Uploaded image for project: 'HBase'
  1. HBase
  2. HBASE-6055 Offline Snapshots in HBase 0.96
  3. HBASE-7339

Splitting a hfilelink causes region servers to go down.

    XMLWordPrintableJSON

    Details

    • Type: Sub-task
    • Status: Closed
    • Priority: Blocker
    • Resolution: Fixed
    • Affects Version/s: hbase-6055
    • Fix Version/s: 0.95.0
    • Component/s: snapshots
    • Labels:
      None
    • Hadoop Flags:
      Reviewed

      Description

      Steps:

      • Have a single region table t with 15 hfiles in it.
      • Snapshot it. (was done using online snapshot from HBASE-7321)
      • Clone a snapshot to table t'.
      • t' has its region do a post-open task that attempts to compact region. policy does not compact all files. (default seems to be 10)
      • after compaction we have hfile links and real hfiles mixed in the region
      • t' starts splitting
      • creating split references, opening daughers fails
      • hfile links are "split", creating hfile link daughter refs. <<hfile>-<region>-<table>>.<parentregion>
      • these "split" hfile links are interpreted as hfile links with table <table>.<parentregion> -> <<hfile>-<region>>-<<table>.<parentregion>> (groupings interpreted incorrectly)
      • Since this is after the splitting PONR, this aborts the server. It then spreads to the next server.

        Attachments

        1. hbase-7339.patch
          19 kB
          Jonathan Hsieh
        2. pre-hbase-7339.patch
          834 kB
          Jonathan Hsieh
        3. hbase-7339.v2.patch
          19 kB
          Jonathan Hsieh
        4. pre-hbase-7339.v2.patch
          1.04 MB
          Jonathan Hsieh

          Activity

            People

            • Assignee:
              jmhsieh Jonathan Hsieh
              Reporter:
              jmhsieh Jonathan Hsieh
            • Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved: