On 10/09/2015 7:26 PM, Tony Breeds wrote:
Hi all,
In trying to fix a few stable/juno issues we need to release a new version
of ceilometerclient for stable/juno. This email is to try and raise awareness
so that if the proposal is bonkers [1] we can come up with something better.
This isn't currently possible due to the current caps in juno and kilo.
The proposed fix is to:
. update g-r in master (liberty): python-ceilometerclient>=1.2
https://review.openstack.org/#/c/222386/
. update g-r in stable/kilo: python-ceilometerclient>=1.1.1,<1.2
. release a sync of stable/kilo g-r to stable/kilo python-ceilometerclient as
1.1.1
. update g-r in stable/juno: python-ceilometerclient<1.1.0,!=1.0.13,!=1.0.14
. release 1.0.15 with a sync of stable/juno g-r
The point is, leave 1.0.x for juno, 1.1.x for kilo and >=1.2 for liberty
This is being tracked as:
https://bugs.launchpad.net/python-ceilometerclient/+bug/1494516
There is a secondary issue if getting the (juno) gate in a shape where we can
actually do all of that.
Yours Tony.
[1] Bonkers is a recognized technical term right?
i commented on patch already but to reiterate, this sounds sane to me.
we tagged stuff improperly during juno/kilo timespan so our versioning
became an issue[1] and it looks like it caught up to us.
as it stands, version 1.1.0 is the rough equivalent to 1.0.14 (but with
a requirement updates). this seems to solve all the requirements issues
so i'm content with the solution. thanks to both of you for figuring out
the requirements logistics.
[1]
http://eavesdrop.openstack.org/irclogs/%23openstack-ceilometer/%23openstack-ceilometer.2015-04-14.log.html#t2015-04-14T21:50:49
cheers,
--
gord
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: [email protected]?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
