|
On 02/02/2015 12:25 PM, Niels de Vos
wrote:
On Sat, Jan 31, 2015 at 07:43:39AM -0500, Justin Clift wrote:Hi all, One of the things which has been a fair drag for the developer part of the Gluster Community, is maintaining our own Jenkins infrastructure. When chatting with the CentOS guys (pre-FOSDEM) the other day, there is the possibility we could use their Jenkins infrastructure instead. They have a *bunch* of physical nodes (100+), and several other projects are using it as well (Ceph, and others). This would potentially make things easier for us, and they're not limited to running things on CentOS either. They already have FreeBSD nodes, and don't foresee problems with getting NetBSD up and running. At the moment, I'm kind of having the opinion this could be a good way forward, and we should probably try out our smoke/regression scripting on their setup to see if it's workable. And if it turns out to be, then great, and we migrate all of our jenkins there then decommission ours (including the master. :) What do people reckon?Sounds good to me in general. Is there a document that describes their Jenkins setupi (like provisioning of the slaves)? Note that we do not (yet) have a way of automated multi-host tests in Jenkins, but that should be looked into one day. Also, is it possible for Gluster Community members to log into a Jenkins slave and troubleshoot specific failures? Thanks, Niels +1. We just need to make sure, CentOS would give required access to gluster community members even if they are not doing anything in CentOS community. Thanks, Lala
|
_______________________________________________ Gluster-devel mailing list Gluster-devel@xxxxxxxxxxx http://www.gluster.org/mailman/listinfo/gluster-devel
