Hello all.
I'm running a new install of PVE (8.2.4) and am experiencing some weird reboots.
Reboots are not graceful, and there's no logging as to what went wrong. I just get "-- Reboot --". Sometimes it will last hours, some times it will last minutes.
Memtest has not been able to complete. But no explicit errors have been seen on the console. Sometimes it runs for 15 minutes, sometimes it can run for hours. I just have not seen it complete. Memtest passed.
I ran prime95 for 4 hours without issues. I'm not sure if it would've completed, I stopped it early.
I turned off auto-start on all VMs and LXCs and rebooted, and it ran fine all night (12+ hours). I started up a VM (truenas) and it crashed within 15 minutes. I let it come back up and waited an hour without reboots, and started an LXC (heimdall) and it crashed within 30 minutes.
I'm running journalctl -f and nothing gets logged out before the reboot.
My hardware:
I read through this post: https://forum.proxmox.com/threads/proxmox-mystery-random-reboots.125001/ and changed my /etc/default/grum to
Attached are some system logs from the GUI, some journalctl -f logs, and then the pveversion -v.
What I haven't had the time to try yet:
Test individual sticks of RAM, in different slots (mobo is used).
Test PSU (it's a few years old, but nothing has been concerning up to this point).
Hooking it up to a monitor to see if the video out shows anything.
Here you can see the moments where it reboots:
I'm running a new install of PVE (8.2.4) and am experiencing some weird reboots.
Reboots are not graceful, and there's no logging as to what went wrong. I just get "-- Reboot --". Sometimes it will last hours, some times it will last minutes.
I ran prime95 for 4 hours without issues. I'm not sure if it would've completed, I stopped it early.
I turned off auto-start on all VMs and LXCs and rebooted, and it ran fine all night (12+ hours). I started up a VM (truenas) and it crashed within 15 minutes. I let it come back up and waited an hour without reboots, and started an LXC (heimdall) and it crashed within 30 minutes.
I'm running journalctl -f and nothing gets logged out before the reboot.
My hardware:
Intel i5 12400 - *NEW*
ASRock Z690m PG RIPTIDE mATX - *USED*
Crucial CT2K32G48C40U5 2x32GB DDR5-4800 - *NEW*
1x Seagate IronWolf 6TB 7200rpm - *NEW*
1x Kingston NV2 500GB M.2 - *NEW*
LEPA G1200 PSU - *USED*
Fractal Node 804 Case - *USED*
I read through this post: https://forum.proxmox.com/threads/proxmox-mystery-random-reboots.125001/ and changed my /etc/default/grum to
GRUB_CMDLINE_LINUX_DEFAULT="quiet pci=assign-busses apicmaintimer idle=poll reboot=cold,hard"
That did not fix it.Attached are some system logs from the GUI, some journalctl -f logs, and then the pveversion -v.
What I haven't had the time to try yet:
Test individual sticks of RAM, in different slots (mobo is used).
Test PSU (it's a few years old, but nothing has been concerning up to this point).
Hooking it up to a monitor to see if the video out shows anything.
Here you can see the moments where it reboots:
Attachments
Last edited: