HuangZhenQiu commented on a change in pull request #7356:
[FLINK-10868][flink-yarn] Enforce maximum failed TMs in YarnResourceManager
URL: https://github.com/apache/flink/pull/7356#discussion_r245161384
##########
File path:
flink-runtime/src/main/java/org/apache/flink/runtime/resourcemanager/ResourceManager.java
##########
@@ -145,6 +146,9 @@
/** All registered listeners for status updates of the ResourceManager.
*/
private ConcurrentMap<String, InfoMessageListenerRpcGateway>
infoMessageListeners;
+ /** The number of failed containers since the master became active. */
+ protected AtomicInteger failedContainerSoFar = new AtomicInteger(0);
Review comment:
There are three code paths that will create failed container. One is the
container can't be started from YARN, which is notified by NMClient callback,
the other two are failed to registerTastManager and disconnectTaskManager (When
tasks fail). Thus, the concurrent is between akka thread and main thread?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
