Created attachment 1721222 [details]
Picture of Dashboard
Description of problem:
Ovirt server is showing using almost all available physical memory even though running vms which account for less than half of used memory. Running 6 vms which should be around 128 Gigs of physical memory, showing 220 GBs of used memory after five days.
Version-Release number of selected component (if applicable):
CentOS Linux release 8.2.2004 (Core)
After restart, server memory usage returns to normal after about three to five days after restart of ovirt server, memory usage spikes to almost all of physical memory. I have been running this new version for about 6 weeks and has been consistent each time. I do maintenance every two weeks, with updates to vms and Ovirt.
Steps to Reproduce:
1. Restart Ovirt-Engine
2. Memory usage spikes after two days.
Created attachment 1721223 [details]
Tar of /var/log/ovirt-engine/
Created attachment 1721224 [details]
Tar of /var/log/vdsm/
Created attachment 1721225 [details]
Picture of free command.
The documentation text flag should only be set after 'doc text' field is provided. Please provide the documentation text and set the flag to '?' again.
Could be related to memory leak on the engine (bz 1891332)
Could you tell us if it's an Hosted Engine environment (from the logs I figure it doesn't, but still - to be sure).
Also - The free output is from the host? The memory leak sounds in beast.rts-llc.com, which is the host, right?
If so, does restarting the ovirt-engine service "refresh" it, dropping the memory usage? It is unlikely since they are on different machines.
Do you mean restarting something on the host?
We suspect this might be fixed by bz 1891332 but we need to get more information about the environment, specifically whether it's an hosted engine environment.
Feel free to reopen if more data can be provided