Starting at 2025-01-15T01:02:00 UTC some Atlas Serverless instances in AWS Singapore region (ap-southeast-1) experienced downtime. The issue was due to unexpected memory exhaustion of the system that caused outage for a subset of the Atlas infrastructure in that region. A small number of serverless instances running on those infrastructure would have experienced downtime. Our engineers were alerted and the system was restored to health around 2025-01-15T01:44:00 UTC.
Microsoft Azure has updated their status post and indicated that impacted services are now recovered. Atlas cluster operations are no longer experiencing delays.