Hi Sachin, Yes, that's correct. To resume from a savepoint, use the command bin/flink run -s <savepointPath> <runArgs>. You can find more details in the Flink documentation on [1].
Additionally, information on how to trigger a savepoint can be found in the section for triggering savepoints [2]. [1] https://nightlies.apache.org/flink/flink-docs-master/docs/ops/state/savepoints/#resuming-from-savepoints [2] https://nightlies.apache.org/flink/flink-docs-master/docs/ops/state/savepoints/#triggering-savepoints Best, Junrui Sachin Mittal <sjmit...@gmail.com> 于2024年5月25日周六 20:35写道： > Hi, > I have a long running yarn cluster and I submit my streaming job using the > following command: > > flink run -m yarn-cluster -yid application_1473169569237_0001 > /usr/lib/flink/examples/streaming/WordCount.jar --input file:///input.txt > --output file:///output/ > > Let's say I want to stop this job, make updates to the jar and some new > input arguments and restart the job from the savepoint. How would I do the > same ? > > Would this be the right command ? > > flink run -s /tmp/flink-savepoints/savepoint-cca7bc-bb1e257f0dab -m > yarn-cluster -yid application_1473169569237_0001 > /usr/lib/flink/examples/streaming/WordCount-Updated.jar --input > file:///input1.txt --output file:///output1/ --newarg value123 > > Thanks > Sachin > >
