That configuration definitely works, we have production Confluence Server linked to JIRA Cloud. I cannot recall any gotchas related to JIRA links, except users need to authenticate to JIRA once migrated. Other gotchas:
If you are using JIRA for authentication, you will need to manage those users separately in the new Confluence Server once migrated. Atlassian Crowd will not fix this, as JIRA Cloud does not support. Users/Passwords will be copied at the time you backup, so you do not need to re-enter into Confluence Server, but moving forward they will not be linked.
Cloud instances are upgraded weekly, so you need to do migration, UAT, cutover in between those upgrades. We migrated, tested, and then had a problem where backups were broken in Cloud (it happens - https://jira.atlassian.com/browse/CONF-39685, https://jira.atlassian.com/browse/CONF-40357, https://jira.atlassian.com/browse/CONF-40687) and so that caused a week (or so) delay and a missed GoLive before the bug was fixed. Then, we got our backup, restored on an OD version running on our server, and the OD version had a severe performance issue (not seen in UAT) which after 100 man-hours, turned out to be an Atlassian bug, and we had to wait for next release before that was fixed.
So, on one hand you need to quickly backup/restore before OD updates - and your data/server version are in sync. On the other hand, you need to take time to (load) test the OD version running on your server. If you have a small number of users, this is less worrisome.
In any case, it can be done! HTH
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.