During the deployment to a new cluster, I witnessed that requests to analyzers just hang forever.

We need to add timeouts for that and maybe to some other places (like communication with GH api, downloading repositories) if we don't have timeouts there yet.