[SOLVED] Many unknown crashes/freezes

[Potatoes1921]

Hi all, I recently started up a homeserver with a bunch of replaced parts from my main Rig. I currently have Proxmox 7.3-6 installed on an m.2ssd, with another m.2 1tb for VM disks and backups. I have 3 4tb HDD passed through to my OMV VM that I am using as a NAS in RaidZ1 for files, backups, and media. Along with an inactive homeassistant VM and a plex server.

However, periodically with no rhyme or reason I can tell so far both the OMV and plex VMs would just be turned off and crashed, while (at the time I had the HA vm up) HA would still be running fine. I also noticed any time I tried to shutdown the OMV it would time out, and I would have to force Stop it to turn the VM off. On top of that, more recently Proxmox itself has been freezing to where I have to hold the power button on the machine to reboot it.

This lead me to believe it was most likely something to do with the ZFS system as it was the 2 vms that were creating it or using it. I suspect it also could be some of the hardware as its fairly old, a ryzen5 1600x, Asrock x370 killer, and I have an Nvidia geforce 1060 6gb installed but not in use. All the drives are brand new.

I currently have all VMs shutdown and just proxmox running. After 2days so far no crash, will update if it changes.

I have a fair bit of experience with proxmox from working in IT but this is my first home rig on non-server hardware. Any guidance or direction would be greatly appreciated.

 

[Chris]

HI,
have you checked the journal for errors?

Also check memory with memtest and smart status of disks...

 

[Potatoes1921]

HI,
have you checked the journal for errors?

Also check memory with memtest and smart status of disks...

I couldn't find much in journal but maybe I wasn't looking properly. I did notice some errors but it was hard to say what was actually an issue. The ones I did find I think also pointed to memory but its been a couple weeks of on and off testing now to remember properly..

Apparently its not though, memtest86+ passed 10 times with 0 failures. Very sad I wish it was as simple as that for once. Smart status is also all good, which they're all new drives so is expected

 

[Potatoes1921]

Update: Had just the OMV VM up to copy some files off of it to create a backup, and it crashed itself. Heres the .conf for it:

 

[Neobin]

What kernel version are you running (pveversion)?
In case it is the 6.1 opt-in, there were/are reports about problems with disk-passthrough on 6.1.6 and 6.1.10:
https://forum.proxmox.com/threads/o...r-proxmox-ve-7-x-available.119483/post-533639 and following.

 

[Potatoes1921]

What kernel version are you running (pveversion)?
In case it is the 6.1 opt-in, there were/are reports about problems with disk-passthrough on 6.1.6 and 6.1.10:
https://forum.proxmox.com/threads/o...r-proxmox-ve-7-x-available.119483/post-533639 and following.

pve-manager/7.3-6/723bb6ec (running kernel: 6.1.2-1-pve). Might still be worth trying a different kernel. Think i should try older since hardware is older, or newer?

 

[Neobin]

6.1.2 seems to be fine from the reports, but yeah, might make sense to try (a) different one(s) anyway.
6.1 is the currently most recent opt-in PVE-kernel.
Other options are the older (and afaik meanwhile not much (if at all?) updated) 5.19 opt-in kernel: [1] or simply the default 5.15 kernel.

[1] https://forum.proxmox.com/threads/opt-in-linux-5-19-kernel-for-proxmox-ve-7-x-available.115090

 

[Potatoes1921]

6.1.2 seems to be fine from the reports, but yeah, might make sense to try (a) different one(s) anyway.
6.1 is the currently most recent opt-in PVE-kernel.
Other options are the older (and afaik meanwhile not much (if at all?) updated) 5.19 opt-in kernel: [1] or simply the default 5.15 kernel.

[1] https://forum.proxmox.com/threads/opt-in-linux-5-19-kernel-for-proxmox-ve-7-x-available.115090

Okay I went all the way to the default 5.15.85-1. The UI is pretty painful but as long as it works that's all that matters to me, won't need to do much configuration anymore.

 

[Potatoes1921]

Guess it didn't work, both VMs still crashed with 5.15.85-1 as my kernel. Feel like it has to be with the OMV install, or the pass through drives since the VM also struggles to shut down on its own unless force stopped.

 

[Potatoes1921]

I tried removing the gpu and that didn't work either. Didn't think so but wasn't using it anyways. I might try and use a different NAS software instead of OMV since that seems to be where its originating from.

 

[Potatoes1921]

Okay, so I moved off of the OMV VM, recreated the ZFS pool and all. I had 1 crash within a day or so of doing it, but I've since been at about 5 days of uptime which is a new record. Part of the issue might be that I didn't give the OMV install enough ram, since I'm now at about 12/16gb of ram used when server is idle, when I only gave the VM 2gb, assuming that it needed more ram to run properly and that was the issues, I'll mark this as solved... as long as it stays stable.