Apache FlinkPDF version

Migrating Flink jobs without state

When you run Flink application without state, you only need to stop your currently running Flink jobs, and restart them on your upgraded Data Hub clusters.

  1. Connect to your host where you run your Flink jobs using ssh. 
    ssh <workload_username>@<master_host>
Password: <your_workload_password>
  2. Determine which jobs you want to stop by listing the Flink job IDs. 
    flink list
  3. Stop your Flink jobs using command line interface (CLI). 
    flink cancel <flink_job_ID>
  4. Resubmit the stopped Flink jobs on your new cluster.
    1. Connect to your host where you intend to run your Flink jobs using ssh. 
      ssh <workload_username>@<master_host>
Password: <your_workload_password>
    2. Resubmit your Flink application. 
      flink run <run_arguments> \
  <jar_file> <app_arguments>

We want your opinion

How can we improve this page?

What kind of feedback do you have?