This script can be used to automate backups of Atlassian Cloud JIRA and Confluence instances. It is based on Atlassian Labs' automatic-cloud-backup script:
This repository contains several fixes to the original backup.sh script, which is not even functional in it current state.
This fork was created because Atlassian Labs had ignored all community pull requests and issues for several months.
The script uses curl and jq.
You will need to create a configuration file for the script to work. To do this, copy backup.sh.vars.example to $HOME/.backup.sh.vars and edit it to match your environment.
Using the script is straightforward. To backup JIRA:
./backup.sh --source jiraTo backup Confluence (wiki):
./backup.sh --source wikiTo backup JIRA without attachments:
./backup.sh --source jira --attachments falseArguments overview:
-a, --attachments
Set if attachments should be in the backup. Values are "true" (default) or "false".
-s, --source
Set what backup should be created. Default source is Jira. Set the argument to "wiki" or "confluence" to backup Confluence.
-t, --timestamp
Set if we should overwrite the previous backup or append a timestamp (default) to prevent just that. The former is useful when an external backup program handles backup rotation. Set the argument to "false" if there should be no timestamp in the filename.
The script generates an authorization cookie using your Atlassian credentials if
Atlassian Cloud only allows creation of one backup per day, counted separately for JIRA and Confluence. However, reuse of the cookie allows running the backup script several times in the row; if a new backup could not be created, then the previously created backup is fetched.
The authorization cookie is used for three things:
If you find a problem please file an issue, or better yet, a pull request.
The license of the original script is unclear. All contributions by me (Samuli Seppänen) are licensed under the 2-Clause BSD License: