How to make qemu pause the VM when I/O hangs on disk?

Product Support Operations
1

Hello.

Using a network filesystem for datastores to store qcow2/raw disks, I wonder why qemu does not pause the VM when I/O hangs on the disk.

I remember quite a long time ago that a switch failure causes all the VM to pause, waiting for I/O to resume and it worked: when the network was OK again, the VM unpaused and continued to work.

I’ll try to setup a lab to reproduce but if someone has hints I’m interested.

NB: the disks are of the form

    <disk type='file' device='disk'>
      <driver name='qemu' type='qcow2' cache='none' discard='unmap'/>
      <source file='/var/lib/one//datastores/0/1674504/disk.0' index='3'/>
      <backingStore/>
      <target dev='sda' bus='scsi'/>
      <boot order='1'/>
      <alias name='scsi0-0-0-0'/>
      <address type='drive' controller='0' bus='0' target='0' unit='0'/>
    </disk>

Thanks.

2

Hello Daniel,

You should take a look at error_policy in libvirt’s domain XML specs. As disk definitions are handled by the OpenNebula core, there are no other options than post-processing the generated domain XMLs…

Best Regards,
Anton Todorov

3

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.