[openstack-dev] [devstack] etcd3 enabled breaks s390x arch

Davanum Srinivas davanum at gmail.com
Wed May 24 12:57:37 UTC 2017


Thanks for the help Mikhail,

So just FYI for others, the etcd 3.2.0 is in RC1, We will get a full
set of arch(es) covered once that goes GA

Thanks,
Dims

On Wed, May 24, 2017 at 8:45 AM, Mikhail S Medvedev <mihailmed at gmail.com> wrote:
>
> On 05/24/2017 06:59 AM, Sean Dague wrote:
>>
>> On 05/24/2017 07:48 AM, Andreas Scheuring wrote:
>> > Hi together,
>> >
>> > recently etcd3 was enabled as service in devstack [1]. This breaks
>> > devstack on s390x Linux, as there are no s390x binaries availabe and
>> > there's no way to disable the etcd3 service.
>> >
>> > I pushed a patch to allow disabling the etcd3 service in local.conf [2].
>> > It would be great if we could get that merged soon to get devstack going
>> > again. It seems like that is not used by any of the default services
>> > (nova, neutron, cinder,...) right now.
>> >
>> > In the long run I would like to understand the plans of etcd3 in
>> > devstack. Are the plans to make the default services dependent on etcd3
>> > in the future?
>> >
>> > Thanks a lot!
>> >
>> > Andreas
>> >
>> >
>> > [1]
>> >
>> > https://github.com/openstack-dev/devstack/commit/546656fc0543ec2bc5b422fd9eee17f1b8122758
>> > [2] https://review.openstack.org/467597
>>
>> Yes, it is designed to be required by base services. See -
>> http://lists.openstack.org/pipermail/openstack-dev/2017-May/117370.html
>>
>>     -Sean
>>
> It is designed to be required, but please be aware of other arches. E.g. the
> original change do DevStack [3] did not allow much flexibility, and only
> worked on x86 and aarch. The d-g change [4] broke d-g for non-x86 arches. I
> have submitted [5] to add some flexibility to be able to specify a different
> mirror from which to pull non-x86 etcd3.
>
> In the last couple days I am playing a whack-a-mole with all of that and
> more. At some point I did request a permission to add PowerKVM CI (ppc64) to
> devstack-gate patches, which might have helped to identify the problem
> earlier. Maybe it should be revisited?
>
> [3] https://review.openstack.org/#/c/445432/
> [4] https://review.openstack.org/#/c/466817/
> [5] https://review.openstack.org/#/c/467437/
>
> ---
> Mikhail Medvedev (mmedvede)
> OpenStack CI for KVM on Power
> IBM
>
>
> __________________________________________________________________________
> OpenStack Development Mailing List (not for usage questions)
> Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev



-- 
Davanum Srinivas :: https://twitter.com/dims



More information about the OpenStack-dev mailing list