I have my vRealize Automation appliance 7.6 + HF 19 with 8vCPU & 128 GB of Memory configured. The Database Size is around 76 GB. Please find the attachment.
We have almost 10K PayLaod virtual machines so far provisioned by the vRA and many got destroy as well on a regular daily basis.
We have seen, a pattern of Memory utilization consistently on the higher side all the time and it is not coming down like the way CPU utilization. You can see the Graph I have attached.
Once the appliance reaches 95 % of utilization the vRA Deployment stopped getting provisioned and we only have to reboot the appliance to resolve the issue. After reboot, everything work in the normal way.
We have increased the memory from 64 GB to 96 GB & then increase it to 128 GB. Where we saw, when the appliance was configured with 64 GB then it reaches 95 % utilization in only 5 days. When it was configured with 96 GB then to reach 95 % utilization it took 14 days.
Then we again increase both CPU and Memory resources to 128 GB and 8 CPU ( 8 Core) on a single socket than to reach 95 % utilization the vRA appliance took 4 weeks of time.
And once it reaches that much of peak of 95 % of memory then we only have an option to reboot the appliance again and need to wait for 4 weeks again to get fail. I don't understand this pattern of utilization for Memory resources and why it can not release back the memory back to the free pool.
At present, it has already reached 89.8 % as per the screenshot.
Is anyone has to face such a similar issue, how to reclaim the used memory or how to clear the memory cache of vRA 7.6 Appliance using a command-line option? Can anyone suggest, please?
Just an FYI, the Heap Dump size for Java did not go beyond 4 GB compared with the configured size of 6 GB at the appliance level.
Any help or pointers in regard to that will be appreciated.
