UPDATE 02:15 : it looks like the power supply is bad. The drives were moved to a comparable chassis and all the instances are back online.

-

We do not have any error messages. Looking into it.