Restoring Self-Hosted Instance - Login Issue
We are running a self-hosted Windmill Instance with Docker on a Linux server spun up in Azure. Very small env. I am in the process of testing restores on it and when I restore the server, I can get to it via http://azure-server-ip:8000 but I cannot login. I have MS SSO turned on, but also have a password for my user account. When I try to sign in with the non-third party it never signs in, just refreshes the page -- I don't usually get credential errors when attempting to sign in either.
I have the base_url set up with our subdomain that we set up in our DNS. Could this be causing the issue? Should it be the default in the UI & then set up in the .env file so that it can be changed.. maybe?
I don't think that having a backup user account for restore specifically is the answer.. but maybe it is?
I don't know if anyone else has had this issue. I tried looking but it's such a niche issue I think. Any help would be appreciated!
4 Replies
what does the login request respond? Are you on same domain as before?
The last time I tried to restore the server, it was not giving me an 'invalid credential' error, but this time it was. I'm just trying to test a restore without having to take the windmill server that is currently running down or change the DNS record to point to the new IP. If this isn't feasible.. I understand, I'm just trying to create a recovery plan & test recovery.
I mean it should just work, there are no reasons why the credentials would be affected
the only reason possibly if if you're restoring at another domain, then you might have a conflict of cookie on your browser
Okay, so I was able to get it to work finally by creating my own cert & having caddy manage it. I guess it's because it wasn't using https -- which is the only difference. Of course it was a different domain because I was trying to just use my server's IP since it's just a temporary restore to confirm that it could be done .. for disaster reasons. Anywho.. thank you!