Strange dmesg and syslog messages

[lpallard]

Hello,

PVE 3.2 is installed on our local server running dual socket Opteron 4334 on a Supermicro H8DCL-iF motherboard.

VM #101 running CentOS 6.5 x86_64 displays strange messages right after GUI has loaded (or early after loading up the basic stuff)

bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dffff000-e0000000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dffff000-e0000000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dffff000-e0000000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back
bmc-watchdog:2863 map pfn expected mapping type uncached-minus for dfffd000-dfffe000, got write-back

VM hardware:
59GB/64GB assigned
CPU: kvm64 2 socket / 4 cores each = 8 cores
HDD1: local RAW format, 120GB VIRTIO
HDD2: software RAID5 passed to the VM via vm.conf / VIRTIO
NIC: 1X e1000

About as the same time the VM is booted up, proxmox's SYSLOG displays

Apr 3 22:08:11 proxmox kernel: kvm: 3866: cpu0 unhandled rdmsr: 0xc001100d
Apr 3 22:08:11 proxmox kernel: kvm: 3866: cpu0 unhandled rdmsr: 0xc0010112
Apr 3 22:08:12 proxmox kernel: kvm: 3866: cpu0 unhandled rdmsr: 0xc0010001
Apr 3 22:08:12 proxmox kernel: kvm: 3866: cpu1 unhandled rdmsr: 0xc001100d
Apr 3 22:08:12 proxmox kernel: kvm: 3866: cpu2 unhandled rdmsr: 0xc001100d
Apr 3 22:08:12 proxmox kernel: kvm: 3866: cpu3 unhandled rdmsr: 0xc001100d
Apr 3 22:08:12 proxmox kernel: kvm: 3866: cpu4 unhandled rdmsr: 0xc001100d
Apr 3 22:08:12 proxmox kernel: kvm: 3866: cpu5 unhandled rdmsr: 0xc001100d
Apr 3 22:08:12 proxmox kernel: kvm: 3866: cpu6 unhandled rdmsr: 0xc001100d
Apr 3 22:08:12 proxmox kernel: kvm: 3866: cpu7 unhandled rdmsr: 0xc001100d

I have confirmed the timing and YES, these errors are time related.

Anybody can provide insight on how to consider these occurrences?

Any help appreciated

Thanks for your support

 

[adamb]

The cpu messages are nothing to be concerned with. The other messages though I am unsure.

 

[lpallard]

According to my google research, the bmc-watchdog mesages were coming from the FreeIPMI packages, and being installed on a CentOS VM, I assume the package had a hard time communicating (or maybe it couldn't at all) with the IPMI platform....

Anyways, no need for IPMI management from CentOS VM. I uninstalled the FreeIPMI package and its dependencies, all is quiet now..

Anybody with more insight, please share..

Defective motherboard?!