Avoid returning early on agent join failures

When a gossip join failure happens do not return early in the call chain because a join failure is most likely transient and the retry logic built in the networkdb is going to retry and succeed. Returning early makes the initialization of ingress network/sandbox to not happen which causes a problem even after the gossip join on retry is successful. Signed-off-by: Jana Radhakrishnan <[email protected]>
moby · Sep 27, 2016 · 23a782b · 23a782b
1 parent bf3d9cc
commit 23a782b
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 2 deletions.
diff --git a/agent.go b/agent.go
@@ -191,8 +191,7 @@ func (c *controller) agentSetup() error {
 
 	if remoteAddr != "" {
 		if err := c.agentJoin(remoteAddr); err != nil {
-			logrus.Errorf("Error in agentJoin : %v", err)
-			return nil
+			logrus.Errorf("Error in joining gossip cluster : %v(join will be retried in background)", err)
 		}
 	}
 

diff --git a/networkdb/cluster.go b/networkdb/cluster.go
@@ -161,6 +161,10 @@ func (nDB *NetworkDB) retryJoin(members []string, stop <-chan struct{}) {
 				logrus.Errorf("Failed to join memberlist %s on retry: %v", members, err)
 				continue
 			}
+			if err := nDB.sendNodeEvent(NodeEventTypeJoin); err != nil {
+				logrus.Errorf("failed to send node join on retry: %v", err)
+				continue
+			}
 			return
 		case <-stop:
 			return