[#316] endpoint / replication protocol: more robust step-holds and replication cursor management

- drop HintMostRecentCommonAncestor rpc call
    - it is wrong to put faith into the active side of the replication to always make that call
      (we might not trust it, ref pull setup)
- clean up step holds + step bookmarks + replication cursor bookmarks on
  send RPC instead
    - this makes it symmetric with Receive RPC
- use a cache (endpoint.sendAbstractionsCache) to avoid the cost of
  listing the on-disk endpoint abstractions state on every step

The "create" methods for endpoint abstractions (CreateReplicationCursor, HoldStep) are now fully
idempotent and return an Abstraction.

Notes about endpoint.sendAbstractionsCache:
- fills lazily from disk state on first `Get` operation
- fill from disk is generally only attempted once
    - unless the `ListAbstractions` fails, in which case the fill from
      disk is retried on next `Get` (the current `Get` will observe a
      subset of the actual on-disk abstractions)
    - the `Invalidate` method is called
- it is a global (zrepl process-wide) cache

fixes #316

daemon/daemon.go

@@ -13,6 +13,7 @@ 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/config"
 	"github.com/zrepl/zrepl/daemon/job"
@@ -95,6 +96,7 @@ func Run(ctx context.Context, conf *config.Config) error {
 	// register global (=non job-local) metrics
 	zfscmd.RegisterMetrics(prometheus.DefaultRegisterer)
 	trace.RegisterMetrics(prometheus.DefaultRegisterer)
+	endpoint.RegisterMetrics(prometheus.DefaultRegisterer)
 
 	log.Info("starting daemon")