[GitHub] [spark] holdenk commented on a diff in pull request #38518: [SPARK-33349][K8S] Reset the executor pods watcher when we receive a version changed from k8s

[GitBox <gi...@apache.org>]

holdenk commented on code in PR #38518:
URL: https://github.com/apache/spark/pull/38518#discussion_r1036510431


##########
resource-managers/kubernetes/core/src/main/scala/org/apache/spark/scheduler/cluster/k8s/ExecutorPodsWatchSnapshotSource.scala:
##########
@@ -86,8 +97,14 @@ class ExecutorPodsWatchSnapshotSource(
     }
 
     override def onClose(e: WatcherException): Unit = {
-      logWarning("Kubernetes client has been closed (this is expected if the application is" +
-        " shutting down.)", e)
+      if (e != null && e.isHttpGone) {

Review Comment:
   I'm _slightly_ worried about an infinite loop here, I don't see any clear doc on when isHttpGone is set true in fabric8, can we have a max resets in K minutes counter or something similar?
   
   Or if there is good docs on the behaviour of isHttpGone that would be good too.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org