[libvirt-users] domain has active block job

[Kashyap Chamarthy]

On Tue, Jan 13, 2015 at 08:49:53AM -0700, Eric Blake wrote:
> On 01/13/2015 07:21 AM, Kashyap Chamarthy wrote:
 
[. . .]

> > Seems like you're hitting an old bug[1] where 'blockcopy' (or
> > 'blockcommit') missed to execute a cleanup routine which destroys a
> > reference to the active block operation -- resulting in the error you're
> > seeing when you attempted to 'abort' the block operation manually.
> > 
> > This bug is fixed in libvirt-1.2.8 and above. I see you're using
> > libvirt-1.2.7, if you can update libvirt in your environment, that
> > should fix your issue.
> 
> Are you using a pre-built distro libvirt?  If so, which one?  We should
> figure out how to get that vendor to backport the right fix for this issue.
> 
> Also, I just now committed another related fix; so even the latest
> 1.2.11 release has an issue where libvirt can get into weird states if
> parallel block job attempts are made.  See commit e1125ce.
> 
> > [1] https://bugzilla.redhat.com/show_bug.cgi?id=1135169 -- blockcopy job
> > was cancel by "CTRL+C" while it show there still be one block job in
> > background
> 
> That was against RHEL 7; but I don't know if any Fedora releases suffer
> from the same issue.

Right, I looked for a Fedora bug before referring to it. Even the
libvirt master git history refers to this bug with the fixed commit:

  commit 8e23e0e977fbcc4a7880e187a63c509d6e6879c6
  Author: Erik Skultety <eskultet at redhat.com>
  Date:   Thu Nov 27 13:29:42 2014 +0100
  
      qemu: fix block{commit,copy} abort handling
      
      When a block{commit,copy} job was aborted on a domain, block job handler
      did not process it correctly, leaving a phantom job in the background.
      Any further calls to any blockjob causes "block <jobtype> still active"
      error. This patch fixes the blockjob handler so that it checks not only
      for VIR_DOMAIN_BLOCK_JOB_FAILED status, but VIR_DOMAIN_BLOCK_JOB_CANCELED
      status as well, followed by our existing cleanup routine.
      
      Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1135169
      
      Signed-off-by: Jiri Denemark <jdenemar at redhat.com>

-- 
/kashyap