[openstack-dev] [puppet] What's up in our CI?
emilien at redhat.com
Thu Feb 25 02:37:57 UTC 2016
On 02/24/2016 09:28 PM, Emilien Macchi wrote:
> 1/ CI failures
> For those who contributed to Puppet OpenStack modules over the last
> days, you might have noticed our CI was pretty unstable.
> The main reason is that our 2 scenarios are overloaded by the number of
> services & tests, so they randomly timeout.
> Here are some actions that should bring back our CI in a good shape:
> * Add scenario003 and move some services from 001/002 to 003
> * Optimize Zuul layout to consume less resources in OpenStack Infra
> * Continue to reduce to 2 the number of workers when possible
> * Investigate why `gem install r10k` takes 12 minutes on Internap cloud
> (instead of a few seconds on other clouds)
> 2/ scenario001 is the RBD scenario
> scenario001 is now deploying Glance, Nova, Cinder and Gnocchi with RBD
> backend (Ceph).
> We can call it a Compute + Telemetry + Ceph scenario.
> 3/ What's next
> Currently in the pipe:
> * running tempest with plugins
> * add Zaqar
> * add Neutron FWaaS, LBaaS
> * on scenario002: use Swift for Glance backand and Neutron ML2
> linuxbridge (instead of ML2 OVS).
> * more (please bring ideas & feedback)
> Any questions / suggestions are welcome,
I realize we have a new thing:
4/ Canonical updated UCA packaging for Mitaka (last update was in
December). Our integration jobs running on Trusty are all broken now.
We're investigating that but if we don't find a fix by tomorrow morning
I'll propose to disable voting on Ubuntu.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 473 bytes
Desc: OpenPGP digital signature
More information about the OpenStack-dev