[openstack-dev] Dependency freeze exception for happybase (I would like version 0.8)

Sean Dague sean at dague.net
Fri Mar 28 21:01:54 UTC 2014


So how did Ceilometer get into this situation? Because the ceilometer
requirements are happybase>=0.4,<=0.6

So we told distributions at Feature Freeze that was adequate version for
ceilometer, and now we're telling them it's not at RC. And that's
honestly too late for most of them.

So I remain -1 on this at this point (if ttx want's to override, that's
his call). And we need to figure out how we don't end up in the same
position in Juno.

	-Sean

On 03/28/2014 03:56 PM, Nadya Privalova wrote:
> Sean, please see my comment in bug regarding 0.7. Unfortunately I have
> nothing to suggest here. If distributive contains only 0.7 we cannot
> work with HBase through happybase. And I don't see any solutions.
> But anyway if you change requirements for happybase (I guess we need to
> get 0.7 back, no?) could you please remove 0.4? 0.4 doesn't allow to
> work with pools but we need to.
> 
> Thanks,
> Nadya
> 
> 
> On Fri, Mar 28, 2014 at 9:07 PM, Sean Dague <sean at dague.net
> <mailto:sean at dague.net>> wrote:
> 
>     On 03/28/2014 12:50 PM, Thierry Carrez wrote:
>     > Thierry Carrez wrote:
>     >> Julien Danjou wrote:
>     >>> On Thu, Mar 27 2014, Thomas Goirand wrote:
>     >>>
>     >>>> -happybase>=0.4,<=0.6
>     >>>> +happybase>=0.8
>     >>>
>     >>> Good for me, and Ceilometer is the only one using happybase as
>     far as I
>     >>> know so that shouldn't be a problem.
>     >>
>     >> OK so I would be fine with "happybase>=0.4,!=0.6,!=0.7" as it
>     allows 0.8
>     >> to be run without adversely impacting downstreams who don't have it
>     >> available.
>     >
>     > Actually that should be "happybase>=0.4,!=0.7" since 0.6 was
>     allowed before.
> 
>     The review has comments about 0.7 being the version in Debian and Ubuntu
>     right now https://review.openstack.org/#/c/82438/.
> 
>     Which is problematic, as that's a version that's specifically being
>     called out by the ceilometer team as not useable. Do we know why, and if
>     it can be worked around?
> 
>             -Sean
> 
>     --
>     Sean Dague
>     Samsung Research America
>     sean at dague.net <mailto:sean at dague.net> / sean.dague at samsung.com
>     <mailto:sean.dague at samsung.com>
>     http://dague.net
> 
> 
>     _______________________________________________
>     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
> 
> 
> 
> 
> _______________________________________________
> 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/20140328/8e2b687f/attachment.pgp>


More information about the OpenStack-dev mailing list