Uploaded image for project: 'Jira Server and Data Center'
  1. Jira Server and Data Center
  2. JRASERVER-67732

Copying index snapshots between Data Center nodes should be faster

    XMLWordPrintable

    Details

    • UIS:
      1
    • Support reference count:
      1
    • Feedback Policy:
      We collect Jira feedback from various sources, and we evaluate what we've collected when planning our product roadmap. To understand how this piece of feedback will be reviewed, see our Implementation of New Features Policy.

      Description

      Problem

      Data Center nodes can ask another node to provide a full index snapshot when:

      1. A new node is started and has no index
      2. Index replication has fallen sufficiently far behind
      3. The copy request has been triggered manually

      Currently, this can take upwards of an hour, making the node where the index is broken or missing unusable. The chief contributor to this is the time it takes to zip the index, and this can be shortened by using a faster compression algorithm as long as the I/O device used for shared storage is fast enough.

      Suggested fix
      • The compression method used can be configured with the jira.index.snapshot.copy.archiver application property.
      • Format: jira.index.snapshot.copy.archiver=(snappy|zip|tar)
      • setting the jira.index.snapshot.copy.archiver application property to zip will revert to the old behavior.

        Attachments

          Issue Links

            Activity

              People

              Assignee:
              klopacinski Karol Lopacinski
              Reporter:
              klopacinski Karol Lopacinski
              Votes:
              2 Vote for this issue
              Watchers:
              10 Start watching this issue

                Dates

                Created:
                Updated:
                Resolved: