02, 0x2e, 0x1c, 0xfc, 0x48, 0xcc, 0xe1, 0x6d,
-	0xd9, 0x35, 0xa5, 0x29, 0x3e, 0x2a, 0x83, 0x38, 0xef, 0xab, 0xab, 0xf1, 0x9d, 0x2a, 0xfb, 0x2a,
-	0x5f, 0x9f, 0x0d, 0xa0, 0x1d, 0xe8, 0x28, 0x1f, 0x31, 0x97, 0x2a, 0xc9, 0x86, 0xa1, 0xa6, 0xfe,
-	0x1e, 0x76, 0xa1, 0x73, 0x1d, 0xc6, 0x4c, 0x7d, 0x0b, 0x1d, 0xb0, 0x03, 0x3d, 0xa3, 0x7e, 0xeb,
-	0xfc, 0xdf, 0x56, 0x3e, 0x00, 0x96, 0x8f, 0x40, 0x0f, 0xec, 0xfc, 0x60, 0x74, 0xfc, 0x22, 0x09,
-	0xaf, 0x5c, 0x31, 0x7e, 0x03, 0x8f, 0xd7, 0xe7, 0x38, 0x23, 0xfa, 0xb5, 0x8f, 0x9f, 0x57, 0xc7,
-	0x18, 0x47, 0xf0, 0xa4, 0xf9, 0x13, 0x80, 0x9e, 0xbf, 0xf5, 0xc3, 0xe2, 0x6d, 0xf7, 0x31, 0x3e,
-	0x87, 0xfe, 0x66, 0xe9, 0xf1, 0xc4, 0x6f, 0x90, 0xb4, 0xd7, 0x84, 0x32, 0x7e, 0x07, 0x47, 0xb5,
-	0xe2, 0xe1, 0xfb, 0x7e, 0x93, 0x10, 0xbc, 0x46, 0x98, 0xf1, 0x2b, 0x78, 0xb4, 0xd6, 0xe2, 0x78,
-	0xe4, 0x6f, 0x8e, 0x0c, 0xaf, 0x06, 0xf1, 0x45, 0xe7, 0x6d, 0x3b, 0x9d, 0xcc, 0x7e, 0xdd, 0x37,
-	0xff, 0x1f, 0xbe, 0xf8, 0x2f, 0x00, 0x00, 0xff, 0xff, 0x27, 0x95, 0xc1, 0x78, 0x4c, 0x08, 0x00,
-	0x00,
+var fileDescriptor_pdu_616c27178643eca4 = []byte{
+	// 995 bytes of a gzipped FileDescriptorProto
+	0x1f, 0x8b, 0x08, 0x00, 0x00, 0x00, 0x00, 0x00, 0x02, 0xff, 0x94, 0x56, 0xcf, 0x6e, 0xdb, 0xc6,
+	0x13, 0x36, 0x25, 0xda, 0xa2, 0x46, 0xce, 0x2f, 0xf4, 0xd8, 0x09, 0x68, 0xfd, 0xd2, 0xd4, 0xd8,
+	0x14, 0x85, 0x63, 0xa0, 0x44, 0xe1, 0xb4, 0x05, 0x8a, 0x14, 0x41, 0xeb, 0xbf, 0x31, 0xd2, 0xba,
+	0xea, 0x5a, 0x0d, 0x8a, 0xdc, 0x18, 0x69, 0x2a, 0x2f, 0x4c, 0x71, 0xe9, 0x5d, 0x2a, 0x88, 0x7a,
+	0xec, 0xa1, 0x87, 0x5e, 0x7a, 0xea, 0xeb, 0xf4, 0x29, 0xfa, 0x20, 0x7d, 0x84, 0x82, 0x6b, 0x92,
+	0xa2, 0x44, 0xca, 0x70, 0x4f, 0xda, 0xf9, 0xe6, 0xdb, 0xd9, 0xd9, 0xd1, 0x37, 0xb3, 0x84, 0x76,
+	0x3c, 0x9c, 0xf8, 0xb1, 0x92, 0x89, 0x64, 0x9b, 0xb0, 0xf1, 0xad, 0xd0, 0xc9, 0x89, 0x08, 0x49,
+	0x4f, 0x75, 0x42, 0x63, 0x4e, 0xd7, 0xec, 0xa0, 0x0a, 0x6a, 0xfc, 0x04, 0x3a, 0x33, 0x40, 0x7b,
+	0xd6, 0x4e, 0x73, 0xb7, 0xb3, 0xdf, 0xf1, 0x4b, 0xa4, 0xb2, 0x9f, 0xfd, 0x6e, 0x01, 0xcc, 0x6c,
+	0x44, 0xb0, 0x7b, 0x41, 0x72, 0xe9, 0x59, 0x3b, 0xd6, 0x6e, 0x9b, 0x9b, 0x35, 0xee, 0x40, 0x87,
+	0x93, 0x9e, 0x8c, 0xa9, 0x2f, 0xaf, 0x28, 0xf2, 0x1a, 0xc6, 0x55, 0x86, 0xf0, 0x23, 0xb8, 0x77,
+	0xa6, 0x7b, 0x61, 0x30, 0xa0, 0x4b, 0x19, 0x0e, 0x49, 0x79, 0xcd, 0x1d, 0x6b, 0xd7, 0xe1, 0xf3,
+	0x60, 0x1a, 0xe7, 0x4c, 0x1f, 0x47, 0x03, 0x35, 0x8d, 0x13, 0x1a, 0x7a, 0xb6, 0xe1, 0x94, 0x21,
+	0xf6, 0x1c, 0xb6, 0xe7, 0x2f, 0xf4, 0x9a, 0x94, 0x16, 0x32, 0xd2, 0x9c, 0xae, 0xf1, 0x71, 0x39,
+	0xd1, 0x2c, 0xc1, 0x12, 0xc2, 0x5e, 0x2d, 0xdf, 0xac, 0xd1, 0x07, 0x27, 0x37, 0xb3, 0x92, 0xa0,
+	0x5f, 0x61, 0xf2, 0x82, 0xc3, 0xfe, 0xb6, 0x60, 0xa3, 0xe2, 0xc7, 0x7d, 0xb0, 0xfb, 0xd3, 0x98,
+	0xcc, 0xe1, 0xff, 0xdb, 0x7f, 0x5c, 0x8d, 0xe0, 0x67, 0xbf, 0x29, 0x8b, 0x1b, 0x6e, 0x5a, 0xd1,
+	0xf3, 0x60, 0x4c, 0x59, 0xd9, 0xcc, 0x3a, 0xc5, 0x4e, 0x27, 0x62, 0x68, 0xca, 0x64, 0x73, 0xb3,
+	0xc6, 0x47, 0xd0, 0x3e, 0x54, 0x14, 0x24, 0xd4, 0xff, 0xe9, 0xd4, 0xd4, 0xc6, 0xe6, 0x33, 0x00,
+	0xbb, 0xe0, 0x18, 0x43, 0xc8, 0xc8, 0x5b, 0x35, 0x91, 0x0a, 0x9b, 0x3d, 0x85, 0x4e, 0xe9, 0x58,
+	0x5c, 0x07, 0xe7, 0x22, 0x0a, 0x62, 0x7d, 0x29, 0x13, 0x77, 0x25, 0xb5, 0x0e, 0xa4, 0xbc, 0x1a,
+	0x07, 0xea, 0xca, 0xb5, 0xd8, 0x9f, 0x0d, 0x68, 0x5d, 0x50, 0x34, 0xbc, 0x43, 0x3d, 0xf1, 0x63,
+	0xb0, 0x4f, 0x94, 0x1c, 0x9b, 0xc4, 0xeb, 0xcb, 0x65, 0xfc, 0xc8, 0xa0, 0xd1, 0x97, 0xe6, 0x2a,
+	0xf5, 0xac, 0x46, 0x5f, 0x2e, 0x4a, 0xc8, 0xae, 0x4a, 0x88, 0x41, 0x7b, 0x26, 0x8d, 0x55, 0x53,
+	0x5f, 0xdb, 0xef, 0x2b, 0xc1, 0x67, 0x30, 0x3e, 0x84, 0xb5, 0x23, 0x35, 0xe5, 0x93, 0xc8, 0x5b,
+	0x33, 0xda, 0xc9, 0x2c, 0xfc, 0x1a, 0x36, 0x38, 0xc5, 0xa1, 0x18, 0x98, 0x7a, 0x1c, 0xca, 0xe8,
+	0x67, 0x31, 0xf2, 0x5a, 0x59, 0x42, 0x15, 0x0f, 0xaf, 0x92, 0xd9, 0x0f, 0x35, 0x11, 0xf0, 0x2b,
+	0x80, 0xb4, 0xf9, 0x68, 0x60, 0xaa, 0x6e, 0x99, 0x78, 0x8f, 0xaa, 0xf1, 0x7a, 0x05, 0x87, 0x97,
+	0xf8, 0xec, 0x0f, 0x0b, 0xfe, 0x7f, 0x0b, 0x17, 0x9f, 0x41, 0xeb, 0x2c, 0x12, 0x89, 0x08, 0xc2,
+	0x4c, 0x4e, 0xdb, 0xe5, 0xd0, 0xa7, 0x93, 0x40, 0x05, 0x51, 0x42, 0xf4, 0x4a, 0x44, 0x43, 0x9e,
+	0x33, 0xf1, 0x39, 0x74, 0xce, 0xa2, 0x81, 0xa2, 0x31, 0x45, 0x49, 0x10, 0x9a, 0xbf, 0xe6, 0xd6,
+	0x8d, 0x65, 0x36, 0xfb, 0x0c, 0x9c, 0x9e, 0x92, 0x31, 0xa9, 0x64, 0x5a, 0xa8, 0xd2, 0x2a, 0xa9,
+	0x72, 0x0b, 0x56, 0x5f, 0x07, 0xe1, 0x24, 0x97, 0xea, 0x8d, 0xc1, 0x7e, 0xb5, 0x72, 0xc9, 0x68,
+	0xdc, 0x85, 0xfb, 0x3f, 0x6a, 0x1a, 0x2e, 0x4e, 0x03, 0x87, 0x2f, 0xc2, 0xc8, 0x60, 0xfd, 0xf8,
+	0x7d, 0x4c, 0x83, 0x84, 0x86, 0x17, 0xe2, 0x17, 0x32, 0xf2, 0x68, 0xf2, 0x39, 0x0c, 0x9f, 0x02,
+	0x64, 0xf9, 0x08, 0xd2, 0x9e, 0x6d, 0xba, 0xb2, 0xed, 0xe7, 0x29, 0xf2, 0x92, 0x93, 0xbd, 0x00,
+	0x37, 0xcd, 0xe1, 0x50, 0x8e, 0xe3, 0x90, 0x12, 0x32, 0xfa, 0xdd, 0x83, 0xce, 0xf7, 0x4a, 0x8c,
+	0x44, 0x14, 0x84, 0x9c, 0xae, 0x33, 0x99, 0x3a, 0x7e, 0x26, 0x6f, 0x5e, 0x76, 0x32, 0xac, 0xec,
+	0xd7, 0xec, 0x2f, 0x0b, 0x80, 0xd3, 0x80, 0xc4, 0x3b, 0xba, 0x4b, 0x3b, 0xdc, 0xc8, 0xbc, 0x71,
+	0xab, 0xcc, 0xf7, 0xc0, 0x3d, 0x0c, 0x29, 0x50, 0xe5, 0x02, 0xdd, 0x8c, 0xc2, 0x0a, 0x5e, 0x2f,
+	0x5a, 0xfb, 0xbf, 0x88, 0x76, 0xbd, 0x94, 0xbf, 0x66, 0x23, 0xd8, 0x3c, 0x22, 0x9d, 0x28, 0x39,
+	0xcd, 0xbb, 0xff, 0x2e, 0x53, 0x13, 0x3f, 0x85, 0x76, 0xc1, 0xf7, 0x1a, 0x4b, 0x27, 0xe3, 0x8c,
+	0xc4, 0xde, 0x00, 0x2e, 0x1c, 0x94, 0x0d, 0xd8, 0xdc, 0xcc, 0x5a, 0xa5, 0x76, 0xc0, 0xe6, 0x9c,
+	0x54, 0x6c, 0xc7, 0x4a, 0x49, 0x95, 0x8b, 0xcd, 0x18, 0xec, 0xa8, 0xee, 0x12, 0xe9, 0x9b, 0xd6,
+	0x4a, 0x4b, 0x17, 0x26, 0xf9, 0xf0, 0xde, 0xf4, 0xab, 0x29, 0xf0, 0x9c, 0xc3, 0xbe, 0x80, 0xad,
+	0x72, 0xb5, 0x26, 0x4a, 0x4b, 0x75, 0x97, 0x17, 0xa4, 0x5f, 0xbb, 0x4f, 0xe3, 0x56, 0x36, 0xae,
+	0xd3, 0x1d, 0xf6, 0xcb, 0x95, 0x62, 0x60, 0x3b, 0xe7, 0x32, 0xa1, 0xf7, 0x42, 0x27, 0x37, 0x5d,
+	0xf0, 0x72, 0x85, 0x17, 0xc8, 0x81, 0x03, 0x6b, 0x37, 0xe9, 0xb0, 0x27, 0xd0, 0xea, 0x89, 0x68,
+	0x94, 0x26, 0xe0, 0x41, 0xeb, 0x3b, 0xd2, 0x3a, 0x18, 0xe5, 0x8d, 0x97, 0x9b, 0xec, 0x83, 0x9c,
+	0xa4, 0xd3, 0xd6, 0x3c, 0x1e, 0x5c, 0xca, 0xbc, 0x35, 0xd3, 0xf5, 0xde, 0x2e, 0x34, 0xfb, 0x4a,
+	0xa4, 0xc3, 0xfc, 0x48, 0x46, 0xc9, 0x61, 0xa0, 0xc8, 0x5d, 0xc1, 0x36, 0xac, 0x9e, 0x04, 0xa1,
+	0x26, 0xd7, 0x42, 0x07, 0xec, 0xbe, 0x9a, 0x90, 0xdb, 0xd8, 0xfb, 0xcd, 0x02, 0x6f, 0xd9, 0x38,
+	0xc0, 0x2d, 0x70, 0x0b, 0xe0, 0x2c, 0x7a, 0x17, 0x84, 0x62, 0xe8, 0xae, 0xe0, 0x36, 0x3c, 0x28,
+	0x50, 0xa3, 0xd0, 0xe0, 0xad, 0x08, 0x45, 0x32, 0x75, 0x2d, 0x7c, 0x02, 0x1f, 0x96, 0x36, 0x14,
+	0xa3, 0xa4, 0x74, 0x80, 0xdb, 0x98, 0x8b, 0x7a, 0x2e, 0x93, 0x4b, 0x11, 0x8d, 0xdc, 0xe6, 0xfe,
+	0x3f, 0x8d, 0x74, 0xe6, 0x17, 0x3c, 0xec, 0x82, 0x9d, 0xde, 0x10, 0x1d, 0x3f, 0xab, 0x46, 0x37,
+	0x5f, 0x69, 0xfc, 0x12, 0xee, 0xcf, 0x3f, 0xdd, 0x1a, 0xd1, 0xaf, 0x7c, 0xef, 0x74, 0xab, 0x98,
+	0xc6, 0x1e, 0x3c, 0xac, 0x7f, 0xf5, 0xb1, 0xeb, 0x2f, 0xfd, 0x96, 0xe8, 0x2e, 0xf7, 0x69, 0x7c,
+	0x01, 0xee, 0xa2, 0x06, 0x71, 0xcb, 0xaf, 0xe9, 0xad, 0x6e, 0x1d, 0xaa, 0xf1, 0x9b, 0xf9, 0xc6,
+	0x36, 0x2a, 0xc2, 0x07, 0x7e, 0x9d, 0x22, 0xbb, 0xb5, 0xb0, 0xc6, 0xcf, 0xe1, 0xde, 0xdc, 0xb8,
+	0xc2, 0x0d, 0x7f, 0x71, 0xfc, 0x75, 0x2b, 0x90, 0x3e, 0x58, 0x7d, 0xd3, 0x8c, 0x87, 0x93, 0xb7,
+	0x6b, 0xe6, 0x93, 0xf1, 0xd9, 0xbf, 0x01, 0x00, 0x00, 0xff, 0xff, 0x7e, 0xa4, 0xfb, 0xca, 0x3f,
+	0x0a, 0x00, 0x00,
 }
@@ -61,6 +61,25 @@ message SendReq {
  Tri Encrypted = 5;

  bool DryRun = 6;
+
+  ReplicationConfig ReplicationConfig = 7;
+}
+
+message ReplicationConfig {
+  ReplicationConfigProtection protection = 1;
+}
+
+
+message ReplicationConfigProtection {
+  ReplicationGuaranteeKind Initial = 1;
+  ReplicationGuaranteeKind Incremental = 2;
+}
+
+enum ReplicationGuaranteeKind {
+  GuaranteeInvalid = 0;
+  GuaranteeResumability = 1;
+  GuaranteeIncrementalReplication = 2;
+  GuaranteeNothing = 3; 
 }

 message Property {
@@ -93,6 +112,8 @@ message ReceiveReq {
  // If true, the receiver should clear the resume token before performing the
  // zfs recv of the stream in the request
  bool ClearResumeToken = 3;
+
+  ReplicationConfig ReplicationConfig = 4;
 }

 message ReceiveRes {}
@@ -83,3 +83,10 @@ func (v *FilesystemVersion) ZFSFilesystemVersion() (*zfs.FilesystemVersion, erro
 		Creation:  ct,
 	}, nil
 }
+
+func ReplicationConfigProtectionWithKind(both ReplicationGuaranteeKind) *ReplicationConfigProtection {
+	return &ReplicationConfigProtection{
+		Initial:     both,
+		Incremental: both,
+	}
+}
@@ -9,6 +9,7 @@ import (
 	"time"

 	"github.com/prometheus/client_golang/prometheus"
+
 	"github.com/zrepl/zrepl/daemon/logging/trace"

 	"github.com/zrepl/zrepl/logger"
@@ -52,10 +53,6 @@ type Receiver interface {
 	Receive(ctx context.Context, req *pdu.ReceiveReq, receive io.ReadCloser) (*pdu.ReceiveRes, error)
 }

-type PlannerPolicy struct {
-	EncryptedSend tri // all sends must be encrypted (send -w, and encryption!=off)
-}
-
 type Planner struct {
 	sender   Sender
 	receiver Receiver
@@ -554,19 +551,25 @@ func (s *Step) updateSizeEstimate(ctx context.Context) error {
 		log.WithError(err).Error("dry run send request failed")
 		return err
 	}
-	s.expectedSize = sres.ExpectedSize
+	if sres == nil {
+		err := fmt.Errorf("dry run send request returned nil send result")
+		log.Error(err.Error())
+		return err
+	}
+	s.expectedSize = sres.GetExpectedSize()
 	return nil
 }

 func (s *Step) buildSendRequest(dryRun bool) (sr *pdu.SendReq) {
 	fs := s.parent.Path
 	sr = &pdu.SendReq{
-		Filesystem:  fs,
-		From:        s.from, // may be nil
-		To:          s.to,
-		Encrypted:   s.encrypt.ToPDU(),
-		ResumeToken: s.resumeToken,
-		DryRun:      dryRun,
+		Filesystem:        fs,
+		From:              s.from, // may be nil
+		To:                s.to,
+		Encrypted:         s.encrypt.ToPDU(),
+		ResumeToken:       s.resumeToken,
+		DryRun:            dryRun,
+		ReplicationConfig: &s.parent.policy.ReplicationConfig,
 	}
 	return sr
 }
@@ -584,6 +587,11 @@ func (s *Step) doReplication(ctx context.Context) error {
 		log.WithError(err).Error("send request failed")
 		return err
 	}
+	if sres == nil {
+		err := fmt.Errorf("send request returned nil send result")
+		log.Error(err.Error())
+		return err
+	}
 	if stream == nil {
 		err := errors.New("send request did not return a stream, broken endpoint implementation")
 		return err
@@ -603,9 +611,10 @@ func (s *Step) doReplication(ctx context.Context) error {
 	}()

 	rr := &pdu.ReceiveReq{
-		Filesystem:       fs,
-		To:               sr.GetTo(),
-		ClearResumeToken: !sres.UsedResumeToken,
+		Filesystem:        fs,
+		To:                sr.GetTo(),
+		ClearResumeToken:  !sres.UsedResumeToken,
+		ReplicationConfig: &s.parent.policy.ReplicationConfig,
 	}
 	log.Debug("initiate receive request")
 	_, err = s.receiver.Receive(ctx, rr, byteCountingStream)
@@ -0,0 +1,43 @@
+package logic
+
+import (
+	"github.com/pkg/errors"
+
+	"github.com/zrepl/zrepl/config"
+	"github.com/zrepl/zrepl/replication/logic/pdu"
+)
+
+type PlannerPolicy struct {
