System Status

All affected hosts are back to normal.

Due to an upstream provider issue, IPv6 egress is unavailable on a subset of IAD hosts. Internet-facing IPv6 services are unaffected.

The affected hosts are back online

One of our upstream providers is performing emergency network maintenance on a subset of hosts in ORD. Some machines in ORD may be unreachable during the incident. A small subset of MPG clusters are running with reduced HA due to their replicas being on the affected hosts.

We have fixed the underlying issue caused failed Machine starts in ARN

We're currently investigating an elevated level of machine start failures. Workaround in the meantime would be to [replace affected machines ](https://docs.fly.io/apps/trouble-host-unavailable#apps-with-one-machine-and-no-volumes)with a new one: `fly machine clone <machineID>`

This maintenance is complete.

We are performing an emergency network maintenance in IAD region. Machines on a subset of hosts will be unreachable for a few minutes starting from the publication of this incident.

We've fixed affected regions and everything is back to normal.

We've rolled out fixes to the affected regions and are monitoring for any additional issues.

We have identified the issue and are working to resolve in the affected regions. Newly created static egress IPs may continue to see problems while we roll out fixes.

We are working to resolve an issue with outbound IPv4 connectivity failures affecting machines with in the ARN, JNB, and GRU regions.