We have started noticing memory issues on one stick of memory in our cloud cluster master node, we will perform emergency maintenance to replace the faulty memory stick.
All vms have been live migrated to other nodes during the maintenance, all cloud API requests will be failing while maintenance is occuring.