We are considering migrating from Jira Cloud to Jira Data Center. Will automations transfer successfully, what difficulties may arise, and how can they be addressed? Thank you!
Be warned: virtually every ID will be different (only the same as an extremely rare coincidence).
The gap here includes IDs for Projects, Statuses, Fields, Transitions, Components, Issue Types, Issue Link Types, Priorities, Resolutions, User IDs, etc.
Since automation rules internally use IDs specific to your Jira instance, that's what's exported into the JSON. Importing that same JSON into another instance results in all those IDs no longer matching.
Inspecting an imported rule might show a problem (unknown ID), which is the best case scenario. But often IDs will exist in both instances but map to different things, and then the imported rule does not show an error. But the imported rule will then work incorrectly.
We are in the midst of trying to mitigate about 200 automation rules while migrating from Data Center to Cloud. Unfortunately we didn't learn about this "limitation" until about 3 months before our planned go-live date.
If you use Jira automation rules much at all, this is a major gap (as of March 2025) with migrating to Jira Cloud from Server or Data Center. Our migration vendor didn't have a way to mitigate this, nor did Atlassian provide anything to help (e.g. JCMA functionality). And that's sad since it's just JSON data and IDs with known types in their rule contexts. Mapping IDs by name from one instance to another is not rocket science.
There are still a handful of things in automation rules that couldn't be fixed even with the postulated additional JCMA functionality. However, an automated process could highlight the remaining suspects for manual review -- a huge improvement.
I'm creating some custom scripts to rewrite our exported rule JSON before importing to Cloud, and I expect to mitigate 95% of these gaps. Unfortunately it's nothing I can share due to business IP limitations. As a Jira admin and a former software developer, this challenge happened to be in my wheelhouse.
I agree, if you have a lot of auatomation rules validating/ replacing the IDs is challenging.
I hope your script turns out well!
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi @DenJr
You can export all rules from Cloud as a single JSON file and import that JSON in your DC instance. This will import all rules as disabled and you can enable them 1 by 1.
While the import is certainly helpful it does not take care of everything. IDs can be different on both environments. e.g. an automation rules creates an issue of type Task with ID 5. If issue type with ID 5 is a bug on your DC instance it will create a bug instead.
So be careful, you need to validate your rules where it uses IDs.
The export is a JSON so you could also prepare/ script the changes in the JSON file before importing.
Also make sure the "rule actor" of the automation rules have the necessary permissions on DC.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.