Why is my real memory > than allocated

I am running Centos with Open Nebula 5.8.5
In the last 6 months I see my real memory is greater than my allocated
When I check the I 2.2 T allocated and 2.7 used.
When I run top on the KVM nodes I see some, but not all of the nodes have ~150 G more used than allocated.

Where is this extra memory usage coming from? How can I reclaim the memory?

Could you share the output of onehost show, for one of the hypervisors with this problem?

Thanks!

Thanks Ruben,

Here’s the output. So where should I run this from?

[oneadmin@dppl-nebprod01 root]$onehost show
bash: onehost: command not found…

Hi

I’m assuming there is a kind of inconsistency between the allocated resources and those actually used in the OpenNebula host. The command should be run at the OpenNebula frontend for example.

Cheers

OK I got in now