Investigate failing uptime checks for staging repo and server

[manuelwedler]

GCP sent alerts for staging repo quite frequently in the last weeks. I've also seen staging server failing, but less frequently than repo. We should check what is wrong.

I think this is happening since we switched from the SSD to the HDD google-managed NFS. One idea I have is that it might be related to the sourcify-staging-repository-archive-gen job. As far as I checked, the uptime alerts happened at the same time as this job was running. This job is using the NFS heavily, so might be worth to investigate this thought a bit more.