The partial outage for databases with primaries in the AMS region has been completely resolved.

Migration of primaries located in the AMS region was performed to avoid downtime due to a maintenance window of our cloud provider.

Due to the misconfiguration of clusters that had at least two locations during maintenance, replicas were not updated, which led to unavailability when requests arrived at replicas. If the request arrived at the primary, no issues were experienced.

As soon as we became aware of the issue, a fix was applied, and the clusters returned to normal operation.