Uploaded image for project: 'HBase'
  1. HBase
  2. HBASE-28538

BackupHFileCleaner.loadHFileRefs is very expensive

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Open
    • Major
    • Resolution: Unresolved
    • None
    • None
    • backup&restore
    • None

    Description

      I noticed some odd CPU spikes on the hmasters of one of our clusters. Turns out it had been getting lots of bulkoads (30k) and processing them was expensive. The method scans hbase and then parses the paths. Surprisingly the parsing is more expensive than the reading hbase, with the vast majority of time spent in org/apache/hadoop/fs/Path.<init>.

      We should see if this is possible to be optimized. Attaching profile.

      Attachments

        1. async-prof-pid-7-cpu-2.html
          299 kB
          Bryan Beaudreault

        Activity

          People

            Unassigned Unassigned
            bbeaudreault Bryan Beaudreault
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

            Dates

              Created:
              Updated: