When promoting a learner member we should not count already a voting
member, but take only into account the number of existing voting
members and their current status (started, unstarted) when taking the
decision whether a learner member can be promoted.
Before this change, it was impossible to grow from a quorum N to a N+1
through promoting a learning member.
Fixes: #11633
In case of URLs that are synonyms, the current lexicographic sorting
and compare of the URLs fails with frustrating errors. Make sure to do
a full comparison between every set of PeerURLs before failing.
Fixes#11013
Useful for deciding when to terminate the unhealthy follower.
If the follower is receiving a leader snapshot, operator may wait.
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
This change introduces joint quorums by changing the Node and RawNode
API to accept pb.ConfChangeV2 (on top of pb.ConfChange).
pb.ConfChange continues to work as today: it allows carrying out a
single configuration change. A pb.ConfChange proposal gets added to
the Raft log as such and is thus also observed by the app during Ready
handling, and fed back to ApplyConfChange.
ConfChangeV2 allows joint configuration changes but will continue to
carry out configuration changes in "one phase" (i.e. without ever
entering a joint config) when this is possible.
This is the first (maybe not last) step in cleaning up the bootstrap
code around StartNode.
Initializing a Raft group for the first time is awkward, since a
configuration has to be pulled from thin air. The way this is solved
today is unclean: The app is supposed to pass peers to StartNode(),
we add configuration changes for them to the log, immediately pretend
that they are applied, but actually leave them unapplied (to give the
app a chance to observe them, though if the app did decide to not apply
them things would really go off the rails), and then return control to
the app. The app will then process the initial Readys and as a result
the configuration will be persisted to disk; restarts of the node then
use RestartNode which doesn't take any peers.
The code that did this lived awkwardly in two places fairly deep down
the callstack, though it was really only necessary in StartNode(). This
commit refactors things to make this more obvious: only StartNode does
this dance now. In particular, RawNode does not support this at all any
more; it expects the app to set up its Storage correctly.
Future work may provide helpers to make this "preseeding" of the Storage
more user-friendly. It isn't entirely straightforward to do so since
the Storage interface doesn't provide the right accessors for this
purpose. Briefly speaking, we want to make sure that a non-bootstrapped
node can never catch up via the log so that we can implicitly use one
of the "skipped" log entries to represent the configuration change into
the bootstrap configuration. This is an invasive change that affects
all consumers of raft, and it is of lower urgency since the code (post
this commit) already encapsulates the complexity sufficiently.
Mechanically extract `progressTracker`, `Progress`, and `inflights`
to their own package named `tracker`. Add lots of comments in the
progress, and take the opportunity to rename and clarify various
fields.
Update to Go 1.12.5 testing. Remove deprecated unused and gosimple
pacakges, and mask staticcheck 1006. Also, fix unconvert errors related
to unnecessary type conversions and following staticcheck errors:
- remove redundant return statements
- use for range instead of for select
- use time.Since instead of time.Now().Sub
- omit comparison to bool constant
- replace T.Fatal and T.Fatalf in tests with T.Error and T.Fatalf respectively because the goroutine calls T.Fatal must be called in the same goroutine as the test
- fix error strings that should not be capitalized
- use sort.Strings(...) instead of sort.Sort(sort.StringSlice(...))
- use he status code of Canceled instead of grpc.ErrClientConnClosing which is deprecated
- use use status.Errorf instead of grpc.Errorf which is deprecated
Related #10528#10438
