We have increased memory capacity on the gateway servers and added additional monitoring around those servers to prevent this from occurring again.

We are continuing to monitor for any further issues.

Availability has been restored for the system.

We have identified the root cause as a memory issue on a gateway server. The standby server was automatically used and the problematic server recovered automatically.

We are investigating a permanent fix to this issue.

The system was briefly unavailable as a service crashed. We are investigating the root cause.