Degraded performance
Slow instance provisioning in US East (Virginia)
Launches took about 9 minutes in US East (Virginia) for 57 minutes.
- Status
- ResolvedResolved — last update 31 Jul 2025, 15:01 UTC
- Started
- 31 Jul, 14:04Thursday 31 July 2025 UTC
- Duration
- 57 minEnded 31 Jul 2025, 15:01 UTC
- Components
- 1Instance launch and lifecycle — US East (Virginia)
Timeline
What we posted, as we posted it.
Resolved5 updates · UTC
- ResolvedThis incident has been resolved. Boot time is never billed.31 Jul, 15:01
- MonitoringLaunch times have been back under a minute since 14:49 UTC. We are monitoring.31 Jul, 14:49
- IdentifiedA batch of nodes was reimaged overnight and came back with an empty image cache, so every launch on them pulls the template from the registry.31 Jul, 14:23
- UpdateWe have narrowed the cause down and are applying a fix. Next update within 30 minutes.31 Jul, 14:15
- InvestigatingInstances in US East (Virginia) are taking longer than usual to become reachable: median launch time is about 9 minutes instead of under one.31 Jul, 14:04
Affected components
Details
- Type
- Degraded performance
- Started
- 31 Jul 2025, 14:04 UTC
- Ended
- 31 Jul 2025, 15:01 UTC
- Region
- US East (Virginia)
- Impact
- Launches took about 9 minutes in US East (Virginia) for 57 minutes.
- Reference
- 6b934f6549ff