There was a Fly release to our production DB this morning which has caused our DB to crash. While attempting to resolve the issue, our Dashboard on Fly started showing a status of 500. Any one else seeing these issues or have a way of solving them?
It looks like you have 2 volumes tied to your app. When a deploy is issued, Nomad will choose an available volume to allocate, which is mostly random if there are multiple, unallocated volumes. That being said, it’s likely that Nomad chose the volume that wasn’t holding the primary.
To resolve this, I would work to scale your app up to two and see if things become healthy. If you only want a single node, then I would work to remove the volume associated with the replica and scale back down to one after.
My DB also crashed this morning: see Postgres DB Dead and Can't Create a New Deployment
The 500s on the dashboards happened for me as well, but that looks to be resolved.