JIRA to Confluence links broken after migration to Cloud

[Mark Hanson]

After migrating from Server to Cloud (both  JIRA and Confluence), the running the JIRA Macro Repair fixes the links from Confluence to JIRA, but not from JIRA to Confluence, which continue to point to the server.

Does anyone know the cause and how to fix?

[Dario B]

@Mark Hanson @sa 

You are facing some known migration related issues:

• JRACLOUD-62850 - Confluence Links in JIRA issues should be updated automatically based on the new applink configurations
• CONFCLOUD-57382 - Jira Issues Macro not able to retrieve Jira's metadata
• CONFCLOUD-55999 - JIRA Issues macros don't load "Data cannot be retrieved due to an unexpected error." OAuth authentication failed: consumer_key_unknown
• CONFCLOUD-53830 - Links to JIRA Server in Confluence notifications is unchanged after the JIRA Server Base URL was changed

If you are affected by above issues you can open a ticket with Atlassian support in https://support.atlassian.com in order to have the links fixed. 

[Sara Allam]

Thanks Dario,

That's what I did actually, I raised a support ticket and they helped me through it. It is now working fine.

[Jack [AppFox]]

The solution to this is not too difficult if you understand the data structure of the integration and a bit of scripting experience.

Note you only need to do this if the Page ID's are going to change (e.g. when you import a Space individually).

The easiest way to perform this task is to iterate over the imported Jira Issues like follows - for each Jira Issue imported (e.g. from JQL search):

1. Get the applicable links from the Jira Issue
2. For each link:
   1. Ask the 'Original Confluence' instance for the Space Key and Page Name
   2. Using the Space Key and Page Name, ask the 'Target Confluence' instance what the Page ID is for the migrated page
   3. Update the link on the Jira Issue with the new Base URL, Context Path and Page ID.

The following resources should be referenced:

• Jira REST API for Issue Links
• Confluence REST API get Page using Title/Space Key

I recommend using Python for this, as it is quick to sketch out the code, I'd say also use the following to simplify the amount of work needed:

• Jira Python Library
• Confluence Python Library

If you were migrating Server-to-Server, I would recommend dumping the CONTENT table and saving on authenticating all those REST Requests, but in this sort of case (Cloud), we would leave the script running for hours/days etc. and this would be a 'bleed-over' migration activity, which doesn't block people using the system.

[Kian Stack Mumo Systems]

Hey Mark, 

The cause is that the XML file that uploads JIRA is still pointing at the old Confluence server. 

We actually just addressed this issue recently when migrating a client from Server to Cloud.

When performing a migration from one Server instance to another, the issue is relatively easy to solve as a simple "find and replace" of the old domain name with the new domain name will suffice. The issue with moving to cloud is that all of the links that point to Confluence are in the form of https://<domainname>/pages/viewpage.action?pageId=<contentID>. When you move to Cloud, all of the ContentID's are rewritten. 

I was able to fix this issue, but it is a pretty technical process. If you aren't comfortable working with shell scripts, the DB and Rest API, fixing the issue is going to be quite challenging. If you're up for it I can attempt to describe the process, but I don't know how well I can explain it in this format. 

Thanks, 

Kian

[Saeedeh Ghoreishi]

Hey Kian

I faced the same problem, could you please tell me the process?

I'm wondering when the contentID is rewritten then how possible to fix and find the pages in correct order.

Best

Saeedeh

[Kian Stack Mumo Systems]

Have you already performed the migration? Once it's been done, there isn't a way that I've found to perform the fix. I was told by Atlassian support that they have some steps to perform after an instance has been migrated, but it isn't "technically" support.

[Saeedeh Ghoreishi]

Yes, I've done it before but just find out the issue. I opened a support ticket as well but no response yet, I don't know why I had the impression (!!) that the id will be remains the same after migration then didn't check so far. 

Thanks for your response

[Kian Stack Mumo Systems]

Yeah, sorry. I discovered the solution after a test instance of ours went sideways on a restore. I don't have any ability to fix it once it's been pushed up.

[Laxmi Narayana Yepuri]

Kian,

We are migrating from Server to Cloud and identified this issue before actual migration (on a trial run). Our main issue is Confluence links in JIRA are broken with re-written content ID along with the domain change for cloud. Can you share any scripts that might make our life easy? I understand we have more control on the server which is our own to do updates. 

If Atllasian support can give a work around post migration, i am happy to take that as well. 

Appreciate your help.

[Dario B]

Hi @Laxmi Narayana Yepuri ,

Welcome to the Atlassian Community!

This is a known issue. Please get in touch with Atlassian Support by using the following link: https://support.atlassian.com/contact

We have a dedicated team helping customers to migrate to Cloud that will help you fixing the links.

Even if not too explicit this is also mentioned in the Confluence Cloud Migration Assistant documentation page page (second link):

More information and support

We have a number of channels available to help you with your migration.

• For more migration planning information and FAQs, visit the Atlassian Cloud Migration Center.
• Have a technical issue or need more support with strategy and best practices? Get in touch.
• Looking for peer advice? Ask the Atlassian Community.

• Want expert guidance? Work with an Atlassian Partner.

Cheers,
Dario

[Ceifhlb Jiraconf]

@Dario

Unfortunately we get a response from Atlassian support that users have to manually Edit confluence links in JIRA. This is not a solution that i was looking for. Looks like we are on our own based on that response. Hope you can review following support request and help.  

https://getsupport.atlassian.com/servicedesk/customer/portal/49/MOVE-3448

Thanks

[Dario B]

@Ceifhlb Jiraconf , 

The issue for which we have a workaround is the one described in:

• JRACLOUD-62850 - Confluence Links in JIRA issues should be updated automatically based on the new applink configurations

I had a look at the MOVE ticket and I can see you don't want to change the issue links to confluence pages, but links you have entered manually in the issue description.

In this case I am afraid there is not much to do and the answer from the engineers assigned to the ticket is correct.

[Cengiz Akyildiz]

Atlassian support is able to fix those Confluence link in Jira (mention in/on) after migration, but not those shortcut links directly embedded by user into comments, or any text field in Jira. And the difficult part they are not always in the same format. So far I've  come across below link types regarding the info in the url, 

• Content/Page ID type
• Tiny url type
• Page title type (there are two url format versions) 

The last one is relatively easier, you just need to modify the base url in entities.xml for those links.

The second one, yes this one I've found the hardest, could not find a solution. As there were not much, I just neglected in my case.

The first one needs working on it. In my case I had more than 8k page links, so needed to build content id mapping to manipulate entities.xml file. You can get the page/content ids from server, but for getting the same from cloud need either a tool (like Bob-swift), or write you REST API code, to export all the page IDs from cloud. Then create a mapping table to replace all those page/content IDs in your entities.xml file. Well there are several steps and details here...

My suggestion, if you don't have much embedded links like in my case, just ignore of manually fix it and leave the mention in/on or Wiki page links to Atlassian support to fix, which can be done also after migration.

I wish, the Page/Content IDs could have been kept the same after migration, so the problem would be easily solved by just replacing the base url...

Good luck to everyone going through this path...

[Sara Allam]

Same problem here! Jira issues are still linked to the pages on the server instead of the cloud.