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

Fix the bug that consul is always in the leaving state #663

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Fix the bug that consul is always in the leaving state #663

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
25 changes: 22 additions & 3 deletions serf/serf.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -108,6 +108,9 @@ type Serf struct {

// metricLabels is the slice of labels to put on all emitted metrics
metricLabels []metrics.Label

// lamport clock time of last received join intent message
memberJoinIntentstatusLTime map[string]LamportTime
}

// SerfState is the state of the Serf instance.
Expand Down Expand Up @@ -376,6 +379,9 @@ func Create(conf *Config) (*Serf, error) {
// Create the buffer for recent intents
serf.recentIntents = make(map[string]nodeIntent)

//Create the buffer for recent JoinIntent statusLTime
serf.memberJoinIntentstatusLTime = make(map[string]LamportTime)

// Create a buffer for events and queries
serf.eventBuffer = make([]*userEvents, conf.EventBuffer)
serf.queryBuffer = make([]*queries, conf.QueryBuffer)
Expand Down Expand Up @@ -965,6 +971,9 @@ func (s *Serf) handleNodeJoin(n *memberlist.Node) {
metrics.IncrCounterWithLabels([]string{"serf", "member", "flap"}, 1, s.metricLabels)
}

if join, ok := s.memberJoinIntentstatusLTime[n.Name]; ok {
member.statusLTime = join
}
member.Status = StatusAlive
member.leaveTime = time.Time{}
member.Addr = n.Addr
Expand Down Expand Up @@ -1116,6 +1125,10 @@ func (s *Serf) handleNodeLeaveIntent(leaveMsg *messageLeave) bool {
return false
}

if join, ok := s.memberJoinIntentstatusLTime[leaveMsg.Node]; ok && leaveMsg.LTime <= join {
return false
}

// Refute us leaving if we are in the alive state
// Must be done in another goroutine since we have the memberLock
if leaveMsg.Node == s.config.NodeName && state == SerfAlive {
Expand Down Expand Up @@ -1226,12 +1239,18 @@ func (s *Serf) handleNodeJoinIntent(joinMsg *messageJoin) bool {
}

// Update the LTime
member.statusLTime = joinMsg.LTime
if join, ok := s.memberJoinIntentstatusLTime[joinMsg.Node]; ok && joinMsg.LTime <= join {
return false
}
s.memberJoinIntentstatusLTime[joinMsg.Node] = joinMsg.LTime

// If we are in the leaving state, we should go back to alive,
// since the leaving message must have been for an older time
if member.Status == StatusLeaving {
// Since default member.Status is StatusAlive and member.statusLTime is zero,
// we should update member.statusLTime here
if member.Status == StatusLeaving || member.Status == StatusAlive {
member.Status = StatusAlive
member.statusLTime = joinMsg.LTime
}
return true
}
Expand Down Expand Up @@ -1528,7 +1547,7 @@ func (s *Serf) resolveNodeConflict() {
}
}

//eraseNode takes a node completely out of the member list
// eraseNode takes a node completely out of the member list
func (s *Serf) eraseNode(m *memberState) {
// Delete from members
delete(s.members, m.Name)
Expand Down