Hi
I configure two VMS one with 8GB RAM and one with 4GB
But obviously they taken much more memory. This could not be all for management. And this nearly killed my System with an OOM. So I think this is an real issue with 13.3…
It is much better than in 13.0 but still worse
Please read the two bottom links in my signature to know how to ask a question if you are looking for help.
From what very little you posted, you are using 108MB of SWAP space. This tells me you are out of physical memory for TrueNAS to run on. SWAP utilization should be zero or possibly a few KB for a properly configured system. SWAP is for emergency use in my opinion, not for routine use.
Please explain.
He’s refering to this prevoius post, but without know hardware specs at this point Is hard.
At least, why not start tryng with a minimal VM? Like 1/1/1 CPU and 1gb RAM
I’m running Truenas Core 13.3 (TrueNAS-13.3-RELEASE) at
AMD Ryzen 5 PRO 4650G - 32GB RAM
Upgraded from 13.0 U2
Boot pool is at ada0p2
I would expect that If I configured the VM with 8GB it will not raise the reserved or virtualized memory over this amount plus a bit for internal stuff (9GB would be ok from my point of view.) But in this case as soon as I start memory intensive actions (large tgz/zip file extraction or file loading) the amount of reserved and virtual memory where increased immediately of this amount and brings the system near to OOM for ZFS Cache and Stable operation of middlewared. Even the external IO is reduced and the system starts to use the swap which is not expected from a system with 32GB RAM without dedup and ~12GB RAM reserved for VMs.
TrueNas Core 13.0 byhve had an nearly issue but there BOTH VM Memory were increased till the system (middlewared as first) crashed (12 GB SWAP fully used as well) if one VM starts to eat memory a lot.
If the vm internal task is eating memory it should use his own swap not the systems and should never ever instabilize the Storage system.
I would be happy to give more details and help to solve this but dont know how.
How can I tag this with Tag Core or better Tag byhve ?
Sorry for the delay but i didn’t receive the alert of your response
This Is probably what everyone expect 
on that u totally right.
And not seems you are allocating too much RAM on VM…
Have you still try with a minimal VM, if this continue happen? If both yes, i suggest to open a bug report on jyra, hope they can help you
