On Jan 09, 2023 at 19:28 UTC, we identified a subset of newly provisioned Managed Databases that were created without expected DNS entries. The primary effect of this was that these clusters were not accessible via their public and private hostnames. We found that this issue would have started as early as 11:00 UTC 01-07-2023, and could have potentially affected new clusters from that time until approximately Jan 10, 2023 - 14:45 UTC, when we temporarily disabled provisioning of new clusters to mitigate the impact of this issue. After continuing to investigate we were able to implement a resolution on Jan 10, 2023 - 20:25 UTC, at which time we re-enabled cluster provisioning and continued to monitor before moving the incident to a resolved status.
To prevent repeat occurrences of this issue, we are actively working to improve our monitoring infrastructure to specifically address this scenario.