JIRA is now available as three separate applications, JIRA Software, JIRA Service Desk, and JIRA Core. For more information on administering these applications, refer to the Administering JIRA Applications documentation.

Automating JIRA Backups

JIRA can be configured to automatically create an XML backup of JIRA data on a routine basis.

(info) Please Note:

For production use or large JIRA installations, it is strongly recommended that you use native database-specific tools instead of the XML backup service. XML backups are not guaranteed to be consistent, as the database may be updated during the backup process. Inconsistent backups are created successfully without any warnings or error messages, but fail during the restore process. Database-native tools offer a much more consistent and reliable means of storing data.

To configure automated JIRA backups:

  1. Log in as a user with the JIRA System Administrators global permission.
  2. Select Administration > System > Advanced > Services (tab) to open the Services page, which lists the current services running on this system. By default, there should be at least one 'Mail Queue Service' running, which cannot be deleted.
    (tick) Keyboard shortcut: g + g + start typing services
  3. In the Add Service form towards the end of the page, complete the following fields:
    • Name — a descriptive name for the backup service, such as Backup Service.
    • Class — the appropriate fully-qualified class name for the Backup service using either of the following methods:
      • Select the Backup service from the list of JIRA's Built-in Services. To do this:
        1. Click the Built-in Services link below the Class field to expand the list of JIRA's built-in service classes.
        2. Click the Backup service link. The Class field will automatically be populated with the following class text string 'com.atlassian.jira.service.services.export.ExportService'
      • Type the fully-qualified class name 'com.atlassian.jira.service.services.export.ExportService' into the Class field.
    • Delay — enter the number of minutes between backups. A good default for this would be 720 minutes (12 hours) or 1440 minutes (24 hours).
      (info) Please Note: The interval specified in the Backup Service Delay (mins) is the time when the next backup job will run since the last server restart. Backup services cannot be scheduled to run at a specific time of day - please see JRA-1865 for more on this.
  4. Click the Add Service button. The Edit Service page is displayed.
  5. Complete the following items on this page:
    • For the Date format field, specify the format which JIRA will use to name the individual backup files. This format can be anything that SimpleDateFormat can parse. A good default is 'yyyy-MMM-dd-HHmm', which would generate files named like this: '2007-Mar-05-1322'.
    • For the Delay field, modify the number of minutes between backups if necessary.
    • If the Use Default Directory check box is displayed, see the note below.
  6. Click the Update button. Your backup service is now configured. XML backups will be performed according to the schedule you specified in the Delay field.
    • For every successful backup, a zipped file of your XML backup will be saved in the backup directory.
    • If a scheduled backup fails for any reason, the zipped XML backup file will be saved into the 'corrupted' directory, which is directly under your nominated backup directory. A file explaining the reason for the failure will be written to the 'corrupted' directory. This file will have the same name as the backup file, but with the extension '.failure.txt'.
      (info) JIRA will create the 'corrupted' directory if required - you do not need to create it.

About custom backup directories

The Use Default Directory check box (not shown in screenshot above) is for legacy JIRA installations (prior to JIRA 4.2), which have backup services that use custom directories.

If you are using JIRA 5.1.0 or earlier, the Use Default Directory will always be displayed, as the option of using custom directories has been deprecated. If you are using JIRA 5.1.1 or later, the Use Default Directory check box will only be displayed if you upgraded from a version of JIRA prior to 4.2 and you are editing an existing backup service which used a custom directory.

  • If you are not using a legacy backup service with a custom directory, select the the Use Default Directory check box. If you do not, your backup service may not work correctly
  • If you are using a legacy backup service with a custom directory, you can choose between using the default directory or your custom directory (cannot be edited). Note, if you choose the default directory option, you will not be able to choose the custom directory option.

The default directory location is the export subdirectory of the JIRA Home Directory.

Was this helpful?

Thanks for your feedback!

Why was this unhelpful?

13 Archived comments

  1. User avatar


    Is there an option somewhere to do an automatic cleanup of backups? Say - keep only recent backups for one last week or month etc. Thank you.

    25 Jun 2012
    1. User avatar


      I would also be interested in it..


      03 Jul 2012
    1. User avatar

      Omar Shehab

      That would be nice.

      19 Jun 2013
    1. User avatar

      Ed Graham

      create a cron file in /etc/cron.d/mycompany with the following contents

      ### delete old backups from JIRA's automated backups dir that are 15 days old or older
      5 6 * * * root /usr/bin/find /var/atlassian/application-data/jira/export/ -mtime +15 -delete
      14 Nov 2014
  2. User avatar


    under Windows use following command line script to remove files older than 7 days:

    set jira_backup_dir="C:\Program Files\Atlassian\Application Data\JIRA\export" :: set your JIRA backup directory here
    forfiles /P %jira_backup_dir% /M *.zip /D -7 /C "cmd /c del @path >> jira_backup.log"

    put this into a .bat file and configure Task Scheduler to run that file once a day.
    You may also run forfiles /? from the command line for more options.

    18 Jul 2012
  3. User avatar


    Have you tried to set the SimpleDateFormat to "E" (means Day of week).

    What I don't know is if the old files will be overwriten.

    07 Dec 2012
  4. User avatar


    I thought I would post this here... but I found this post: http://alexsuslin.com/blog/2012/02/06/daily-backup-atlassian-jira-and-confluence-on-windows/


    That script helps a lot for windows backups.

    29 Jan 2013
  5. User avatar

    John Desha

    How do you turn off the default every 12 hour backups? A link to those instructions (if they exist) would be great in the "For production use or large JIRA installations" warning at the top.

    10 Jul 2013
    1. User avatar

      John Desha

      I found it. My screen was not wide enough to see the Delay column and Delete/Edit links on the Services page.

      10 Jul 2013
  6. User avatar


    I just use the date format "EEEE" the process with take place every day and will be named Monday, Tuesday, Wednesday .... So by Monday of the next week the last weeks file of Monday is overwritten. 

    13 Sep 2013
  7. User avatar

    Jim Gust

    I don't know about the newer versions, but 6.1 does not overwrite older backup files if you have a simple date format of "E" or "EEEE". I'm in Ubuntu so it may be a permissions issue, but I haven't been able to get it to work correctly. 

    21 Apr 2014
  8. User avatar

    Yuriy Vasylchenko

    Where can I disable the compression for backup archives?

    I use the de-duplicated file system for backups and effect of compression is absolutely negative.

    15 May 2014
  9. User avatar

    Dovid Bender

    Will simply taking a copy of:

    1) Install directory

    2) Data directory

    3) A dump of PostGres DB

    Work as a backup solution? It seems to be a lot easier to do run a cron job that will jsut scp all the files to another server nightly.


    23 Mar 2015
Powered by Confluence and Scroll Viewport