Uploaded image for project: 'Cassandra'
  1. Cassandra
  2. CASSANDRA-14084

Disks can be imbalanced during replace of same address when using JBOD

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Resolved
    • Normal
    • Resolution: Fixed
    • 3.11.2, 4.0-alpha1, 4.0
    • None
    • None
    • Normal

    Description

      While investigating CASSANDRA-14083, I noticed that we use the pending ranges to calculate the disk boundaries when the node is bootstrapping.

      The problem is that when the node is replacing a node with the same address, it sets itself as normal locally (for other unrelated reasons), so the local ranges will be null and consequently the disk boundaries will be null. This will cause the sstables to be randomly spread across disks potentially causing imbalance.

      Attachments

        1. dtest14084.png
          166 kB
          Paulo Motta

        Issue Links

          relates to

          Improvement - An improvement or enhancement to an existing feature or task. CASSANDRA-12344 Forward writes to replacement node with same address during replace

          • Normal -
          • Open

          Improvement - An improvement or enhancement to an existing feature or task. CASSANDRA-14083 Avoid invalidating disk boundaries unnecessarily

          • Normal -
          • Resolved

          Activity

            People

              pauloricardomg Paulo Motta
              pauloricardomg Paulo Motta
              Paulo Motta
              Marcus Eriksson
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: