How to diagnose and repair attachement file links in jira tickets to the attachment data storage?

[Mirko]

Hi all,

I run Jira from a docker container and have used a symbolic link on the linux machine to store my attachements outside the JiraHome directory.

So I have a /JiraData/JiraHome, which ist configured as a starting parameter in the docker container,

and in /JiraData/JiraHome/data/attachments is a symbolic link to /JiraData/JiraAttachments.

This has worked fine so far.

A few days ago I updated the JiraDocker Container to 8.20, it didn't work the just start the new container with the existing /JiraData/JiraHome, so I had to delete it, create an empty JiraHome to install to and then did a System Restore from the Backup I had made before the update.

So far Jira is working, all my projekts have been restored, but I cannot access any of the attachements!

I tried to copy the attachements from the /JiraData/JiraAttachments folder to /JiraData/JiraHome/data/attachments to se if the symbolic link ist the issue but with no luck.

So alle the attachements are there but Jira does not find them.

How can I diagnose or repair the file links?

Thank you very much!

[Florian PEREZ]

Hello,

I would suggest you 2 things:

• First try to recreate the index
• If it don't work you can try to use this procedure. I've done this before and it was very helpful!

https://community.atlassian.com/t5/Jira-questions/JIRA-Script-fix-missing-attachments-in-database/qaq-p/663710

Hope this help! 

[Mirko]

Hi florianper,

thank you for your hints. I read through the post and am not shure if I understood right.

This quote

I have reworked gathering script provided by Atlassian to make him remove database metadata for not existed attachments ONLY!

to me sounds as if it would only delete dead filelinks, but does not "repair" them. So I still could not access my attachements right?

Then could you please explain how recreating the index will help and how to do that? I have to ask since I can hack a few linux commands but am not deep enough in Jira to grasp what this is really doing... thank you!