Proxmox 8.3 gets hung after few hours

S

ske357

New Member
Feb 12, 2025
5
0
1
I've installed Proxmox 8.3 on a barebones PC I have. After about 4-6 hours of just letting it run the PC would freeze or get hung and the PC is not responsive forcing me to reboot manually. I've had this PC for about 6 years or more running WIN10 with no freezing issues. I took the PC to my local shop to have hardware checked and everything came back working fine. They installed an older version of Proxmox VE (8.0.3) and the PC ran all weekend no issues. Once I got it back home and ran apt update && apt upgrade the system is now on 8.3.3 and the freeze happened again. This seems like an software issue but im not sure. Im not a noob at linux just not a guru. Any commands you would like for me to run just let me know. Any help would be appreciated.

AMD Ryzen 7 1700 8-cores
32GB RAM
2TB SSD (proxmox installed here)
6TB WD HDD
 
Hi @ske357

this might be problem o incompatibility of Kernel with your HW, might will help to revert to different version

I am afraid that you will not avoid console:

apt search pve-kernel to get info of available kernels

pet get install pve-kernel-6.2.16-2-pve you cen install different kernel, example



And this from our Proxmox training, hope is fine to post it here :)

Proxmox-Boot-Tool

Use the proxmox-boot-tool to pin a certain kernel version.

proxmox-boot-tool kernel list will show available kernel versions.

If an older kernel is installed, but not shown in the list, you can manually add it with
proxmox-boot-tool kernel add <version>

To pin a kernel version run:
proxmox-boot-tool kernel pin <version>
With the additional --next-boot flag you can choose to boot a specific kernel on the
next boot only.

Before you start, what is on monitor, are you able to login on console locally or SSH ?
Before start experimenting, backup all your data
 
Lukas Moravek said:
Hi @ske357

this might be problem o incompatibility of Kernel with your HW, might will help to revert to different version

I am afraid that you will not avoid console:

apt search pve-kernel to get info of available kernels

pet get install pve-kernel-6.2.16-2-pve you cen install different kernel, example



And this from our Proxmox training, hope is fine to post it here :)

Proxmox-Boot-Tool

Use the proxmox-boot-tool to pin a certain kernel version.

proxmox-boot-tool kernel list will show available kernel versions.

If an older kernel is installed, but not shown in the list, you can manually add it with
proxmox-boot-tool kernel add <version>

To pin a kernel version run:
proxmox-boot-tool kernel pin <version>
With the additional --next-boot flag you can choose to boot a specific kernel on the
next boot only.

Before you start, what is on monitor, are you able to login on console locally or SSH ?
Before start experimenting, backup all your data
Click to expand...
Once I reboot i can log back into it via console or ssh... there is no data yet and anything can be deleted if necessary.
 
ske357 said:
Once I reboot i can log back into it via console or ssh... there is no data yet and anything can be deleted if necessary.
Click to expand...
When the machine freeze, is there any log on monitor screen, error message ?

also some hints

View logs from the last boot before the current one
journalctl -b -1

View logs from a specific time period before the freeze
journalctl --since "1 hour ago" --until "now"#

View kernel messages leading up to the freeze
journalctl -k -b -1

View the last 1000 lines before the current
bootjournalctl -b -1 -n 1000

View logs from the last boot before the current one and filter errors
journalctl -b -1 | grep -i "error|fail|fatal"

Export to the file, if you want to upload, so I can look
bootjournalctl -b -1 -n 1000 > freeze_logs.txt
 
Last edited:
  • Like
Reactions: BinaryCrash
Lukas Moravek said:
When the machine freeze, is there any log on monitor screen, error message ?

also some hints

View logs from the last boot before the current one
journalctl -b -1

View logs from a specific time period before the freeze
journalctl --since "1 hour ago" --until "now"#

View kernel messages leading up to the freeze
journalctl -k -b -1

View the last 1000 lines before the current
bootjournalctl -b -1 -n 1000

View logs from the last boot before the current one and filter errors
journalctl -b -1 | grep -i "error|fail|fatal"

Export to the file, if you want to upload, so I can look
bootjournalctl -b -1 -n 1000 > freeze_logs.txt
Click to expand...
I did the commands above and nothing stands out to me, here is the last 1000 lines.
 

Attachments

Lukas Moravek said:
No clue based on the log
- Bios firmware update ? 6 year old I do not belive
- Try disable Powermangement features in BIOS

When the machine freeze, is there any log on monitor screen, error message ?

If nothing helped, I suggest to downgrade version of Kernel.
Click to expand...
I flashed a BIOS update last week and still froze. I just disabled all the power management features in the BIOS, ill let it run for a few hours to see if that makes a difference. When the machine freezes there are no visible errors, the way i can tell is my cursor on the console is not longer visible/ blinking and then i ping with no response. You are probably right about 6 years... more like 7-8. PC was bought after I moved into my current hm in '16.

Update: Power management may have been the issue, been running for about 7.5 hours on the bare install of 8.3. Going to update and upgrade and let it run over night to see if it freezes.

Update2: After update and upgrading, the system ran about 13 min before freezing. This morning i downgraded the kernel as you suggested and voila, the PC has been running for almost 9 hrs.
 
Last edited:
@Lukas Moravek Thank you for your assistance. I believe I now have a working lab environment.

Things I did:
disabled all power management in BIOS
downgraded kernel to 6.8.12-4 (previously 6.8.12-8)
 
You must log in or register to reply here.
Top Bottom