Uploaded image for project: 'MariaDB ColumnStore'
  1. MariaDB ColumnStore
  2. MCOL-786

redistribute remove moved data from removed node to one node

    XMLWordPrintable

    Details

    • Type: Bug
    • Status: Closed (View Workflow)
    • Priority: Major
    • Resolution: Fixed
    • Affects Version/s: 1.0.9
    • Fix Version/s: 1.1.1
    • Component/s: writeengine
    • Labels:
      None
    • Sprint:
      2017-13, 2017-14, 2017-15, 2017-16, 2017-17, 2017-18, 2017-19, 2017-20, 2017-21, 2017-22

      Description

      Single server node with 5 dbroots with large amount of data loaded. Uncovered that dbroot2 was using the same physical storage as dbroot1. So ran:
      Fri Jun 23 14:18:30 2017: redistributedata start remove 2
      Fri Jun 23 15:02:33 2017: unassignDbrootPmConfig 2 pm1

      However after completion the volume for dbroot1 still was near capacity. Most of the data from dbroot2 was moved to dbroot1 rather than being spread out across dbroot1, 3,4,5 as expected.

      Why did this happen , there is enough data that this should not be an edge case due to small number of partitions i believe.

      Also subsequently redistribute does nothing despite dbroot1 very clearly having more partitions than the others.

        Attachments

          Issue Links

            Activity

              People

              Assignee:
              dleeyh Daniel Lee
              Reporter:
              dthompson David Thompson (Inactive)
              Votes:
              1 Vote for this issue
              Watchers:
              3 Start watching this issue

                Dates

                Created:
                Updated:
                Resolved:

                  Git Integration

                  Error rendering 'com.xiplink.jira.git.jira_git_plugin:git-issue-webpanel'. Please contact your Jira administrators.