[ovs-dev] OVS 2.6.1 release?

[Thadeu Lima de Souza Cascardo]

On Wed, Oct 19, 2016 at 06:08:12PM +0200, Justin Pettit wrote:
> 
> > On Oct 19, 2016, at 5:55 PM, Thadeu Lima de Souza Cascardo <cascardo at redhat.com> wrote:
> > 
> >> On Tue, Oct 18, 2016 at 05:35:27PM +0200, Justin Pettit wrote:
> >> 
> >>>> On Oct 18, 2016, at 5:31 PM, Russell Bryant <russell at ovn.org> wrote:
> >>>> 
> >>>>> On Tue, Oct 18, 2016 at 4:54 AM, Justin Pettit <jpettit at ovn.org> wrote:
> >>>>>> On Oct 17, 2016, at 5:54 PM, Ben Pfaff <blp at ovn.org> wrote:
> >>>>>> 
> >>>>>> On Mon, Oct 17, 2016 at 09:14:33AM -0400, Russell Bryant wrote:
> >>>>>> Can we do a 2.6.1 release in the near future?  The latest Python commit is
> >>>>>> needed for OpenStack.
> >>>>> 
> >>>>> I'd support that.
> >>>>> 
> >>>>> Justin usually makes the releases, and he's traveling on vacation this
> >>>>> week, so we may not hear from him for a while.
> >>>> 
> >>>> I landed a few hours ago and can now do a release. Do we want to give people a day or two to get their changes in or are we ready now?
> >>> 
> >>> A couple of days is fine with me.  I don't have anything else in mind, but others may.
> >> 
> >> If I don't hear any requests to hold off today, I'll kick off the process.
> >> 
> >> --Justin
> > 
> > I have a fix for a bug that causes 100% CPU use when no ovsdb-server is up and
> > there are link notifications.
> > 
> > It's pending submission because I am testing on FreeBSD and found an issue that
> > might or not be caused by my patch. And it would be nice to fix it anyway.
> 
> Sounds good. Just let me know in a day or two the status. 
> 
> --Justin

Hi, Justin.

Thanks a lot for your patience.

Sorry to be late about the status. I had some travel arranged and thought I
could do this while traveling but this turned out to be wrong.

So, it turns out my fix does not introduce any new problems. I have just posted
it, so consider if you want to wait to apply it before releasing.

The other bug will take much more time to fix and would be much more invasive.
So, maybe not worth waiting for. Basically, after removing an interface when
using the netdev datapath, it won't be closed and adding it back won't reopen
it, requiring a restart or database change, just like it was before ifnotifier
for the kernel datapath. I might work on a fix, but that might take a while.

Regards.
Cascardo.