Is there any way to retain ID's with a migration?

[Mathew Lederman]

We're working on migrating from a Server/Datacenter instance that has been used for 5+ years to the Atlassian Cloud. Is there any way to retain ID's related to Jira and/or Confluence?

The two biggest concerns are Jira Custom Field IDs and Confluence Page IDs. These IDs have been built into API's, integrations, communications, etc. for dozens of teams. Changing these IDs will break just about every integration we have and cause dozens, possibly hundreds of hours of work from various teams to appropriately handle these changes without severely interrupting day-to-day business.

 Any suggestions, recommendations or options would be greatly appreciated!

[Jimmy Van · Accepted solution]

Hi Matthew,

For Jira, if you migrate soon using the Migrate all at once option, then the custom field id's will remain the same. Migrating by any other method will cause the id's to change. We note that it may be necessary to use project by project migration, for example, if one of your marketplace apps mandates it.

For Confluence, there isn't a simple solution as underneath the hood, it's migrating space by space and generating new id's for each space and piece of content. Typically for integrations, it's about redirecting customers to the right documentation. Please check out https://community.atlassian.com/t5/Atlassian-Migration-Program/Don-t-Let-Your-Users-End-up-in-a-Dead-End/ba-p/2508545 if you specially need customers to be redirected from other applications.

You can otherwise easily fix Confluence (and Jira) links inside Confluence (and Jira) by using Update product links.

Cheers,
Jimmy

[Julien MALLARD]

Hi Matthew,

This is very interesting. Thanks !

Does all at once option preserve other IDs, such as those for boards, sprints, filters, projects, resolutions, statuses, issue types, versions, components, etc. ?

By the way, thanks for the Confluence related tips, will be useful !

Regards,

Julien

[Jimmy Van]

Hi Julien,

Yes, Migrate all at once will preserve all the other id's. It overwrites your existing Jira instance and keeps all your old id's. This is another reason you may need to use project-by-project migration - if you have existing data, it will wipe all of it. If there's existing data, we can't create the same entity (project, board, sprint, etc) with the same id as something that already exists in Jira Cloud.

Cheers,
Jimmy

[Julien MALLARD]

Hi Jimmy,

Thank you so much for this confirmation. Obviously duplicate IDs would be problematic.
We are actually migrating to an empty cloud :)

The good thing here is that updating integrated systems with Jira becomes much simpler. For example, for a link pointing to a particular board, including its ID, we just have to modify the host...
This is a major advantage compared to the Project by Project method !

Cheers,

Julien