[ClusterLabs] [Pacemaker] [HA] RFC: moving Pacemaker openstack-resource-agents to stackforge

[Digimer]

<Resending to the Cluster Labs mailing list, this list is deprecated>

On 23/06/15 06:27 AM, Adam Spiers wrote:
> [cross-posting to openstack-dev and pacemaker lists; please consider
> trimming the recipients list if your reply is not relevant to both
> communities]
> 
> Hi all,
> 
> https://github.com/madkiss/openstack-resource-agents/ is a nice
> repository of Pacemaker High Availability resource agents (RAs) for
> OpenStack, usage of which has been officially recommended in the
> OpenStack High Availability guide.  Here is one of several examples:
> 
>     http://docs.openstack.org/high-availability-guide/content/_add_openstack_identity_resource_to_pacemaker.html
> 
> Martin Loschwitz, who owns this repository, has since moved away from
> OpenStack, and no longer maintains it.  I recently proposed moving the
> repository to StackForge, and he gave his consent and in fact said
> that he had the same intention but hadn't got round to it:
> 
>     https://github.com/madkiss/openstack-resource-agents/issues/22#issuecomment-113386505
> 
> You can see from that same github issue that several key members of
> the OpenStack Pacemaker sub-community are all in favour.  Therefore
> I am volunteering to do the move to StackForge.

There is a "CusterLabs" group on github that most of the HA cluster
projects have or are moving under. Why not use that?

> Another possibility would be to move each RA to its corresponding
> OpenStack project, although this makes a lot less sense to me, since
> it would require the core members of every OpenStack project to care
> enough about Pacemaker to agree to maintain an RA for it.
> 
> This raises the question of maintainership.  SUSE has a vested
> interest in these resource agents, so we would be happy to help
> maintain them.  I believe Red Hat is also using these, so any
> volunteers from there or indeed anywhere else to co-maintain would be
> welcome.  They are already fairly complete, and I don't expect there
> will be a huge amount of change.
> 
> I'm probably getting ahead of myself, but the other big question is
> regarding CI.  Currently there are no tests at all.  Of course we
> could add bashate, and maybe even some functional tests, but
> ultimately some integration tests would be really nice.  However for
> now I propose we focus on the move and defer CI work till later.
> 
> Thoughts?
> 
> Thanks!
> Adam
> 
> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
> 
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org
> 


-- 
Digimer
Papers and Projects: https://alteeve.ca/w/
What if the cure for cancer is trapped in the mind of a person without
access to education?