[ovs-discuss] [OVN] OVN chassis processes terminate problem

[Hui Kang]

Han Zhou <zhouhan at gmail.com> wrote on 06/07/2016 08:17:38 PM:

> From: Han Zhou <zhouhan at gmail.com>
> To: Hui Kang/Watson/IBM at IBMUS
> Cc: Ben Pfaff <blp at ovn.org>, "discuss at openvswitch.org"
> <discuss at openvswitch.org>
> Date: 06/07/2016 08:17 PM
> Subject: Re: [ovs-discuss] [OVN] OVN chassis processes terminate problem
>
> On Fri, May 27, 2016 at 9:46 AM, Hui Kang <kangh at us.ibm.com> wrote:
> >
> > Ben Pfaff <blp at ovn.org> wrote on 05/27/2016 10:51:09 AM:
> >
> > > From: Ben Pfaff <blp at ovn.org>
> > > To: Hui Kang/Watson/IBM at IBMUS
> > > Cc: discuss at openvswitch.org
> > > Date: 05/27/2016 10:51 AM
> > > Subject: Re: [ovs-discuss] [OVN] OVN chassis processes terminate
problem
> > >
> > > On Thu, May 26, 2016 at 07:48:07AM +0000, Hui Kang wrote:
> > > > I have a testbed, including 100 hypervisors running ovn-controller,
> > > > ovsdb-server, and ovs-vswitchd. When VM instances are created
> and ports are
> > > > bounded to the ovsdb-server, the OVN chassis process on certain
> > > node terminates
> > > > unexpected after the hundreds of ports are created and bounded.
> > > Following are
> > > > the log messages after the processes terminate.
> > >
> > > Is there any evidence of a crash?  You might want to turn on core
dumps,
> > > if they are not already enabled.
> >
> > Thanks for the suggestions. I will enable the core dumps and check it
out.
> >
> > BTW, the crash does not happen when I used an earlier commit of OVS
code
> > (git commit id: 54529e94). So I doubt that some code committed after
that
> > date causes this scalability issue in OVN.
> >
> Is it still a problem? By chassis processes do you mean ovn-
> controller process?

Yes to both questions. Thanks.

- Hui


> Han
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://openvswitch.org/pipermail/ovs-discuss/attachments/20160608/193b7898/attachment-0002.html>