Observe resumed heartbeat after failures (hashicorp#467)

benbuzbee · Jun 22, 2021 · 7bba491 · 7bba491
1 parent f7242e7
commit 7bba491
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 0 deletions.
diff --git a/observer.go b/observer.go
@@ -34,6 +34,11 @@ type FailedHeartbeatObservation struct {
 	LastContact time.Time
 }
 
+// ResumedHeartbeatObservation is sent when a node resumes to heartbeat with the leader following failures
+type ResumedHeartbeatObservation struct {
+	PeerID ServerID
+}
+
 // nextObserverId is used to provide a unique ID for each observer to aid in
 // deregistration.
 var nextObserverID uint64

diff --git a/replication.go b/replication.go
@@ -388,6 +388,9 @@ func (r *Raft) heartbeat(s *followerReplication, stopCh chan struct{}) {
 			case <-stopCh:
 			}
 		} else {
+			if failures > 0 {
+				r.observe(ResumedHeartbeatObservation{PeerID: s.peer.ID})
+			}
 			s.setLastContact()
 			failures = 0
 			labels := []metrics.Label{{Name: "peer_id", Value: string(s.peer.ID)}}