The Root cause of this incident for a small subset of stores was due to a bad/degraded AWS EC2 instance. Due to this degradation we have activated HA failover mechanism and balanced stores on new hardware nodes from our fleet.
For real time HA recovery additional High Availability Strategies are available: https://magemojo.com/kb/best-practices/high-availability-strategies/
Please feel free to contact support if you have any questions.