[FLINK-26279] Provide mailbox delay metric

• Type: New Feature
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Fix Version/s: 1.16.0
• Component/s: Runtime / Metrics
• Labels: pull-request-available
• Assignee: Sebastian Mattheis
• Reporter: Sebastian Mattheis
• Created: 21/Feb/22 10:02
• Updated: 21/Mar/22 11:16
• Resolved: 21/Mar/22 11:12

Description

Related to FLINK-25414. In order to better debug Task being blocked in busy state, we can periodically enqueue a mailbox action and check how long it took for the task thread to execute such action. We can report this delay as task thread delay/mailbox delay. Would be helpful for debugging why unaligned checkpoints are taking so long to trigger/execute.

Issue Links

• links to: GitHub Pull Request #18991