I did what no devOps person should ever do… update Gitlab on a friday afternoon…
Worst, I did it the wrong way!
What I did:
We were on gitlab-ee:14.0.0, and our setup was flagged using “latest”. So when I rebooted the stack, it tried to update to 14.10.3
The update failed and the stack would not reboot… at all. I managed to revert to forcing 14.0.0 and the server was back online (I remember running gitlab rake commands as suggested in many posts)
My colleagues were happy because they can use gitlab again, and I told myself I would attempt the update another time.
A few days later, I realized that all admin users could not login to gitlab, at all! They were getting ERROR 500 (I did not log in after the service was back online, so I did not realize right away).
We, admins, can’t reach any section of the site. The backend works fine, I can still push and pull data from the repos in the terminal as I always did… This is really weird and I can’t find how toi fix the issue.
I spent around 30 hours with a colleague trying to fix it before asking.
If my tone or the way I asked my question does not comply with the ways of this community, please feel free to correct me, I am new here!