Skip to content

[flink] Optimize failover recovery for idempotent aggregations#3706

Open
platinumhamburg wants to merge 2 commits into
apache:mainfrom
platinumhamburg:feature/merge-engine-connector-failover-recovery
Open

[flink] Optimize failover recovery for idempotent aggregations#3706
platinumhamburg wants to merge 2 commits into
apache:mainfrom
platinumhamburg:feature/merge-engine-connector-failover-recovery

Conversation

@platinumhamburg

Copy link
Copy Markdown
Contributor

Purpose

Linked issue: close #3703

Brief change log

Tests

API and Format

Documentation

@platinumhamburg
platinumhamburg force-pushed the feature/merge-engine-connector-failover-recovery branch 2 times, most recently from dd720ea to 631ec14 Compare July 20, 2026 18:06
Expose duplicate-application idempotence as aggregation metadata.

Verify every aggregation implementation with a parameterized real-merger contract.
Use no-op recovery when all effective aggregation operations are idempotent, and retain undo recovery otherwise.

Preserve operator topology and checkpoint compatibility, clean stale recovery metadata, and cover SQL and DataStream failover paths.
@platinumhamburg
platinumhamburg force-pushed the feature/merge-engine-connector-failover-recovery branch from 631ec14 to c55d63c Compare July 21, 2026 04:43
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[flink] Skip unnecessary undo recovery for idempotent aggregations

1 participant