Site repository changed after first deploy

I have encountered a weird issue. I created a site manually using netlify-cli:

netlify sites:create -c -m

And used a git url pointing to a self-hosted gitlab instance in the form:

ssh://git@mygitlab.com:2222/group-name/repo-name.git

The first deploy was successful. After making some changes and pushing them, the build failed in the Initializing step with the error below:

3:54:20 PM: Failed during stage "preparing repo": For more information, visit https://docs.netlify.com/configure-builds/troubleshooting-tips/#build-fails-with-error-128: fatal: "group-name/repo-name" does not appear to be a git repository
fatal: Could not read from remote repository.

Checking the build settings, the git repo has changed without any interaction from me.

Initial repo: ssh://git@mygitlab.com:2222/group-name/repo-name.git
Current repo: group-name/repo-name

Since I’m on the starter plan, I can’t change the repo back to the correct one. Any help would be appreciated. Thanks!

1 Like

Could you share the name of the site?

Thanks for the reply! The name of the site is statica-eu-casa-aelia

Could you share your repo’s URL? I can update it from our end.

please share the url

So was this a bug, is this issue going to happen with other sites?

Since that is a private repo in a private gitlab instance, I would not like to share it in a public forum. You can see the correct repo in the logs of the first deployment. If you can’t access the logs, please let me know how I can share the repo privately.

I don’t think I have been able to confirm that. I don’t have a self-hosted instance to check or verify.

I have tried to update the path from our end, let us know if it helps or if this happens again.

@hrishikesh I’m having the same issue with a new site that I created today

Also the repo that you changed should start with ssh://git@. Apologies for not informing you that the repo in the logs is not correct and it is missing the git@ part.

@jonian I’ve escalated this bug to the devs, they’re currently investigating. Thanks for updating us with more context!

@jonian This should now be fixed, if you continue to have issues please let us know!