+	EncryptedSend     tri // all sends must be encrypted (send -w, and encryption!=off)
+	ReplicationConfig pdu.ReplicationConfig
+}
+
+func ReplicationConfigFromConfig(in *config.Replication) (*pdu.ReplicationConfig, error) {
+	initial, err := pduReplicationGuaranteeKindFromConfig(in.Protection.Initial)
+	if err != nil {
+		return nil, errors.Wrap(err, "field 'initial'")
+	}
+	incremental, err := pduReplicationGuaranteeKindFromConfig(in.Protection.Incremental)
+	if err != nil {
+		return nil, errors.Wrap(err, "field 'incremental'")
+	}
+	return &pdu.ReplicationConfig{
+		Protection: &pdu.ReplicationConfigProtection{
+			Initial:     initial,
+			Incremental: incremental,
+		},
+	}, nil
+}
+
+func pduReplicationGuaranteeKindFromConfig(in string) (k pdu.ReplicationGuaranteeKind, _ error) {
+	switch in {
+	case "guarantee_nothing":
+		return pdu.ReplicationGuaranteeKind_GuaranteeNothing, nil
+	case "guarantee_incremental":
+		return pdu.ReplicationGuaranteeKind_GuaranteeIncrementalReplication, nil
+	case "guarantee_resumability":
+		return pdu.ReplicationGuaranteeKind_GuaranteeResumability, nil
+	default:
+		return k, errors.Errorf("%q is not in guarantee_{nothing,incremental,resumability}", in)
+	}
+}
@@ -162,3 +162,30 @@ func (f *FilesystemReport) NextStep() *StepReport {
 func (f *StepReport) IsIncremental() bool {
 	return f.Info.From != ""
 }
+
+// Returns, for the latest replication attempt,
+// 0  if there have not been any replication attempts,
+// -1 if the replication failed while enumerating file systems
+// N  if N filesystems could not not be replicated successfully
+func (r *Report) GetFailedFilesystemsCountInLatestAttempt() int {
+
+	if len(r.Attempts) == 0 {
+		return 0
+	}
+
+	a := r.Attempts[len(r.Attempts)-1]
+	switch a.State {
+	case AttemptPlanningError:
+		return -1
+	case AttemptFanOutError:
+		var count int
+		for _, f := range a.Filesystems {
+			if f.Error() != nil {
+				count++
+			}
+		}
+		return count
+	default:
+		return 0
+	}
+}
@@ -66,7 +66,7 @@ type ProtocolError struct {
 }

 func (e *ProtocolError) Error() string {
-	return fmt.Sprintf("protocol error: %s", e)
+	return fmt.Sprintf("protocol error: %s", e.cause)
 }

 func (c *Client) recv(ctx context.Context, conn *stream.Conn, res proto.Message) error {
@@ -239,6 +239,10 @@ func (s *Server) serveConnRequest(ctx context.Context, endpoint string, c *strea

 	if sendStream != nil {
 		err := c.SendStream(ctx, sendStream, ZFSStream)
+		closeErr := sendStream.Close()
+		if closeErr != nil {
+			s.log.WithError(err).Error("cannot close send stream")
+		}
 		if err != nil {
 			s.log.WithError(err).Error("cannot write send stream")
 		}
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
Christian Schwarz	bd30afaaf9	experiment: zync: an rsync-like tool for ZFS datasets (re-uses zrepl abstractions) Local sync and sync over SSH seems to work. The sshdirect transport might be interesting / re-usable for zrepl as well, although we'd still need some kind of locking in that case.	2021-01-14 01:33:24 +01:00
Rafał Bugajewski	96d5288667	docs: fix typos	2020-12-17 12:00:29 +01:00
Christian Schwarz	b8cd3c59f1	docs: update supporters	2020-11-11 16:06:54 +01:00
Christian Schwarz	91632b52bb	build: circleci: periodic full build: only build each of the matching branches once	2020-11-04 17:53:54 +01:00
Christian Schwarz	d39c0e3745	docs + readme: actually fix Patreon badge see https://github.com/endel/shieldsio-patreon/issues/8#issuecomment-700144629	2020-11-01 16:10:48 +01:00
Christian Schwarz	180eaea195	docs: 0.3.1 changelog	2020-11-01 14:19:03 +01:00
Christian Schwarz	69ed2d7117	docs + readme: fix Patreon badge	2020-11-01 14:18:36 +01:00
