Micro release exception for Nova, Glance, Horizon, Keystone

[Steve Langasek]

Hello there,

On Wed, Jun 20, 2012 at 08:44:26PM +0000, Matt Zimmerman wrote:

> If the SRU team feels it's going well, I'm in favor.

On Wed, Jun 20, 2012 at 07:57:29PM +0000, Kees Cook wrote:

> Can someone from the SRU team vouch for this package?

On Fri, Jun 22, 2012 at 00:07:52AM +0000, Mark Shuttleworth wrote:

> Is it worth considering a formal delegation of these requests /
> decisions to SRU+Security teams, with TB as an escalation path?

As a member of the SRU team, I can say that I am in favor of the nova
package being granted an MRE based on the stated upstream update policy and
the stated committment of the server team to do additional regression
testing via deployment of the packages in -proposed.

I don't say that I /vouch/ for this package however.  This is comparatively
new software that has never undergone a successful SRU for a new upstream
point release, so we don't really have any practical experience that lets us
judge the quality of the upstream regression testing.  From reading the TB
list archives of the past months, that seems to be one of the things that
some members of the TB are requesting as a prerequisite for an MRE.

That's not consistent with the documented policy on
<https://wiki.ubuntu.com/StableReleaseUpdates/MicroReleaseExceptions>
though, and I don't think that's the right way to structure our handling of
MREs.  The high demand for SRUs following the 12.04 LTS release has led to
the SRU team members individually being pressured to accept SRUs that
*don't* fit the stated policy of patches that are "as small and unintrusive
as possible".  If the SRU team is put in the position of having the
authority of being able to sign off on these exceptions directly, I think an
inevitable outcome is that the SRU team will cave to that pressure, and
accept SRUs against their better judgement and without the level of scrutiny
of the upstream regression testing that the MRE process calls for.

I recognize that the TB doesn't have an infinite amount of time for
processing MRE requests and so you might prefer to delegate this to the SRU
team, but I don't think the SRU team is in a much better position there, in
point of fact.  We've been working this month to expand the SRU team to
address a manpower gap; I think that having the SRU team approving
exceptions would make that gap much worse and lead to poor decisions around
SRU acceptance.  OTOH, if the TB is going to rely on SRU team input before
granting MREs, I think the exception requests are likely to stall precisely
because the SRU team often doesn't have a basis for an informed opinion.

So there's my two cents in advance of today's TB meeting, which has on the
agenda to discuss the MRE process. :)

In the meantime, on the basis of the timbre of this thread, I'm going to go
ahead and accept the current nova SRU to precise-proposed to unblock folks
for testing.  I would appreciate it if some member of the TB would formally
grant an MRE here (if that's the intent) and add it to the list on
<https://wiki.ubuntu.com/StableReleaseUpdates/MicroReleaseExceptions>.

Thanks,
-- 
Steve Langasek                   Give me a lever long enough and a Free OS
Debian Developer                   to set it on, and I can move the world.
Ubuntu Developer                                    http://www.debian.org/
slangasek at ubuntu.com                                     vorlon at debian.org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <https://lists.ubuntu.com/archives/technical-board/attachments/20120625/e017c143/attachment.pgp>