sd 2:0:0:0: [sda] abort (Proxmox 3.4)

[cbx]

Hi

I have a few problem with an virtual, when backup run, the kvm failed automatically and I revieve error log :

[mar nov 3 04:16:07 2015] sd 2:0:0:0: [sda] abort
[mar nov 3 04:17:08 2015] INFO: rcu_sched detected stalls on CPUs/tasks: { 4} (detected by 7, t=60002 jiffies, g=33686286, c=33686285, q=0)
[mar nov 3 04:17:08 2015] sending NMI to all CPUs:
[mar nov 3 04:17:08 2015] NMI backtrace for cpu 7
[mar nov 3 04:17:08 2015] CPU: 7 PID: 0 Comm: swapper/7 ve: 0 Tainted: PF O-------------- 3.10.0-233.1.2.lve1.3.33.4.el7.x86_64 #1 ovz.4.1
[mar nov 3 04:17:08 2015] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.7.5-0-ge51488c-20140602_164612-nilsson.home.kraxel.org 04/01/2014
[mar nov 3 04:17:08 2015] task: ffff8801f55a4150 ti: ffff8801f55b4000 task.ti: ffff8801f55b4000
[mar nov 3 04:17:08 2015] RIP: 0010:[<ffffffff8104621a>] [<ffffffff8104621a>] native_write_msr_safe+0xa/0x10
[mar nov 3 04:17:08 2015] RSP: 0018:ffff8801ffdc3d80 EFLAGS: 00000046
[mar nov 3 04:17:08 2015] RAX: 0000000000000400 RBX: 0000000000000007 RCX: 0000000000000830
[mar nov 3 04:17:08 2015].....


the problem only appear in one of our KVM of our cluster and the virtual is not accesible during 5/10 min.

I have found this post that seem talk about our problem
http://unix.stackexchange.com/quest...ut-be-disabled-in-linux-attempting-task-abort

I have try it but for now with no better result.
Did someone have similar problem?

Thanks a lot

 

[tom]

is this log from the host or form the guest?

 

[cbx]

is this log from the host or form the guest?

Problem is always in guest logs.... I can still access to host but have an hight load, when load down, I can also access to guest.