[Gluster-devel] [Gluster-Maintainers] More news on 3.7.11

[Kaushal M]

On Fri, Apr 15, 2016 at 2:36 PM, Niels de Vos <ndevos at redhat.com> wrote:
> On Fri, Apr 15, 2016 at 01:32:23PM +0530, Kaushal M wrote:
>> Some more (bad) news on the status of 3.7.11.
>>
>> I've been doing some more tests with release-3.7, and found that the
>> fix for solving daemons failing to start when management encryption is
>> enabled doesn't work in all cases.
>>
>> Now I've got 2 options I can take, and would like some opinions on
>> which I should take.
>>
>> 1. Delay the release a little more, and fix the issue completely. I
>> don't know how long a proper fix is going to take.
>>
>> Or,
>> 2. Revert the IPv6 patch that exposed this problem, and release
>> immediately. We can then work on getting the issue fixed on master,
>> and then backport the IPv6 change again.
>>
>> What do other maintainers feel? Hopefully I get some opinions before
>> the weekend.
>
> I'm all for reverting the patch and release 3.7.11 as soon as possible.
>
> It is also not clear to me how much the IPv6 change was a fix, or a
> feature enhancement. Anything that gets backported to a stable branch
> should be done with extremely high confidence that nothing can break.
> Any patch, even the really simple ones, can have unexpected
> side-effects. A patch that gets backported and breaks the release (or
> the release schedule) might not have been suitable for inclusion in the
> stable branch in the first place.

One of the reasons we didn't catch the failures earlier is because our
regression test VMs have IPv6 disabled IIRC. I don't know why this was
disabled, but it should be enabled again.

>
> Thanks,
> Niels