Christian Schwarz	c420f3c909	[#381 ] zfs: ListFilesystemVersions: make list filesystems version invocation deterministic fixes #381 ref #379	2020-11-01 13:59:21 +01:00
Christian Schwarz	0a2dea05a9	[#385 ] status + replication: warning if replication succeeeded without any filesystem being replicated refs #385 refs #384	2020-11-01 13:51:28 +01:00
Christian Schwarz	293c89d392	[#385 ] replication: report AttemptDone if no filesystems are replicated fixes #385	2020-11-01 13:51:28 +01:00
Jeremy Bryan Smith	bb5ef0c8b2	docs: fix link to template.sh sample hook file	2020-11-01 10:45:17 +01:00
Christian Schwarz	8f44cdc284	build: rpm: don't strip binaries fixes build failure for aarch64 ``` make rpm-docker GOARCH=arm64 GOOS=linux + /usr/lib/rpm/brp-compress + /usr/lib/rpm/brp-strip /usr/bin/strip /usr/bin/strip: Unable to recognise the format of the input file `/build/src/artifacts/rpmbuild/BUILDROOT/zrepl-v0.3.0.38.g53028ed5-1.aarch64/usr/bin/zrepl' error: Bad exit status from /var/tmp/rpm-tmp.g8Fdbs (%install) Bad exit status from /var/tmp/rpm-tmp.g8Fdbs (%install) ```	2020-11-01 10:23:50 +01:00
Christian Schwarz	53028ed50a	docs: gen-sphinx-versioning-flags.py: implement the `stable` branch logic used for 0.3.0 release	2020-09-12 14:03:45 +02:00
Christian Schwarz	17e152c601	docs: build zrepl.github.io on circleci	2020-09-12 14:03:45 +02:00
Christian Schwarz	2ba6aabd4a	build: circleci: quickcheck-docs: upload artifacts	2020-09-06 17:28:16 +02:00
Christian Schwarz	98207c904d	docs + README: document new build & release process	2020-09-06 17:28:16 +02:00
Christian Schwarz	d2b825a8ae	build: daily release builds of branches stable, master, problame/circleci-build	2020-09-06 17:28:16 +02:00
Christian Schwarz	913b8b37fe	build: circleci: include pipeline number in minio url	2020-09-06 17:22:44 +02:00
Christian Schwarz	f494c5ba31	build: pin Go version in circleci	2020-09-06 17:15:32 +02:00
Christian Schwarz	b9503685f0	build: parametrized pipeline (no more approval for release build), embed exact go version info in artifacts refs #378 refs #377	2020-09-06 15:39:34 +02:00
Christian Schwarz	0f4143c0e0	docs: update supporters	2020-09-05 17:49:35 +02:00
Christian Schwarz	8839ed1f95	docs: update multi-job & multi-host setup section	2020-09-05 17:45:18 +02:00
Christian Schwarz	41b4038ad5	docs: add example setup 'local disk backup' to jobs overview table	2020-09-05 17:44:46 +02:00
Christian Schwarz	88d21eb23a	Makefile: docs: treat Sphinx warnings as errors	2020-09-05 16:41:22 +02:00
Christian Schwarz	e0be7e4d4f	docs: installation/rpm-repos: remove ineffective literalinclude directive	2020-09-05 16:40:33 +02:00
Christian Schwarz	0c189265e8	platformtests: fix skipping encryption-only tests on systems that don't support encryption (Or split the test into two tests of which one is skipped depending on encryption support)	2020-09-05 16:04:34 +02:00
Christian Schwarz	b1f8cdf385	[#373 ] pruning: add optional `regex` field to `last_n` rule fixes #373	2020-09-02 22:45:44 +02:00
Christian Schwarz	428a60870a	pruning: cleanup retention grid impl + tests + correct docs package is now at 95% code coverage and the additional tests codify all behavior specified in the docs There is a slight change in behavior: Intervals are now [duration) instead of (duration]. If the leftmost interval is not keep=all, the most recently created snapshot will be destroyed if there are other snapshots within that first interval. Since we recommend keep=all all over the docs, and zrepl 0.3 will put holds on that snapshot if it is being replicated, I feel like this is an acceptable change in behavior. refs #292 fixup of `0bbe2befce`	2020-09-02 22:45:44 +02:00
