[SOLVED] Bootloop after upgrading to PVE 8

Spaylia

Member
May 13, 2020
7
2
8
Hello,

I just upgraded from PVE 7 to 8 following the doc, the check list has no errors nor warnings.

However, when rebooting after the upgrade, I tried to login physically, I didn't have time to finish writing the password and it rebooted by itself.
Second boot I wrote the password and it worked for a few seconds and crashed again.
Third boot I thought maybe those auto start VMs are the problem, I'll try to edit the conf file so they don't start on boot, but it crashes before I can edit the files.
A few boot later I don't even reach the login screen, it crashes while printing the kernel messages during boot.

During those latest boot attempts, I got the messages:
Code:
[1.151016] mce: [Hardware Error]: CPU 10: Machine Check: 0 Bank 5: bea0000000000108
[1.151025] mce: [Hardware Error]: TSC 0 ADDR 1ffffa827498e MISC d012000200000000 SYND 44000000 IPID 5006000000000
[1.151032] mce: [Hardware Error]: PROCESSOR 2:870f10 TIME 1497266781 SOCKET O APIC 6 microcode 8701021

When looking for a solution online, most of them suggest a problem with CPU temps but my BIOS reports 45°C. The memtest passed with no error. However I don't think those errors are the cause, it's been running on this hardware for a long time and nothing happened today. Also I don't think I have seen those during the first few boots.

I'm running a Ryzen 5 3600 on an AsRock X470 Taichi, 64GB of Kingston ECC 2666 memory and I have an HBA, NIC and GT710 plugged in.
 
Last edited:
The GT710 is faulty apparently. I noticed some flickers when trying a Debian Live CD so ai figured it was worth trying removing it.
I don't have any video output now but at least it doesn't crash anymore!