Skip to content

[SPARK] Fix streaming writer overwriting batch data by scoping OVERWRITE_BY_FILTER to batch only#7458

Open
gregdiy wants to merge 1 commit intoapache:masterfrom
gregdiy:fix/streaming-overwrite-capability
Open

[SPARK] Fix streaming writer overwriting batch data by scoping OVERWRITE_BY_FILTER to batch only#7458
gregdiy wants to merge 1 commit intoapache:masterfrom
gregdiy:fix/streaming-overwrite-capability

Conversation

@gregdiy
Copy link

@gregdiy gregdiy commented Mar 18, 2026

Fixes #7457

…ITE_BY_FILTER to batch only

Purpose

Linked issue: close #7457

Tests

API and Format

Documentation

Generative AI tooling

Note: There is no valid streaming use case for OVERWRITE_BY_FILTER.
Streaming overwrites are already handled explicitly via OutputMode.Complete()
in PaimonSink.scala. Advertising OVERWRITE_BY_FILTER to streaming writers
serves no purpose and only introduces this destructive side effect during
writer initialization.

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.

[BUG][SPARK] First streaming micro-batch overwrites batch data on tables with metadata.iceberg.storage enabled

2 participants