Cluster hangs at migration

[Tim vh]

Hi,

I have currently 2 separate Proxmox clusters running with OmniOS napp-it NFS SAN's.
When I migrate a VM, online or offline after a while the cluster gets unresponsive.
The VM that i being migrating hangs.
At that moment all my vm's are starting to hang.
When I try to access my SAN via NFS at that moment it's very slow.
When I stop the VM that's being migrated everything is running fine again.
As soon I start the VM everything hangs again.
When I migrate the VM back to the original node there are no problems anymore.

Does anyone got an idea how to solve this?

 

[Tim vh]

Log on node:

Dec 2 09:43:41 node34 kernel: [14639926.997391] kvm [24172]: vcpu0 unhandled rdmsr: 0xce
Dec 2 09:43:41 node34 kernel: [14639927.130637] kvm [24172]: vcpu0 unhandled rdmsr: 0x345
Dec 2 09:43:41 node34 kernel: [14639927.130709] kvm_set_msr_common: 246 callbacks suppressed
Dec 2 09:43:41 node34 kernel: [14639927.130711] kvm [24172]: vcpu0 unhandled wrmsr: 0x680 data 0
Dec 2 09:43:41 node34 kernel: [14639927.130740] kvm [24172]: vcpu0 unhandled wrmsr: 0x6c0 data 0
Dec 2 09:43:41 node34 kernel: [14639927.130768] kvm [24172]: vcpu0 unhandled wrmsr: 0x681 data 0
Dec 2 09:43:41 node34 kernel: [14639927.130796] kvm [24172]: vcpu0 unhandled wrmsr: 0x6c1 data 0
Dec 2 09:43:41 node34 kernel: [14639927.130824] kvm [24172]: vcpu0 unhandled wrmsr: 0x682 data 0
Dec 2 09:43:41 node34 kernel: [14639927.130851] kvm [24172]: vcpu0 unhandled wrmsr: 0x6c2 data 0


Dec 2 10:16:51 node34 pvestatd[2243]: command '/usr/bin/ssh -o 'BatchMode=yes' -i /etc/pve/priv/zfs/192.168.30.10_id_rsa root@192.168.30.10 zfs get -o value -Hp available,used SSDpoolTwo' failed: exit code 255


Dec 2 09:35:51 node34 kernel: [14639456.835439] kvm: zapping shadow pages for mmio generation wraparound
Dec 2 09:35:51 node34 kernel: [14639456.835888] kvm: zapping shadow pages for mmio generation wraparound