[FLINK-4410] Report more information about operator checkpoints

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Implemented
• Affects Version/s: 1.1.2
• Fix Version/s: 1.2.0, 1.3.0
• Component/s: Runtime / State Backends, Runtime / Web Frontend
• Assignee: Ufuk Celebi
• Reporter: Ufuk Celebi
• Created: 17/Aug/16 09:46
• Updated: 04/Feb/22 10:27
• Resolved: 10/Jan/17 08:53

Description

Checkpoint statistics contain the duration of a checkpoint, measured as from the CheckpointCoordinator's start to the point when the acknowledge message came.

We should additionally expose

• duration of the synchronous part of a checkpoint
• duration of the asynchronous part of a checkpoint
• number of bytes buffered during the stream alignment phase
• duration of the stream alignment phase

Note: In the case of using at-least once semantics, the latter two will always be zero.

Issue Links

• causes: FLINK-25958 OOME Checkpoints & Savepoints were shown as COMPLETE in Flink UI
• contains: FLINK-5069 Pending checkpoint statistics gauge
• links to: GitHub Pull Request #3042

Sub-Tasks

• Gather operator checkpoint durations data sizes from the runtime: Closed
• Gather more detailed checkpoint stats in CheckpointStatsTracker: Closed
• Visualize additional checkpoint information: Closed