dmvk commented on issue #9759: [FLINK-14169][history-server] Cleanup expired jobs from history server URL: https://github.com/apache/flink/pull/9759#issuecomment-542579134 @aljoscha Thanks for the review! The original motivation was [FLINK-13958](https://issues.apache.org/jira/browse/FLINK-13958). As we are forced to run jobs in detached mode, we need to somehow determine that the batch job has completed successfully. We do this by polling the history server. History server became unresponsive with large number of jobs (1000+) and in case of restart it took a really long time to recover as it needed to replay the whole history from hdfs. We only need to keep recent jobs in history server and the expected behavior would be, that the history server "mirrors" the actual state of watched directories. WDYT? Thanks, D.
---------------------------------------------------------------- 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] With regards, Apache Git Services
