Hi ,
Thanks for reaching out to Microsoft Q&A.
This is most likely a platform-side telemetry issue rather than anything wrong with your server. When CPU, memory, and storage metrics disappear but connection and availability metrics remain, it usually means the infrastructure metrics pipeline in Microsoft Azure is lagging or partially failing for that region or cluster. Restarts do not help because the problem is upstream of the db VM.
In practice, the fix is to open a support ticket and reference the known incident reported in the thread you shared. Microsoft typically restores the metrics once backend telemetry sync is corrected. Until then, use query-level monitoring (pg_stat views, extensions, or custom alerts) if you need temporary visibility.
Please 'Upvote'(Thumbs-up) and 'Accept' as answer if the reply was helpful. This will be benefitting other community members who face the same issue.