Flink: Add equality delete conversion committer#16874
Merged
Merged
Conversation
Adds EqualityConvertCommitter, a two input operator for the equality delete conversion task. It buffers the DVWriteResults from the parallel EqualityConvertDVWriter instances and the EqualityConvertPlan from the planner, then commits once both the plan and its done-timestamp watermark have arrived. The committer adds both new staging data files and the writer's merged DVs, removes the superseded DVs, and carries over the remaining staging deletes. It validates against the planner's main snapshot, so external changes on the target branch fail the commit. The commit summary records the processed staging snapshot id, which the planner reads on restart to skip already-committed snapshots, to ensure idempotency. The committer is responsible for sending Trigger records downstream to the TaskResultAggregator of the surrounding table maintenance framework. It emits one after every cycle, including no-op and error, so the task always completes. On an upstream abort or a commit failure it deletes the DVs written this cycle so the failure doesn't leak Puffin files.
wombatu-kun
approved these changes
Jun 20, 2026
Contributor
|
Thanks @mxm for the PR! Thanks @wombatu-kun for the review! |
This was referenced Jun 20, 2026
huaxingao
pushed a commit
that referenced
this pull request
Jun 20, 2026
huaxingao
pushed a commit
to huaxingao/iceberg
that referenced
this pull request
Jun 27, 2026
…OSS path) Adds a runnable proof that the ConvertEqualityDeletes Flink maintenance task converts equality deletes on a staging branch into deletion vectors (DVs) on main, leaving no equality deletes behind, on stock OSS Apache Iceberg. - flink/v2.1 PoC test (TestEqualityDeleteToDVPoc): asserts before/after state explicitly (staging has the eq-delete; after conversion main has a DV, zero eq-deletes, and reads the delete-applied rows). - poc-eq-delete-to-dv/: README, INVOCATION.md (jar build + task invocation), CUSTOMER_FLINK_SETUP.md (customer consumption). Verified passing against apache/iceberg main with all six ConvertEqualityDeletes PRs merged (apache#16831, apache#16844, apache#16858, apache#16874, apache#16889, apache#16948). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
huaxingao
pushed a commit
to huaxingao/iceberg
that referenced
this pull request
Jun 27, 2026
…OSS path) Adds a runnable proof that the ConvertEqualityDeletes Flink maintenance task converts equality deletes on a staging branch into deletion vectors (DVs) on main, leaving no equality deletes behind, on stock OSS Apache Iceberg. - flink/v2.1 PoC test (TestEqualityDeleteToDVPoc): asserts before/after state explicitly (staging has the eq-delete; after conversion main has a DV, zero eq-deletes, and reads the delete-applied rows). - poc-eq-delete-to-dv/: README, INVOCATION.md (jar build + task invocation), CUSTOMER_FLINK_SETUP.md (customer consumption). Verified passing against apache/iceberg main with all six ConvertEqualityDeletes PRs merged (apache#16831, apache#16844, apache#16858, apache#16874, apache#16889, apache#16948). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
jakelong95
pushed a commit
to jakelong95/iceberg
that referenced
this pull request
Jun 29, 2026
* Flink: Add equality delete conversion committer Adds EqualityConvertCommitter, a two input operator for the equality delete conversion task. It buffers the DVWriteResults from the parallel EqualityConvertDVWriter instances and the EqualityConvertPlan from the planner, then commits once both the plan and its done-timestamp watermark have arrived. The committer adds both new staging data files and the writer's merged DVs, removes the superseded DVs, and carries over the remaining staging deletes. It validates against the planner's main snapshot, so external changes on the target branch fail the commit. The commit summary records the processed staging snapshot id, which the planner reads on restart to skip already-committed snapshots, to ensure idempotency. The committer is responsible for sending Trigger records downstream to the TaskResultAggregator of the surrounding table maintenance framework. It emits one after every cycle, including no-op and error, so the task always completes. On an upstream abort or a commit failure it deletes the DVs written this cycle so the failure doesn't leak Puffin files. * fixup! Rename stale DV resolver/merger references to DV writer * fixup! Reword inaccurate watermark-absorption javadoc on the committer * fixup! Add shared-branch DV-merge regression test for the committer
jakelong95
pushed a commit
to jakelong95/iceberg
that referenced
this pull request
Jun 29, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds EqualityConvertCommitter, a two input operator for the equality delete conversion task. It buffers the DVWriteResults from the parallel EqualityConvertDVWriter instances and the EqualityConvertPlan from the planner, then commits once both the plan and its done-timestamp watermark have arrived.
The committer adds both new staging data files and the writer's merged DVs, removes the superseded DVs, and carries over the remaining staging deletes. It validates against the planner's main snapshot, so external changes on the target branch fail the commit. The commit summary records the processed staging snapshot id, which the planner reads on restart to skip already-committed snapshots, to ensure idempotency.
The committer is responsible for sending Trigger records downstream to the TaskResultAggregator of the surrounding table maintenance framework. It emits one after every cycle, including no-op and error, so the task always completes. On an upstream abort or a commit failure it deletes the DVs written this cycle so the failure doesn't leak Puffin files.
This commit is factored out of #15996.