Since 2026-11-10 23:02 UTC, my dashboard’s Status page has shown a Service Interruption for one of my apps: “A host some of your apps are running on is experiencing network issues. Apps may experience network connectivity issues.” fly incidents hosts list shows the same issue, and its “last updated” time is still the start time.
What I’ve checked:
- The app has four machines in iad: one web, one scheduler and two workers. fly machine list --json shows host_status: ok for all of them, all started.
- The app has no volumes.
- I’ve deployed several times since the notice appeared, most recently 2026-11-11 21:16 UTC. Every deploy succeeded, and the app responds normally.
- About 20 minutes after the notice appeared, my Managed Postgres cluster completed its v1-to-v2 migration (cutover around 23:25 UTC). I don’t know whether that’s related.
I read the “host unavailable” troubleshooting doc, but its fix (clone the affected machine and destroy the original) seems to assume a machine with a bad host status, and none of mine has one.
Questions:
- Is the notice stale, for example from a machine or a Postgres node that’s since moved, or is something of mine still on the affected host?
- If something is still on that host, how do I find out which machine or resource it is, given that every machine reports host_status: ok?
- If the notice is stale, is there a way to clear it, or will it time out on its own?
Thanks!