Restarting replication job and finding job ID

Learn how to restart the Kudu replication job from a saved state and how to identify the job ID on YARN or a session cluster.

Restarting from a savepoint

To restart the replication job from a savepoint, you must pass the savepoint path by using the -s flag during job submission.

Run the following command to restart the job:

flink run-application -t yarn-application \
  -s hdfs:///kudu-replication/savepoints/my_table/savepoint-<id> \
  -Dclassloader.parent-first-patterns.additional=org.apache.kudu \
  -c org.apache.kudu.replication.ReplicationJob \
  kudu-replication-<version>.jar \
  --job.sourceMasterAddresses source-master1:7051,source-master2:7051 \
  --job.sinkMasterAddresses   sink-master1:7051,sink-master2:7051 \
  --job.tableName              my_table \
  --job.checkpointsDirectory  hdfs:///kudu-replication/checkpoints/my_table

Finding the job ID

You must identify the job ID to perform management tasks such as stopping the job or creating a savepoint.

On YARN: To find the job ID of a running replication job on YARN, run the following command:
```
flink list -t yarn-application -Dyarn.application.id=<yarn-application-id>
```
On a session cluster: To list all running jobs across a session cluster, run the following command: