Remind you VRM01 memory was rising to near 80%:
Portal has an alarm VM Memory Usage Exceeds the Threshold, alarm object is VRM01
Memory Usage Mechanism
Virtual Resources Management (FusionCompute) collected memory information from OS Kernal which included free, buffers and cached memory status, VRM OS selected cached+ and slab information stored in Cached from memory, if slab is in high usage then FusionCompute will reported in high memory usage.
As shown in the figure:
The picture shown above is descripting VRM is a component that manage physical resources from host, if high memory alarm occurs it wouldn't affect operation of guest VM on hosts.
Solution 1: The high memory usage alarm of VRM01 could be ignored.
Solution 2: Release the memory space in system
Login VRM server as root user, run the following command:
Echo 3> /proc/sys/vm/drop_caches
Wait for 10 minutes and logon to FusionManager Portal check if the alarm is cleaned automatically.
Solution 3: Version upgrade will be released in future.