For our project we ran an instance of GitLab CE for almost six years now (started using v8 back then). The project is about to end and in the future there won’t be any resources to keep the GitLab up and running, take care about updates and so on.
Still it would be sad to just drop all the work we did, so I thought about archiving the content of the GitLab instance in plain HTML and CSS plus the Git repos on a static webserver on the same domain, so that links won’t 404 and people could still browse the old project pages and clone the repos using HTTP.
Has anyone ever done this to a GitLab?
My approach would be to set everything public that is ok to do so, and then do
wget -r … for the HTML part.
For the Git repos, I was going to rsync the directory holding all the repositories somewhere down at
/var/opt/gitlab but it looks like with the rise of Gitaly, this folder has disappeared.
How would I download all the repositories to a folder structure equivalent to the
I don’t care about packages or registries, artifacts or build logs, nor anything else that GitLab has in stock today. Just a shallow clone of the work that has gone into the project.
Who can give advice how to approach this? Any ideas?