Intermittent errors and unavailability on the Platform API
Summary: The Platform API was intermittently unavailable, with several brief windows of full unavailability before service was fully restored.
Root cause: A recent deployment was misconfigured and the affected service ended up running with fewer healthy instances than required. When the remaining healthy instance was replaced, the replacement was unable to start, causing requests to fail.
Resolution: The misconfigured deployment was corrected and healthy instances were restored, returning the service to its expected state.
Customer impact: Intermittent request failures and short windows of full unavailability on the Platform API.
Follow-ups: Investigate why steady-state availability had been running below the configured replica minimum; add validation to prevent unpullable image references reaching production.
Summary: The Platform API was intermittently unavailable, with several brief windows of full unavailability before service was fully restored.
Root cause: A recent deployment was misconfigured and the affected service ended up running with fewer healthy instances than required. When the remaining healthy instance was replaced, the replacement was unable to start, causing requests to fail.
Resolution: The misconfigured deployment was corrected and healthy instances were restored, returning the service to its expected state.
Customer impact: Intermittent request failures and short windows of full unavailability on the Platform API.
Follow-ups: Investigate why steady-state availability had been running below the configured replica minimum; add validation to prevent unpullable image references reaching production.
This incident has been resolved. The Platform API has returned to its expected steady state and availability is holding.
A fix has been deployed and the service is recovering. We are monitoring to confirm it remains stable.
We have identified the cause — a recent deployment left the service running with fewer healthy instances than required. We are rolling out a fix.
We are investigating intermittent errors and unavailability affecting the Platform API. Some requests are failing while others succeed.