Open Stack

Hi again openstack-infra,

We've worked quite a bit on ARA since I first told you guys about it here.
We're pretty much out of PoC and Alpha territory by now and I've made
a formal public announcement about it [1].

Following up on the questions from the thread now that we've progressed a bit:
- Usability with things like ElasticSearch:
  The ARA CLI client uses Cliff and thus has a very similar look and
feel as python-openstackclient.
  Cliff provides several output formats out of the box such as YAML or JSON.
  You could use the client to output the information you want in JSON
format and then send that off to ElasticSearch with already-indexed
fields.

- Static file generation
  This is very high on the list of things to do.
  We will make it possible for the ARA CLI client to generate a static
version of the interface for a single playbook run.
  So far we're aiming more towards a html "tree" rather than just a
single page to keep the same browsing experience.

I see a great opportunity for ARA to be used by the OpenStack community.
The project officially needs a home for code reviews and I think
openstack or openstack-infra with the usual gerrit workflow would be a
great fit if you believe it makes sense.

What do you think ?

Please let me know !

Thanks,

[1]: https://dmsimard.com/2016/05/21/ara-an-idea-to-store-browse-and-troubleshoot-ansible-playbook-runs/

David Moreau Simard
Senior Software Engineer | Openstack RDO

dmsimard = [irc, github, twitter]


On Tue, May 10, 2016 at 9:24 PM, David Moreau Simard <dms at redhat.com> wrote:
> Hi openstack-infra,
>
> I've been hacking on an idea since last friday and I was thinking it
> could be of interest for you guys as Ansible users, but also
> considering the next generation of Zuul jobs will be powered by
> Ansible.
>
> The project I'm working on is called ARA [1]. It is very much an alpha
> at this point but it is hopefully fleshed out enough at this point for
> you to see where I am going with this.
> I could go on about what the project looks like but I've actually
> built a sane amount of docs [2] and even put up a video of what the
> interface looks like on YouTube [3].
>
> ARA was born out of necessity because the vast majority of RDO CI [4]
> is driven by Ansible, either through projects like TripleO-Quickstart
> [5] or WeIRDO [6].
>
> We are equipped in a fairly similar way than upstream, we have Jenkins
> console logs that are large enough to crash your browser and logs that
> are collected and made browsable like on logs.o.o.
> In a nutshell, we're running CI jobs that install and test OpenStack
> deployments through various installers: TripleO, Packstack,
> Puppet-OpenStack, and soon: Kolla and Chef-OpenStack.
>
> The sad reality is that the RDO community has limited resources and
> ARA aims at making it easier and faster to see what failed, where and
> why to make everyone more efficient.
> Another sad truth is that I am much more a system administrator than a
> developer and as such I am not exactly awesome in either Python or
> UI/UX frontends.
>
> As such, if you are interested in the project, I would certainly
> welcome not only feedback but contributions if you feel you could use
> this.
>
> Let me know if you have any questions !
>
> [1]: https://github.com/dmsimard/ara
> [2]: http://ara.readthedocs.io/en/latest/
> [3]: https://www.youtube.com/watch?v=K3jTqgm2YuY
> [4]: https://ci.centos.org/view/rdo/view/promotion-pipeline/
> [5]: https://github.com/openstack/tripleo-quickstart
> [6]: https://github.com/redhat-openstack/weirdo
>
> David Moreau Simard
> Senior Software Engineer | Openstack RDO
>
> dmsimard = [irc, github, twitter]