Christian Schwarz	af2d6579c5	[#347 ] zfscmd: fix dangling trace Task on .Start() failure fixes #347	2020-09-02 22:45:44 +02:00
Christian Schwarz	0f3da73ef1	[#347 ] zfscmd + zfs: define .Start() semantics, apply to call sites in pkg zfs fixes #347	2020-09-02 22:45:44 +02:00
Christian Schwarz	fecc9416ab	[#347 ] package trace: envconst-configurable debug mode	2020-09-02 22:45:44 +02:00
Christian Schwarz	a7915db4c3	[#347 ] package trace: printing debugString before instead of at panic (fixup `e500d9e`)	2020-09-02 22:45:44 +02:00
Christian Schwarz	3a4e841c73	[#292 ] pruning: grid: add all snapshots that do not match the regex to the rule's destroy list Before this patch, multiple grids with disjoint regexes would result in no snapshots being destroyed at all. fixes #292	2020-09-02 22:45:44 +02:00
Christian Schwarz	bcf6ff1c08	[#292 ] pruning: add func MustNewKeepGrid	2020-09-02 22:45:44 +02:00
Christian Schwarz	ad7a104ab4	[#292 ] pruning: simplify table test driver	2020-09-02 22:45:44 +02:00
Christian Schwarz	639359f393	[#292 ] pruning: last_n: use snap name as fallback when creation is equal	2020-09-02 22:45:44 +02:00
Christian Schwarz	91e310b7e3	build: rpm + deb targets, build-in-docker targets, CircleCI pipeline rewrite Co-authored-by: Armin Wehrfritz <dkxls23@gmail.com>	2020-09-02 21:34:52 +02:00
Christian Schwarz	5b30ad01ce	make formatcheck: exit with non-zero status code if unformatted files are found fixup of `7a5883d404`	2020-09-02 21:33:36 +02:00
chenrui	8f4f9338a9	build: remove travis config as I did not find it got referenced in the PR builds	2020-09-01 19:21:59 +02:00
Christian Schwarz	67bbce3c36	Merge pull request #375 from InsanePrawn/make_formatcheck Add `make formatcheck` and reformat all files	2020-09-01 00:12:10 +02:00
InsanePrawn	180c3d9ae1	Reformat all files with `make format`. Signed-off-by: InsanePrawn <insane.prawny@gmail.com>	2020-08-31 23:57:45 +02:00
InsanePrawn	7a5883d404	circleci: Add `make formatcheck` Signed-off-by: InsanePrawn <insane.prawny@gmail.com>	2020-08-31 23:57:45 +02:00
InsanePrawn	c90acefacb	Makefile: Add `formatcheck` target - `make formatcheck` outputs the list of files that need formatting - `make format` applies the formatting and now also outputs the diffs. Signed-off-by: InsanePrawn <insane.prawny@gmail.com>	2020-08-31 23:53:27 +02:00
InsanePrawn	d6799e08d8	lazy.sh: various fixes (#372 ) - Emit warning and return 1 when no cmd is specified - get docdep requirements from lazy.sh parent dir, not $GOPATH/src/zrepl/zrepl - Explicitly warn when gopath is not in PATH - Fix `release` cmd - replace `! -z` with `-n` Signed-off-by: InsanePrawn <insane.prawny@gmail.com>	2020-08-31 16:40:13 +02:00
Christian Schwarz	c44dccc34b	build: circleci: switch to 'machine' executor for test-build-in-docker job refs #356	2020-08-26 11:47:17 +02:00
Christian Schwarz	7f1695c457	docs: transport: fix easyrsa script (fixup of `6b4c6fc`)	2020-08-23 20:36:43 +02:00
Christian Schwarz	e500d9ee26	package trace: track activeChildTasks in a set if debugEnabled=true refs #358	2020-08-23 20:13:58 +02:00
Christian Schwarz	d5ce578929	[#358 ] platformtest: ReceiveForceIntoEncryptedErr: fix `trace` panic caused by dangling open zfs send process ``` panic: end task: 1 active child tasks: end task: task still has active child tasks goroutine 1 [running]: github.com/zrepl/zrepl/daemon/logging/trace.WithTask.func1.1(0xc000020680, 0xc000080f00) /home/cs/zrepl/zrepl/daemon/logging/trace/trace.go:221 +0x2f7 github.com/zrepl/zrepl/daemon/logging/trace.WithTask.func1() /home/cs/zrepl/zrepl/daemon/logging/trace/trace.go:237 +0x38 main.HarnessRun(0x7ffe0b49ad3a, 0x11, 0x7ffe0b49acf2, 0x1a, 0xc800000, 0x7ffe0b49ad19, 0x16, 0x1, 0x0, 0x0, ...) /home/cs/zrepl/zrepl/platformtest/harness/harness.go:168 +0xe54 main.main() /home/cs/zrepl/zrepl/platformtest/harness/harness.go:41 +0x2ef ``` fixes #358	2020-08-23 20:13:58 +02:00
