Postgres was completely innaccessible for over an hour today. a restart did not fix it. It started working again after a second restart but we lost a lot of transactions because of that. If you guys could look into the cause of this that would be great.

do you see any error? can it be a space problem?

It would help if you could provide deployment logs.

no it had no specific errors and plenty of space (~ 1gb used out of 50). its an infrastructure problem with railway. We'll probably have to move off postgres on railway to a more reliable solution honestly. this is not the first time they've just been inexplicably down. Idk why this request created a bounty or whatever. the only people who would be able to see and fix whatever caused this are internal to railway.

As requested above, it would greatly help if you could provide the deployment logs.

i dont have the deployment logs. the last postgres deployment was > 30 days ago so they are no longer stored. This just randomly stopped working. after working for quite a while

I have similar issue, tried Restart/Redeploy multiple times, still cant connect. Then I proceeded with Remove and Deploy, then it works again.