Uploaded image for project: 'HBase'
  1. HBase
  2. HBASE-7006

[MTTR] Improve Region Server Recovery Time - Distributed Log Replay

    XMLWordPrintableJSON

Details

    • New Feature
    • Status: Closed
    • Critical
    • Resolution: Fixed
    • None
    • 0.98.0, 0.95.1
    • MTTR
    • None
    • Reviewed
    • Hide
      Distributed Log Replay Description:

      After a region server fails, we firstly assign a failed region to another region server with recovering state marked in ZooKeeper. Then a SplitLogWorker directly replays edits from WAL(Write-Ahead-Log)s of the failed region server to the region after it's re-opened in the new location. When a region is in recovering state, it can also accept writes but no reads(including Append and Increment), region split or merge.

      The feature piggybacks on existing distributed log splitting framework and directly replay WAL edits to another region server instead of creating recovered.edits files.

      The advantages over existing log splitting recovered edits implementation:
      1) Eliminate the steps to write and read recovered.edits files. There could be thousands of recovered.edits files are created and written concurrently during a region server recovery. Many small random writes could degrade the overall system performance.
      2) Allow writes even when a region is in recovering state. It only takes seconds for a failed over region to accept writes again.

      The feature can be enabled by setting hbase.master.distributed.log.replay to true (by default is false)
      Show
      Distributed Log Replay Description: After a region server fails, we firstly assign a failed region to another region server with recovering state marked in ZooKeeper. Then a SplitLogWorker directly replays edits from WAL(Write-Ahead-Log)s of the failed region server to the region after it's re-opened in the new location. When a region is in recovering state, it can also accept writes but no reads(including Append and Increment), region split or merge. The feature piggybacks on existing distributed log splitting framework and directly replay WAL edits to another region server instead of creating recovered.edits files. The advantages over existing log splitting recovered edits implementation: 1) Eliminate the steps to write and read recovered.edits files. There could be thousands of recovered.edits files are created and written concurrently during a region server recovery. Many small random writes could degrade the overall system performance. 2) Allow writes even when a region is in recovering state. It only takes seconds for a failed over region to accept writes again. The feature can be enabled by setting hbase.master.distributed.log.replay to true (by default is false)

    Description

      Just saw interesting issue where a cluster went down hard and 30 nodes had 1700 WALs to replay. Replay took almost an hour. It looks like it could run faster that much of the time is spent zk'ing and nn'ing.

      Putting in 0.96 so it gets a look at least. Can always punt.

      Attachments

        1. ProposaltoimprovelogsplittingprocessregardingtoHBASE-7006-v2.pdf
          130 kB
          Jeffrey Zhong
        2. LogSplitting Comparison.pdf
          50 kB
          Jeffrey Zhong
        3. hbase-7006-combined-v9.patch
          312 kB
          Jeffrey Zhong
        4. hbase-7006-combined-v8.patch
          311 kB
          Jeffrey Zhong
        5. hbase-7006-combined-v7.patch
          315 kB
          Jeffrey Zhong
        6. hbase-7006-combined-v6.patch
          315 kB
          Jeffrey Zhong
        7. hbase-7006-combined-v5.patch
          307 kB
          Jeffrey Zhong
        8. hbase-7006-combined-v4.patch
          298 kB
          Jeffrey Zhong
        9. hbase-7006-combined-v1.patch
          246 kB
          Jeffrey Zhong
        10. hbase-7006-combined.patch
          234 kB
          Jeffrey Zhong
        11. hbase-7006-addendum.patch
          1.0 kB
          Jeffrey Zhong
        12. 7006-addendum-3.txt
          2 kB
          Ted Yu

        Issue Links

          depends upon

          Bug - A problem which impairs or prevents the functions of the product. HBASE-14028 DistributedLogReplay drops edits when ITBLL 125M

          • Major - Major loss of function.
          • Closed
          is related to

          Bug - A problem which impairs or prevents the functions of the product. HBASE-8560 TestMasterShutdown failing in trunk 0.95/trunk -- "Unable to get data of znode /hbase/meta-region-server because node does not exist (not an error)"

          • Major - Major loss of function.
          • Closed

          Bug - A problem which impairs or prevents the functions of the product. HBASE-8567 TestDistributedLogSplitting#testLogReplayForDisablingTable fails on hadoop 2.0

          • Major - Major loss of function.
          • Closed

          Wish - General wishlist item. HBASE-7825 Retire non distributed log splitting related code

          • Major - Major loss of function.
          • Closed
          is required by

          Umbrella - An overarching type made of sub-tasks HBASE-5843 Improve HBase MTTR - Mean Time To Recover

          • Major - Major loss of function.
          • Closed
          relates to

          Sub-task - The sub-task of the issue HBASE-11280 Document distributed log replay and distributed log splitting

          • Major - Major loss of function.
          • Closed

          Bug - A problem which impairs or prevents the functions of the product. HBASE-8701 distributedLogReplay need to apply wal edits in the receiving order of those edits

          • Major - Major loss of function.
          • Closed

          Bug - A problem which impairs or prevents the functions of the product. HBASE-8729 distributedLogReplay may hang during chained region server failure

          • Major - Major loss of function.
          • Closed

          Bug - A problem which impairs or prevents the functions of the product. HBASE-8568 Test case TestDistributedLogSplitting#testWorkerAbort failed intermittently

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Closed

          Improvement - An improvement or enhancement to an existing feature or task. HBASE-8573 Store last flushed sequence id for each store of region for Distributed Log Replay

          • Major - Major loss of function.
          • Closed

          Improvement - An improvement or enhancement to an existing feature or task. HBASE-8617 Introducing a new config to disable writes during recovering

          • Major - Major loss of function.
          • Closed

          Test - A new unit, integration or system test. HBASE-8575 TestDistributedLogSplitting#testMarkRegionsRecoveringInZK fails intermittently due to lack of online region

          • Major - Major loss of function.
          • Closed
          supercedes

          Sub-task - The sub-task of the issue HBASE-6984 Serve writes during log split

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Closed

          Activity

            People

              jeffreyz Jeffrey Zhong
              stack Michael Stack
              Votes:
              0 Vote for this issue
              Watchers:
              31 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: