The virtual machine was unexpectedly shut down

[tangshuai]

Jan 11 12:48:00 node19 systemd[1]: Starting Proxmox VE replication runner...
Jan 11 12:48:01 node19 systemd[1]: pvesr.service: Succeeded.
Jan 11 12:48:01 node19 systemd[1]: Started Proxmox VE replication runner.
Jan 11 12:48:05 node19 pmxcfs[1775]: [status] notice: received log
Jan 11 12:48:09 node19 kernel: mce: [Hardware Error]: Machine check events logged
Jan 11 12:48:09 node19 kernel: EDAC skx MC2: HANDLING MCE MEMORY ERROR
Jan 11 12:48:09 node19 kernel: EDAC skx MC2: CPU 12: Machine Check Event: 0x0 Bank 7: 0xdc00010001010090
Jan 11 12:48:09 node19 kernel: EDAC skx MC2: TSC 0x0
Jan 11 12:48:09 node19 kernel: EDAC skx MC2: ADDR 0x37f83024c0
Jan 11 12:48:09 node19 kernel: EDAC skx MC2: MISC 0x200001c083001086
Jan 11 12:48:09 node19 kernel: EDAC skx MC2: PROCESSOR 0:0x50654 TIME 1704948489 SOCKET 1 APIC 0x20
Jan 11 12:48:09 node19 kernel: EDAC MC2: 4 CE memory read error on CPU_SrcID#1_MC#0_Chan#0_DIMM#0 (channel:0 slot:0 page:0x37f8302 offset:0x4c0 grain:32 syndrome:0x0 - OVERFLOW err_code:0x0101:0x0090 socket:1 imc:0 rank:0 bg:3 ba:3 row:0x54a2 col:0x3b8)
Jan 11 12:48:12 node19 pmxcfs[1775]: [status] notice: received log
Jan 11 12:48:16 node19 kernel: MCE: Killing kvm:47816 due to hardware memory corruption fault at 56461151ad48
Jan 11 12:48:16 node19 kernel: fwbr3358i0: port 2(tap3358i0) entered disabled state
Jan 11 12:48:16 node19 kernel: fwbr3358i0: port 2(tap3358i0) entered disabled state
Jan 11 12:48:16 node19 pveproxy[18818]: worker exit
Jan 11 12:48:16 node19 systemd[1]: 3358.scope: Succeeded.
Jan 11 12:48:16 node19 pveproxy[1920]: worker 18818 finished
Jan 11 12:48:16 node19 pveproxy[1920]: starting 1 worker(s)
Jan 11 12:48:16 node19 pveproxy[1920]: worker 21319 started
Jan 11 12:48:16 node19 pvestatd[1891]: VM 3358 qmp command failed - VM 3358 not running
Jan 11 12:48:16 node19 kernel: mce: [Hardware Error]: Machine check events logged
Jan 11 12:48:16 node19 kernel: EDAC skx MC2: HANDLING MCE MEMORY ERROR
Jan 11 12:48:16 node19 kernel: EDAC skx MC2: CPU 12: Machine Check Event: 0x0 Bank 7: 0xdc0000c001010090
Jan 11 12:48:16 node19 kernel: EDAC skx MC2: TSC 0x0
Jan 11 12:48:16 node19 kernel: EDAC skx MC2: ADDR 0x37f82ffac0
Jan 11 12:48:16 node19 kernel: EDAC skx MC2: MISC 0x200002c127c01086
Jan 11 12:48:16 node19 kernel: EDAC skx MC2: PROCESSOR 0:0x50654 TIME 1704948496 SOCKET 1 APIC 0x20
Jan 11 12:48:16 node19 kernel: EDAC MC2: 3 CE memory read error on CPU_SrcID#1_MC#0_Chan#0_DIMM#0 (channel:0 slot:0 page:0x37f82ff offset:0xac0 grain:32 syndrome:0x0 - OVERFLOW err_code:0x0101:0x0090 socket:1 imc:0 rank:0 bg:3 ba:3 row:0x54a2 col:0x348)
Jan 11 12:48:17 node19 qmeventd[1490]: Starting cleanup for 3358
Jan 11 12:48:17 node19 kernel: fwbr3358i0: port 1(fwln3358i0) entered disabled state
Jan 11 12:48:17 node19 kernel: vmbr0v15: port 73(fwpr3358p0) entered disabled state
Jan 11 12:48:17 node19 kernel: device fwln3358i0 left promiscuous mode
Jan 11 12:48:17 node19 kernel: fwbr3358i0: port 1(fwln3358i0) entered disabled state
Jan 11 12:48:17 node19 kernel: device fwpr3358p0 left promiscuous mode
Jan 11 12:48:17 node19 kernel: vmbr0v15: port 73(fwpr3358p0) entered disabled state
Jan 11 12:48:17 node19 qmeventd[1490]: Finished cleanup for 3358

 

[tangshuai]

I don't know why my virtual machine was unexpectedly shut down. The above is the log from the system, but there are multiple virtual machines under this host, and only this one was shut down. The hardware management interface of the host did not detect any hardware fault alarms.

 

[sb-jw]

Your RAM seems to be defective, so you should check it with memtest.

 

[tangshuai]

Your RAM seems to be defective, so you should check it with memtest.

But why is it that only a single virtual machine goes down and the other dozens are fine? What I'm curious about is this point

 

[tangshuai]

May I ask if the ProxmoxVirtual Environment 6.4-4 system comes with its own EDAC module,

 

[sb-jw]

But why is it that only a single virtual machine goes down and the other dozens are fine?

After the reboot it may be a different VM. The defective area is and remains defective and whatever wants to use this area is somehow affected by it. So again, do a Memtest and see if and what errors you have. Maybe re-seating the CPU or DIMM will help.