Github user zsxwing commented on a diff in the pull request: https://github.com/apache/spark/pull/14553#discussion_r83965244 --- Diff: sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/StreamExecution.scala --- @@ -336,17 +342,27 @@ class StreamExecution( } if (hasNewData) { reportTimeTaken(OFFSET_WAL_WRITE_LATENCY) { - assert( - offsetLog.add(currentBatchId, availableOffsets.toCompositeOffset(sources)), + assert(offsetLog.add(currentBatchId, availableOffsets.toCompositeOffset(sources)), s"Concurrent update to the log. Multiple streaming jobs detected for $currentBatchId") logInfo(s"Committed offsets for batch $currentBatchId.") + // Now that we've updated the scheduler's persistent checkpoint, it is safe for the + // sources to discard data from *before* the previous batch. + // The scheduler might still request the previous batch from a source in some cases + // if a crash and recovery occured. + val prevBatchOff = offsetLog.get(currentBatchId - 2) --- End diff -- Can we just commit `currentBatchId - 1`? In which case, it will fail?
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org