Christian Schwarz	3f8fe3a368	[#346 ] do not rely on `creation` property for filesystem version diff fixes #346 fixes #358	2020-08-23 19:21:55 +02:00
Christian Schwarz	6b4c6fc062	[#357 ] docs: update quickstart + tls transport to produce keypairs with subject alternative names fixes #357	2020-08-22 03:05:30 +02:00
Christian Schwarz	e239d6f633	build: make platformtest-* usable * use plain `go build` instead of the `go test` + `func TestMain` hack * add target cover-platform and use that to generate coverage reports	2020-08-21 23:01:35 +02:00
InsanePrawn	0bbe2befce	docs: prune: add prune interval visualisation fixes #122 Co-Authored-By: Christian Schwarz <me@cschwarz.com> Signed-off-by: InsanePrawn <insane.prawny@gmail.com>	2020-08-21 22:05:05 +02:00
InsanePrawn	fa4e048169	readme: fix typo Signed-off-by: InsanePrawn <insane.prawny@gmail.com>	2020-08-21 22:05:05 +02:00
Christian Schwarz	4f9f21f7f2	logger: fix go-1.15-discovered conversion from int to string logger/datastructures.go:85:10: conversion from Level (int) to string yields a string of one rune	2020-08-12 21:44:02 +02:00
Christian Schwarz	480176ba2d	rpc/dataconn: fix go1.15-discovered recursive Error() method impl rpc/dataconn/dataconn_client.go:69:9: Sprintf format %s with arg e causes recursive Error method call	2020-08-12 21:41:31 +02:00
Christian Schwarz	1190c0f6d2	docs: supporters: update	2020-08-12 21:38:23 +02:00
Christian Schwarz	720a284db5	dist/grafana: fix endpoint abstractions cache metric panel fixup of `30cdc14`	2020-08-04 01:36:31 +02:00
Hans Schulz	83fdffbcef	replication: prometheus metric for number of failed replications in last attempt - package replication: metric - Grafana panel - wiring - changelog Signed-off-by: Christian Schwarz <me@cschwarz.com> closes #341	2020-08-04 01:19:44 +02:00
Christian Schwarz	0ee7a49d31	[#289 ] zfs: workaround for OpenZFS 0.7 dry send info with zero estimated size fixes #289	2020-07-26 20:32:35 +02:00
Christian Schwarz	02db5994fe	[#345 ] fix broken identification of parent-fs for initial replication ordering fixup of `02807279` fixes #345	2020-07-26 20:32:35 +02:00
Christian Schwarz	1d7a84e8ae	build: extract debian binary packaging workflow trigger into a reusable script	2020-07-26 20:32:35 +02:00
Christian Schwarz	a0e3dc7040	[#348 ] replication: add platformtest to check behavior on recv fail while still sending Regression test for #348	2020-07-26 20:32:35 +02:00
Christian Schwarz	43495d70c7	[#348 ] replication: return errors if .Send rpc returns a nil SendRes discovered while developing the next commit in this series	2020-07-26 20:32:35 +02:00
Christian Schwarz	fa586b493c	[#348 ] fix crash on early-recv error * The SendStream.Close() was not called by dataconn.Server, which left the zfs send process dangling. * When the source job's ctx interceptor closed the task, the dangling zfs send was detect by the trace package and panicked. 020-07-25T19:54:41-04:00 [ERRO][latitude][rpc.data][cyZj$J3Ca$J3Ca.CJwB]: cannot write send stream err="frameconn: shutting down" panic: end task: 1 active child tasks: end task: task still has active child tasks goroutine 196966 [running]: github.com/zrepl/zrepl/daemon/logging/trace.WithTask.func1.1(0xc000320680, 0xcde000) /home/jeremy/go/src/github.com/zrepl/zrepl/daemon/logging/trace/trace.go:221 +0x2f7 github.com/zrepl/zrepl/daemon/logging/trace.WithTask.func1() /home/jeremy/go/src/github.com/zrepl/zrepl/daemon/logging/trace/trace.go:237 +0x38 github.com/zrepl/zrepl/daemon/logging/trace.WithTaskAndSpan.func1() /home/jeremy/go/src/github.com/zrepl/zrepl/daemon/logging/trace/trace_convenience.go:41 +0x37 github.com/zrepl/zrepl/daemon/job.