Proxmox 9 + pfsense 2.8 high memory usage

F

fv92

New Member
Sep 19, 2025
2
0
1
Hi,

In a cluster, I have several nodes with different types of servers. I recently added a server with a Proxmox 9 installation, while I upgraded the others.
This last server has a 25 Gb Intel E810-XXV network card, while the old servers have a 10 Gb Intel X710.

My pfsense 2.8.1 VM runs correctly on the old nodes, with a memory usage of 1.2 Gb (30%) on the VM and 320 Mb (8%) on the pfsense.

If I migrate the VM to the new node, the memory increases to 95 to 99%, causing the host to crash after 5 to 6 days. I have to hard reboot the server because I can't control it. The KVM service is at 100% CPU.

I tried adjusting the VM's hardware settings to find a solution to this problem, but nothing. Perhaps a slight improvement with queues=1 on network cards with 75% memory, but it gradually increases, and even more so if I simply reboot the VM. I tried with or without Balloon, it's the same.

All the drivers are in Virtio. I focused my tests on the network part because that's where there's the biggest difference with the other nodes, but it could be due to other factors.

Other information: Before migrating to Proxmox 9, the VM had the E1000 network driver, but I had to switch to Virtio, otherwise the VM would no longer communicate. I also checked the pfsense prerequisites https://docs.netgate.com/pfsense/en/latest/recipes/virtualize-proxmox-ve.html.

Does anyone have any ideas on how to solve my problem ?
 
Per this thread the 810 driver has a memory leak:

E

Thread 'Slow memory leak in 6.8.12-13-pve'

Hi all,
I have recently added a host to a cluster, it has a newer kernel version than the others and I am noticing a slow and gradual memory leak which is consuming around an additional 500MB a minute. This continues until the host grinds to a halt and becomes un-operational. This host is identical in spec to the others in the cluster, of which they are not experiencing the issue. This is a Ceph cluster, with OSDs on each host. Looking in htop, no processes are using extreme amounts of memory which points to a kernel memory leak. I am using ZFS for boot disks. ARC looks good...
 
Thanks for the feedback.
I don't see any memory leaks on the host. I'm running version 6.14.11-1-pve.
I think my memory leak problem is with my pfsense VM (freebsd).
If I run the VM on a newer node with an E810 network card, the VM memory goes up to 98% on the host side (15% on the VM side).
On an older node (X710 network card), even if there's a significant difference between the two values (as discussed here), I don't have this saturation problem.
It's after a few days that my VM crashes and crashes my host.
A quick clarification: this VM is used to establish VPN connections for NFS mounts on the cluster. My node crash may be the result of VPN connection losses due to my VM crashing because it no longer accesses the NFS share.
Any other ideas ?
 
You must log in or register to reply here.
Top Bottom