Migrating Flink jobs

After upgrading the version of the Data Lake in your Cloudera on cloud environment, you need to create Cloudera Data Hub clusters using the Cloudera Streaming Analytics cluster template that corresponds to the new version of the Data Lake. To restart your existing Flink jobs on the upgraded Cloudera Data Hub clusters, you need to migrate your Flink jobs. The migration process depends on the state of the Flink jobs.

After creating a new Cloudera Data Hub cluster using the latest Cloudera Streaming Analytics cluster template, you need to migrate your Flink jobs to the new cluster. The process of migration depends on the state of your Flink jobs: if you have stateful jobs, you need to stop your Flink applications with creating savepoints. Before starting the job migration, make sure that the job dependies are updated based on the latest version of Flink in Cloudera on cloud.