Open Stack

Dina's analysis is a good one.

Basically oslo.messaging *completely violated* oslo-incubator rules and
updated common code in their tree. Then shipped a release, then broke
anyone trying to use them. This is actually very not cool.

This was not caught in the gate because the only thing that tickles this
is currently something which actually tries to touch all the config
options, and includes oslo.messaging, which is nova's config generator,
in the nova pep8 job.

Currently in the devstack/gate we aren't doing anything with code which
touches the ca_file option, which means python's lazy loading isn't ever
catching / exploding there.

The immediate fix is to block 1.3.0a8 -
https://review.openstack.org/#/c/77844/2/global-requirements.txt

The longer term is to actually come up with a gate job which tests for
these kinds of incompatibilities in a live environment.

I also really think we need to revisit this idea of config options
getting declared inside oslo-incubator. Because at 5 source trees this
was manageable, but with the current number of projects including the
incubator, the upgrade process for something like this becomes crazy
really fast. This is just the tip of the iceberg as we go to carving up
the incubator more.

	-Sean

On 03/04/2014 06:44 AM, Dina Belova wrote:
> Michael, hello.
> 
> I've found this issue some days ago with oslo.messaging master (and now
> it's released, as you see..)
> 
> The main problem here is: *Error importing module
> nova.openstack.common.sslutils: duplicate option: ca_file*
> 
> That's because of https://review.openstack.org/#/c/71997/ merged to
> oslo.messaging - and now released
> 
> Author changed opts not only in oslo.messaging files, but also in
> openstack.compute one - sslutils - and we've got duplicate option error
> with openstack.common.sslutils in nova.
> 
> As I discussed that with Doug Hellmann (and understood him), on
> #openstack-dev some days ago, he proposed to merge the same fix to
> oslo-incubator and then update all other projects using oslo.messaging.
> 
> Here is the fix: https://review.openstack.org/#/c/76300/ to oslo.incubator
> 
> Although, I suppose now your idea with simply not using last
> oslo.messaging release will be much quicker due to code freeze.
> 
> 
> On Tue, Mar 4, 2014 at 2:38 PM, Michael Still <mikal at stillhq.com
> <mailto:mikal at stillhq.com>> wrote:
> 
>     Hi.
> 
>     You might have noticed that the nova gate is currently broken. I
>     believe this is related to an oslo.messaging release today, and have
>     proposed a fix at https://review.openstack.org/#/c/77844/
> 
>     Cheers,
>     Michael
> 
>     --
>     Rackspace Australia
> 
>     _______________________________________________
>     OpenStack-dev mailing list
>     OpenStack-dev at lists.openstack.org
>     <mailto:OpenStack-dev at lists.openstack.org>
>     http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
> 
> 
> 
> 
> -- 
> 
> Best regards,
> 
> Dina Belova
> 
> Software Engineer
> 
> Mirantis Inc.
> 
> 
> 
> _______________________________________________
> OpenStack-dev mailing list
> OpenStack-dev at lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
> 


-- 
Sean Dague
Samsung Research America
sean at dague.net / sean.dague at samsung.com
http://dague.net

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 482 bytes
Desc: OpenPGP digital signature
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20140304/c14992f4/attachment.pgp>