[GitHub] [spark] HeartSaVioR commented on pull request #31989: [WIP][SPARK-34891][SS] Introduce state store manager for session window in streaming query

HeartSaVioR commented on pull request #31989: URL: https://github.com/apache/spark/pull/31989#issuecomment-810697899 I just simply remedied write amplification a bit via determining existing values to be "deleted" or "replaced", instead of blindly judging it as "deleted". -- This is an

[GitHub] [spark] HeartSaVioR commented on pull request #31989: [WIP][SPARK-34891][SS] Introduce state store manager for session window in streaming query

HeartSaVioR commented on pull request #31989: URL: https://github.com/apache/spark/pull/31989#issuecomment-809067233 Except the test suite, one more thing worths to address here is write amplification; we "blindly" replace all start times and all sessions. This could bring unnecessary writ