Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Skip leadership check if the etcd instance is active processing heartbeats #18428

Merged
merged 1 commit into from
Aug 14, 2024
Merged

Skip leadership check if the etcd instance is active processing heartbeats #18428

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 12 additions & 2 deletions server/etcdserver/raft.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -80,7 +80,9 @@ type toApply struct {
type raftNode struct {
lg *zap.Logger

tickMu *sync.Mutex
tickMu *sync.RWMutex
// timestamp of the latest tick
latestTickTs time.Time
raftNodeConfig

// a chan to send/receive snapshot
Expand Down Expand Up @@ -132,8 +134,9 @@ func newRaftNode(cfg raftNodeConfig) *raftNode {
raft.SetLogger(lg)
r := &raftNode{
lg: cfg.lg,
tickMu: new(sync.Mutex),
tickMu: new(sync.RWMutex),
raftNodeConfig: cfg,
latestTickTs: time.Now(),
// set up contention detectors for raft heartbeat message.
// expect to send a heartbeat within 2 heartbeat intervals.
td: contention.NewTimeoutDetector(2 * cfg.heartbeat),
Expand All @@ -155,9 +158,16 @@ func newRaftNode(cfg raftNodeConfig) *raftNode {
func (r *raftNode) tick() {
r.tickMu.Lock()
r.Tick()
r.latestTickTs = time.Now()
r.tickMu.Unlock()
}

func (r *raftNode) getLatestTickTs() time.Time {
r.tickMu.RLock()
defer r.tickMu.RUnlock()
return r.latestTickTs
}

// start prepares and starts raftNode in a new goroutine. It is no longer safe
// to modify the fields after it has been started.
func (r *raftNode) start(rh *raftReadyHandler) {
Expand Down
16 changes: 16 additions & 0 deletions server/etcdserver/server.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -904,10 +904,26 @@ func (s *EtcdServer) revokeExpiredLeases(leases []*lease.Lease) {
})
}

// isActive checks if the etcd instance is still actively processing the
// heartbeat message (ticks). It returns false if no heartbeat has been
// received within 3 * tickMs.
ahrtr marked this conversation as resolved.
Show resolved Hide resolved
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why 3 election ticks? I think the correct value is somewhere between TickMs (default 100ms) and ElectionMs (default 1s), meaning the 3 makes sense for default config, but I would argue it can be problematic for cases where ElectionMS < 3* TicksMs

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why 3 election ticks?

Read #18428 (comment).

I would argue it can be problematic for cases where ElectionMS < 3* TicksMs

I am not worry about this.

  • checking activity should be only related to tickMs.
  • also if ElectionMS < 3* TicksMs, it means it's a very inappropriate election value.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

also if ElectionMS < 3* TicksMs, it means it's a very inappropriate election value.

Then let's codify that is is inappropriate. Either working or failing validation.

func (s *EtcdServer) isActive() bool {
latestTickTs := s.r.getLatestTickTs()
threshold := 3 * time.Duration(s.Cfg.TickMs) * time.Millisecond
return latestTickTs.Add(threshold).After(time.Now())
}

// ensureLeadership checks whether current member is still the leader.
func (s *EtcdServer) ensureLeadership() bool {
lg := s.Logger()

if s.isActive() {
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is incorrect assumption for 5 node clusters, getting a tick from 1 member is not effective way to confirm no other leader was elected. Imagine 3/2 network split with leader on side of 2. With your change the leader can continue think it's active by receiving ticks from 1 member, where on the other side of network 3 members have already elected new leader.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It has nothing to do with how many nodes the cluster has. The intention is to guard the case the node being stuck by itself, e.g stall write.

Imagine 3/2 network split with leader on side of 2. With your change the leader can continue think it's active by receiving ticks from 1 member

It's active, but it won't be a leader any more. It will step down automatically in such case due to losing quorum.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Note that the raft protocol handle the network partition case perfectly.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's active, but it won't be a leader any more.

We are talking about 1 second before network partition, until heath probes fail for ElectionMs, the old leader will still think it's a leader before they resign.

lg.Debug("The member is active, skip checking leadership",
zap.Time("latestTickTs", s.r.getLatestTickTs()),
zap.Time("now", time.Now()))
return true
}

ctx, cancel := context.WithTimeout(s.ctx, s.Cfg.ReqTimeout())
defer cancel()
if err := s.linearizableReadNotify(ctx); err != nil {
Expand Down
43 changes: 43 additions & 0 deletions server/etcdserver/server_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,6 +29,7 @@ import (

"github.com/coreos/go-semver/semver"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
"go.uber.org/zap"
"go.uber.org/zap/zaptest"

Expand Down Expand Up @@ -1538,3 +1539,45 @@ func TestWaitAppliedIndex(t *testing.T) {
})
}
}

func TestIsActive(t *testing.T) {
cases := []struct {
name string
tickMs uint
durationSinceLastTick time.Duration
expectActive bool
}{
{
name: "1.5*tickMs,active",
tickMs: 100,
durationSinceLastTick: 150 * time.Millisecond,
expectActive: true,
},
{
name: "2*tickMs,active",
tickMs: 200,
durationSinceLastTick: 400 * time.Millisecond,
expectActive: true,
},
{
name: "4*tickMs,not active",
tickMs: 150,
durationSinceLastTick: 600 * time.Millisecond,
expectActive: false,
},
}

for _, tc := range cases {
s := EtcdServer{
Cfg: config.ServerConfig{
TickMs: tc.tickMs,
},
r: raftNode{
tickMu: new(sync.RWMutex),
latestTickTs: time.Now().Add(-tc.durationSinceLastTick),
},
}

require.Equal(t, tc.expectActive, s.isActive())
}
}
Loading