Merge pull request #46615 from gmarek/log-monitor

Automatic merge from submit-queue (batch tested with PRs 45488, 45335, 45909, 46074, 46615) Make log-monitor give up on trying to ssh to a dead node after some time Fix #38263

Merge pull request #46615 from gmarek/log-monitor
f1cf1f9c · Kubernetes Submit Queue · GitHub · 4979ae4b · 0cc1999e · f1cf1f9c
Commit f1cf1f9c authored May 30, 2017 by Kubernetes Submit Queue Committed by GitHub May 30, 2017
Hide whitespace changes
Inline Side-by-side

Showing with 6 additions and 2 deletions

log_size_monitoring.go test/e2e/framework/log_size_monitoring.go +6 -2

No files found.
--- a/test/e2e/framework/log_size_monitoring.go
+++ b/test/e2e/framework/log_size_monitoring.go
@@ -249,9 +249,13 @@ func (g *LogSizeGatherer) Work() bool {
 	)
 	if err != nil {
 		Logf("Error while trying to SSH to %v, skipping probe. Error: %v", workItem.ip, err)
-		if workItem.backoffMultiplier < 128 {
-			workItem.backoffMultiplier *= 2
+		// In case of repeated error give up.
+		if workItem.backoffMultiplier >= 128 {
+			Logf("Failed to ssh to a node %v multiple times in a row. Giving up.", workItem.ip)
+			g.wg.Done()
+			return false
 		}
+		workItem.backoffMultiplier *= 2
 		go g.pushWorkItem(workItem)
 		return true
 	}