[FLINK-5230] Safety nets against leaving dysfunctional JobManagers

• Type: Improvement
• Status: Open
• Priority: Not a Priority
• Resolution: Unresolved
• Component/s: Runtime / Coordination
• Labels: auto-deprioritized-major auto-deprioritized-minor
• Assignee: Unassigned
• Reporter: Stephan Ewen
• Created: 02/Dec/16 14:09
• Updated: 07/Jan/22 22:39

Description

There are certain ways that a JobManager can become dysfunctional.

If the JobManager process continues to exist (not restarted by YARN / Mesos) etc, but is not doing its work properly and more, it makes the Streaming Job unavailable.

There some safety nets to bring into place for that, see sub issues.

Sub-Tasks

• JobManager should exit on "OutOfMemoryError": Open