(PassiveSide).Run.func1(0xdcf780, 0xc0000a3560, 0xdc65a0, 0xc00035e620, 0xc0000a34d0) /home/jeremy/go/src/github.com/zrepl/zrepl/daemon/job/passive.go:194 +0x2e7 github.com/zrepl/zrepl/rpc.NewServer.func3(0xdcf780, 0xc0001ce4b0, 0xdc65e0, 0xc00035e600, 0xc0000a34d0) /home/jeremy/go/src/github.com/zrepl/zrepl/rpc/rpc_server.go:82 +0xd5 github.com/zrepl/zrepl/rpc/dataconn.(Server).serveConn(0xc0000a2ba0, 0xc00018eca0) /home/jeremy/go/src/github.com/zrepl/zrepl/rpc/dataconn/dataconn_server.go:149 +0x3be github.com/zrepl/zrepl/rpc/dataconn.(Server).Serve.func3(0xc0000b8180, 0xc0000a2ba0, 0xc00018eca0) /home/jeremy/go/src/github.com/zrepl/zrepl/rpc/dataconn/dataconn_server.go:108 +0x5d created by github.com/zrepl/zrepl/rpc/dataconn.(Server).Serve /home/jeremy/go/src/github.com/zrepl/zrepl/rpc/dataconn/dataconn_server.go:106 +0x24a 2020-07-25T19:58:55-04:00 [ERRO][latitude][rpc.data][Pt4F$gCWT$gCWT.fzhc]: cannot write send stream err="frameconn: shutting down" panic: end task: 1 active child tasks: end task: task still has active child tasks fixes #348	2020-07-26 20:32:35 +02:00
Christian Schwarz	30cdc1430e	replication + endpoint: replication guarantees: guarantee_{resumability,incremental,nothing} This commit - adds a configuration in which no step holds, replication cursors, etc. are created - removes the send.step_holds.disable_incremental setting - creates a new config option `replication` for active-side jobs - adds the replication.protection.{initial,incremental} settings, each of which can have values - `guarantee_resumability` - `guarantee_incremental` - `guarantee_nothing` (refer to docs/configuration/replication.rst for semantics) The `replication` config from an active side is sent to both endpoint.Sender and endpoint.Receiver for each replication step. Sender and Receiver then act accordingly. For `guarantee_incremental`, we add the new `tentative-replication-cursor` abstraction. The necessity for that abstraction is outlined in https://github.com/zrepl/zrepl/issues/340. fixes https://github.com/zrepl/zrepl/issues/340	2020-07-26 20:32:35 +02:00
Christian Schwarz	27673a23e9	config: add test for fromdefaults behavior	2020-07-26 20:32:35 +02:00
Christian Schwarz	95fc299733	daemon/job: test that sample configs are buildable	2020-07-26 20:32:35 +02:00
Christian Schwarz	4e702eedc9	cmd: zfs-abstraction list --json: fix panic (was panicking because `abstractions` is in fact a channel	2020-07-26 20:32:35 +02:00
Christian Schwarz	8ff83f2f1a	[#342 ] endpoint: always create unencrypted placeholder filesystems This "breaks" the use case of receiving an unencrypted send into an encrypted receiver by setting the receiver's `root_fs`'s `encryption=on`. "breaks" in air-quotes because we have not yet released a version of zrepl with encrypted send support. We will bring back the featured outlined above in a future release. See https://github.com/zrepl/zrepl/issues/342#issuecomment-657231818 and following.	2020-07-26 20:32:35 +02:00
Christian Schwarz	4b8f0ad112	docs: supporters: update	2020-06-22 13:36:00 +02:00
Brian Candler	dbc8bbeb6a	docs: config: prune: example: keep manual snapshots on receiver Fixes #335 closes #336 Signed-off-by: Christian Schwarz <me@cschwarz.com>	2020-06-22 12:32:03 +02:00
Christian Schwarz	b3e856f40d	docs: changelog: 0.3: fix broken issue link	2020-06-22 12:30:42 +02:00