Previous incidents

A peek load of API calls was received and handled, although the auto-scaling took a few seconds to kick in. In those seconds the health check was performed and received an error.

The internal health checks of the service discovered an error and restarted the service as a precaution. The underlaying problem is identified and a solution is being worked on.

The service was released which caused a few seconds of downtime. The new zero-downtime deployments are being introduced which will solve this issue.