Uploaded image for project: 'Cassandra'
  1. Cassandra
  2. CASSANDRA-1181

kinder gentler compaction

    XMLWordPrintableJSON

Details

    • Task
    • Status: Resolved
    • Normal
    • Resolution: Fixed
    • 0.6.3
    • None
    • None

    Description

      I suggested this in a ML thread but it seems that nobody actually tried it. I think it's worth following up on:

      You could try setting the compaction thread to a lower priority. You could add a thread priority to NamedThreadPool, and pass that up from CompactionExecutor constructor. According to http://www.javamex.com/tutorials/threads/priority_what.shtml you have to run as root and add a JVM option to get this to work.

      In particular, Brandon saw stress.py read latencies spike to 100ms during [anti]compaction on a 2 core machine. I'd like to see if this can mitigate that.

      Attachments

        1. 1181-trunk.txt
          4 kB
          Brandon Williams
        2. stats.txt
          4 kB
          Edward Capriolo
        3. 1181.txt
          3 kB
          Jonathan Ellis
        4. CompactionManager.java
          24 kB
          Edward Capriolo

        Activity

          People

            brandon.williams Brandon Williams
            jbellis Jonathan Ellis
            Brandon Williams
            Jonathan Ellis
            Votes:
            0 Vote for this issue
            Watchers:
            4 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: