Uploaded image for project: 'Spark'
  1. Spark
  2. SPARK-38178

Correct the logic to measure the memory usage of RocksDB

Attach filesAttach ScreenshotVotersWatch issueWatchersCreate sub-taskLinkCloneUpdate Comment AuthorReplace String in CommentUpdate Comment VisibilityDelete Comments
    XMLWordPrintableJSON

Details

    • Bug
    • Status: Resolved
    • Major
    • Resolution: Fixed
    • 3.2.1
    • 3.3.0, 3.2.2
    • Structured Streaming
    • None

    Description

      Currently, structed streaming would report the memory usage of RocksDB as state store's metrics. However, it only sums from rocksdb.estimate-table-readers-mem and rocksdb.size-all-mem-tables. However, this lacks of the memory usage of block cache. (You can refer to [Memory-usage-in-RocksDB](https://github.com/facebook/rocksdb/wiki/Memory-usage-in-RocksDB) for more details.)

      BTW, as the "block-cache-pinned-usage" is included in "block-cache-usage", we don't need to include the pinned usage.

      Attachments

        Activity

          This comment will be Viewable by All Users Viewable by All Users
          Cancel

          People

            yunta Yun Tang
            yunta Yun Tang
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved:

              Slack

                Issue deployment