Atlassian Cloud instances are using RDS and those instances are being backed up on a daily basis. However, after inquiring about those backups to roll back changes in Atlassian Jira Cloud and/or Atlassian Confluence Cloud, a reply was given that Administrator(s) should implement their own backup strategy based on Track storage and move data across products instead.
The approach above, even if not ideal, will provide users with a way to backup some data and roll back their instances in case of mistakes (project deleted in Jira / space deleted in Confluence) or any other reason. However, since developers can create 3rd party plugins using Atlassian Connect and/or Forge, how a user can determine whether or not the data used by the plugin gets backed up using the backup strategy described in link above?
Furthermore, how the backup strategy mentioned on link above differs from scripts in Bitbucket ? Do they generate the same ZIP file as a backup?
how a user can determine whether or not the data used by the plugin gets backed up using the backup strategy described in link above?
The short answer, restore it. I know this sounds tedious but there’s some wisdom about backups, “Untested backups & DR plans are just wishful thinking.” Otherwise, you might ask the app vendors if their data is part of an export.
Do they generate the same ZIP file as a backup?
Yes. The docs explain how to backup through the UI. The script does it through an otherwise undocumented REST API endpoint. The result is supposed to be the same (browser vs curl differences not withstanding).
@RafaelPintoSperafico and @ibuchanan
As the question is about cloud backup: You need to ask the plugin vendor if they store their data within Jira/Confluence, or in their infrastructure. If it is in Jira/Confluence, it will be backed up. Otherwise not.