AFAIK, the HA data, including Zookeeper meta data and real data on DFS, will only be cleaned up when the Flink cluster reached terminated state.
So if you are using a session cluster, the root cluster node on Zk will be cleaned up after you manually stop the session cluster. The job sub directory will be cleaned up when the job finished/canceled/failed. If you are using a job/application cluster, once the only running job finished/failed, all the HA data will be cleaned up. I think you need to check the job restart strategy you have set. For example, the following configuration will make the Flink cluster terminated after 10 attempts. restart-strategy: fixed-delay restart-strategy.fixed-delay.attempts: 10 Best, Yang Cristian <k...@fastmail.fm> 于2020年9月9日周三 上午12:28写道: > I'm using the standalone script to start the cluster. > > As far as I can tell, it's not easy to reproduce. We found that zookeeper > lost a node around the time this happened, but all of our other 75 Flink > jobs which use the same setup, version and zookeeper, didn't have any > issues. They didn't even restart. > > So unfortunately I don't know how to reproduce this. All I know is I can't > sleep. I have nightmares were my precious state is deleted. I wake up > crying and quickly start manually savepointing all jobs just in case, > because I feel the day of reckon is near. Flinkpocalypse! > > On Tue, Sep 8, 2020, at 5:54 AM, Robert Metzger wrote: > > Thanks a lot for reporting this problem here Cristian! > > I am not super familiar with the involved components, but the behavior you > are describing doesn't sound right to me. > Which entrypoint are you using? This is logged at the beginning, like > this: "2020-09-08 14:45:32,807 INFO > org.apache.flink.runtime.entrypoint.ClusterEntrypoint [] - > Starting StandaloneSessionClusterEntrypoint (Version: 1.11.1, Scala: 2.12, > Rev:7eb514a, Date:2020-07-15T07:02:09+02:00)" > > Do you know by chance if this problem is reproducible? With > the StandaloneSessionClusterEntrypoint I was not able to reproduce the > problem. > > > > > On Tue, Sep 8, 2020 at 4:00 AM Husky Zeng <568793...@qq.com> wrote: > > Hi Cristian, > > > I don't know if it was designed to be like this deliberately. > > So I have already submitted an issue ,and wait for somebody to response. > > https://issues.apache.org/jira/browse/FLINK-19154 > > > > -- > Sent from: > http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/ > >