[openstack-dev] [tripleo] CI Squad Meeting Summary (week 21) - Devmode OVB, RDO Cloud and config management
adarazs at redhat.com
Fri May 26 14:58:27 UTC 2017
If the topics below interest you and you want to contribute to the
discussion, feel free to join the next meeting:
Time: Thursdays, 14:30-15:30 UTC
Full minutes: https://etherpad.openstack.org/p/tripleo-ci-squad-meeting
= Periodic & Promotion OVB jobs Quickstart transition =
We had a lively technical discussions this week. Gabriele's work on
transitioning the periodic & promotion jobs is nearly complete, only
needs reviews at this point. We won't set a transition date for these as
it is not really impacting folks long term if these jobs are failing for
a few days at this point. We'll transition when everything is ready.
= RDO Cloud & Devmode OVB =
We continued planning the introduction of RDO Cloud for the upstream OVB
jobs. We're still at the point of account setup.
The new OVB based devmode seems to be working fine. If you have access
to RDO Cloud, and haven't tried it already, give it a go. It can set up
a full master branch based deployment within 2 hours, including any
pending changes baked into the under & overcloud.
When you have your account info sourced, all it takes is
$ ./devmode.sh --ovb
from your tripleo-quickstart repo! See here for more info.
= Container jobs on nodepool multinode =
Gabriele is stuck with these new Quickstart jobs. We would need a deep
dive into debugging and using the container based TripleO deployments.
Let us know if you can do one!
= How to handle Quickstart configuration =
This a never-ending topic, on which we managed to spend a good chunk of
time this week as well. Where should we put various configs? Should we
duplicate a bunch of variables or cut them into small files?
For now it seems we can agree on 3 levels of configuration:
* nodes config (i.e. how many nodes we want for the deployment)
* envionment + provisioner settings (i.e. you want to run on rdocloud
with ovb, or on a local machine with libvirt)
* featureset (a certain set of features enabled/disabled for the jobs,
like pacemaker and ssl)
This seems rather straightforward until we encounter exceptions. We're
going to figure out the edge cases and rework the current configs to
stick to the rules.
That's it for this week. Thank you for reading the summary.
More information about the OpenStack-dev