Revision history [back]

There could be several reasons why your TaskManager fails and terminates your job after switching to Flink 1.10.0. Here are some possible causes:

1. Incompatible job configuration: Flink 1.10.0 introduced several configuration changes that may not be compatible with your existing job configuration. Check the release notes and migration guide to make sure that your job configuration is updated properly.

2. Memory allocation issues: Flink 1.10.0 introduced a new memory management system that requires more memory allocation than previous versions. Make sure that your TaskManager machines have enough memory to handle the new requirements.

3. Library conflicts: Flink 1.10.0 introduced changes to the libraries used for serialization and data exchange. If you have custom serialization or data exchange code, it may be incompatible with the new libraries and cause TaskManager failures.

4. Bug in Flink 1.10.0: There could be a bug in Flink 1.10.0 that is causing the TaskManager failures. Check the Flink issue tracker and forums for any reported issues and possible workarounds.

To troubleshoot the issue, check the TaskManager logs for any error messages or stack traces. This can help identify the root cause of the problem and guide your troubleshooting efforts.