On 15/01/10 09:18, Marc Haber wrote:
[my apologies for this re-post. I stupidly managed to hide the first
instance of this mail away in an unrelated patch-thread and am thus
afraid that the people who could have answered didn't see it]
Hi,
I have one test host running Debian unstable (kernel 2.6.32.3), and I
would like to virtualize on it using KVM and virsh. Debian unstable
has libvirt 0.7.5.
On this host, I cannot start any KVM domain using console with these
XML parts:
<serial type='pty'>
<target port='0'/>
</serial>
<console type='pty'>
<target port='0'/>
</console>
When I try virsh start $domain, I get the error message:
error: internal error no assigned pty for device serial0
When I remove the serial and console stanza, the machine starts up.
This is the configuration documented everywhere. At first, I suspected
a change in the configuration syntax and filed Debian bug #565145, But
on the #virt IRC channel, people suggested that there was a change in
libvirt parsing KVM's output, which may be a genuine bug breaking the
console.
Is this already a known issue? Can I do anything to help debugging?
Hi, Marc,
Could you please post the full domain XML, and the QEMU command line it
generated? You can get the latter from
/var/log/libvirt/qemu/<domain>.log. Can you also post the versions of
qemu and libvirt?
This should let me see where it's going wrong.
Thanks,
Matt
--
Matthew Booth, RHCA, RHCSS
Red Hat Engineering, Virtualisation Team
M: +44 (0)7977 267231
GPG ID: D33C3490
GPG FPR: 3733 612D 2D05 5458 8A8A 1600 3441 EA19 D33C 3490