[openstack-dev] [kolla][rdo] libvirt 2.0 process is failed during launching qemu progress

Clark Boylan cboylan at sapwetik.org
Mon Jan 2 22:23:22 UTC 2017


On Sun, Jan 1, 2017, at 09:13 PM, Jeffrey Zhang wrote:
> Checked more failed gate, here is the more info:
> 
> 1. only rax-iad has such issue
> 2. the failed and successful gates have the same rpm installed.
> 
> So the only difference between failed gate and successful gate is the xen
> hypervisor.

I don't have any answers but my first thought was maybe it is related to
https://bugs.launchpad.net/nova/+bug/1643911; however, that affects
libvirt 1.3.1 on Ubuntu Xenial and there is at least one instance of
that occurring on a non Xen hypervisor so maybe they are not related.

Also noticed that you appear to have two redundant sets of system logs
in the Kolla centos job(s). See
http://logs.openstack.org/80/415880/1/check/gate-kolla-dsvm-deploy-centos-binary-centos-7-nv/d94c349/logs/system_log/
and
http://logs.openstack.org/80/415880/1/check/gate-kolla-dsvm-deploy-centos-binary-centos-7-nv/d94c349/logs/system_logs/.

As for debugging this you may be able to install libvirt's debugging
symbols
(http://debuginfo.centos.org/7/x86_64/libvirt-debuginfo-2.0.0-10.el7.x86_64.rpm),
capture a core dump, load into gdb and go from there.

Clark



More information about the OpenStack-dev mailing list