Virtual disks with SCSI or virtio hang after some time

Virtual disks with SCSI or virtio hang after some time

[Guido Winkelmann]

Hi,

When using KVM machines with virtual disks that are hooked up to the guest via 
either virtio or SCSI, the virtual disk will often hang completely after a 
short time of operation. When this happens, the network connectivity of the 
machine will usually go down, too. (I.e. it stops responding to pings.)

When this happens when using SCSI, the following messages will appear in 
syslog (in the guest):

sd 0:0:0:0: [sda] ABORT operation started
sd 0:0:0:0: ABORT operation timed out
sd 0:0:0:0: [sda] DEVICE RESET operation started

When using virtio, no log messages appear at the time the hang occurs, but 
some time before, these messages appear:

JBD: barrier-based sync failed on vda1-8
JBD: barrier-based sync failed on vda3-8
JBD: barrier-based sync failed on vda5-8

The hang seems to happen faster with virtio, usually within one minute after 
bootup, while with SCSI, it can take a few more minutes

In the host, no interesting messages appear in syslog.

The host OS is Fedora Core 12, with qemu-kvm 0.11.0, Kernel 
2.6.32.11-99.fc12.x86_64 and libvirt 0.8.1.

The guest kernel is 2.6.32-gentoo-r7.

Other, non-linux, guest operating systems seem to have problems with the 
virtualised scsi disks, too. (I've tried OpenBSD.)

Does anyone have an idea what is happening here and what can be done about it?

Regards,

	Guido

-- 
Too much multitasking isn't good for you.

Re: Virtual disks with SCSI or virtio hang after some time

[Avi Kivity]

On 06/01/2010 06:59 PM, Guido Winkelmann wrote:
> The host OS is Fedora Core 12, with qemu-kvm 0.11.0
>    

Please try with at least qemu-kvm-0.11.1, preferably qemu-kvm-0.12.4.  
Also use Linux 2.6.32.latest in the guest.

-- 
error compiling committee.c: too many arguments to function

Re: Virtual disks with SCSI or virtio hang after some time

[Guido Winkelmann]

Am Dienstag, 1. Juni 2010 schrieben Sie:
> On 06/01/2010 06:59 PM, Guido Winkelmann wrote:
> > The host OS is Fedora Core 12, with qemu-kvm 0.11.0
> 
> Please try with at least qemu-kvm-0.11.1, preferably qemu-kvm-0.12.4.
> Also use Linux 2.6.32.latest in the guest.

Okay, I've upgraded the userspace tools to qemu-kvm-0.12.4, and so far, the 
problem is not resurfacing. Before, it could usually be triggered by doing 
hdparm -t /dev/vda (or sda, respectively) once or twice.

Unfortunately, now I cannot start new VMs with libvirt anymore, and I have no 
idea why... :(

	Guido