The GitHub Actions job "Required Checks" on texera.git/sync-result-bounded-reads has succeeded. Run started by GitHub user aglinxinyuan (triggered by aglinxinyuan).
Head commit for run: dea2e58876f89b47628a48231265068290ece7b1 / Xinyuan Lin <[email protected]> fix(amber): bound sync result reads so visualization fetches release their reader SyncExecutionResource.collectOperatorResult read operator results through an unbounded document.get() iterator; the visualization and oversized-first-tuple branches returned after a single next(), and the catch-all swallowed exceptions mid-drain, leaving the Iceberg Parquet reader / S3 input stream open until GC finalization. Size each read to what the method actually consumes: getRange(0, 1) for the first tuple, getRange(1, totalCount) for the truncation loops (which always drain), and drain the bounded remainder in the catch path. Bounded getRange iterators release their reader inside the next() that serves their last record (#6882), so no close API is needed. Report URL: https://github.com/apache/texera/actions/runs/30143170612 With regards, GitHub Actions via GitBox
