Open Stack

On 3 July 2015 at 21:25, Dave Walker <email at daviey.com> wrote:
> On 3 July 2015 at 10:17, Robert Collins <robertc at robertcollins.net> wrote:
>> On 3 July 2015 at 21:01, Thierry Carrez <thierry at openstack.org> wrote:
> <SNIP>
>>> - Does that (or should that) also affect stable/kilo and stable/juno ?
>>> (there is no upper-constraints there)
>>
>> No, and since the disruption of a new pbr and associated minimum
>> versions throughout stable would be huge, I've no plans to backport
>> it. If folk wanted to (e.g. if stable suffers from lots of firedrills
>> even with the caps we added last time) I can throw up some patches and
>> we can see about it: but its a big effort requiring point releases of
>> /everything/ (because of the pbr version issue).
>>
>
> At the moment, it seems that additional point releases need to be
> created on-demand - but the situation does seem better than it did a
> year ago.  Is it less effort to do this across the board, once for all
> stable/* or continue the current process?
>
> Long term, we'll benefit from this on stable/liberty - but defining a
> process for the old world is probably useful.

Right - so if you look at stable kilo its largely uncapped. I'd still
expect issues there. Like I say, its doable, and I'm happy to prep
patches, but there's an awful lot of repos to update, unless we have
definite buy-in I'd rather not start.

Also, we should wait until we've bedded down all the remaining issues
in master, because then we only have one patch per thing to backport.

-Rob


-- 
Robert Collins <rbtcollins at hp.com>
Distinguished Technologist
HP Converged Cloud