Hi
No i did not finde a direct solution however the problem just disappeared after a few houres with frustration and a few dozen reboots of the server
Hope you mange to find a solution because i have encounter it a few times with out luck on finding out how to fix it
The same issue just popped up for me. Have you guys found a fix for this in the meantime? I realize this thread is old but I’m just hunting the solution.
The issue was that I just renewed my certificates with Let’s encrypt and that the certificates were not yet updated for the registry (which runs on a different instance on AWS).
Removing the older certificates for the gitlab-runner and doing a gitlab-ctl reconfigure triggered a copy of the certificates to the other machine.
Today i got the same issue and error message from my gitlab runner, after upgrading the gitlab main version to the newest.
I am running my gitlab with docker, behind an apache. My full gitlab instance is configured to not use certificates in any way, so my apache is handling all ssl related stuff and the certificates are still valid.
The only active part in my config related to Registry is:
registry_external_url=‘https://mygit.domain.com:63322’
Everything else is commented out (like defaults).
This worked perfectly, until i made the Gitlab upgrade today.
Never mind my problem.
I found out, gitlab generated some letsencrypt certificates and used them.
They expire after 90 days, which indeed expired shortly after i upgraded the gitlab instance yesterday… That’s called bad luck!
i tried to renew the certificates with ‘gitlab-ctl renew-le-certs’ which sadly was not successful and showed multiple errors, which is another issue (which belongs into another thread)!
Now i ended up using my real certificate which i also use in my apache instance.