xintongsong commented on a change in pull request #8952:
URL: https://github.com/apache/flink/pull/8952#discussion_r547825381
##########
File path:
flink-runtime/src/main/java/org/apache/flink/runtime/resourcemanager/active/ActiveResourceManager.java
##########
@@ -246,6 +269,10 @@ private void requestNewWorker(WorkerResourceSpec
workerResourceSpec) {
workerResourceSpec,
count,
exception);
+ if (recordWorkerFailure()) {
+ // if exceed failure
rate try to slow down
+
tryResetWorkerCreationCoolDown();
+ }
Review comment:
What is the concern on having different behaviors in
`onWorkerTerminated`?
IIUC, this feature is a protection agains container start-up failures. The
current implementation counts all the terminated containers as failed,
including those that are terminated normally as expected.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
