<html><head><meta http-equiv="content-type" content="text/html; charset=utf-8"></head><body dir="auto"><div><br></div>Let me share my experience because I’m end user and cloud operator. I’m running 5 large production deployment on openstack-ansible (each cloud has more than 250+ computes nodes so you know scale). Openstack-ansible community is very supportive and helped me a lot to bring up massive deployment with scale. <div><br></div><div>Few month back one of my customer ask me to use kolla-ansible so I use kolla-ansible to reply their cloud. For me both are great but I have more command on openstack-ansible and after deploying 5 large cloud it’s in my blood. </div><div><br></div><div>Both ansible based so I don’t think you will have any difficulty to play. </div><div><br></div><div>You can find me or others on IRC if you have more questions. <br><br><div dir="ltr">Sent from my iPhone</div><div dir="ltr"><br><blockquote type="cite">On Jun 10, 2022, at 10:56 AM, Mark Goddard <mark@stackhpc.com> wrote:<br><br></blockquote></div><blockquote type="cite"><div dir="ltr"><div dir="ltr"><div class="gmail_default" style="font-family:verdana,sans-serif">This has been discussed before, but it would be nice if we as a community had an easier way to answer this question. I know many of us feel conflicted about aggressively promoting our own projects over alternatives, but also want to give them a fair chance. Maybe each project needs a "why pick me" page that could be linked to in a standard response.<br></div><div class="gmail_default" style="font-family:verdana,sans-serif"><br></div><div class="gmail_default" style="font-family:verdana,sans-serif">Mark</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Thu, 9 Jun 2022 at 16:40, Jonathan Rosser <<a href="mailto:jonathan.rosser@rd.bbc.co.uk">jonathan.rosser@rd.bbc.co.uk</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div>
<p>Hi Dave,</p>
<p>I have been hesitating to reply to your mailing list post because
it doesn't feel right to pitch two community tools against each
other here on the mailing list - so i won't do that here.<br>
</p>
<p>I would say that the deployment tool is a means to an end. So you
should look at the technology choices, upgrade paths, support for
"day 2 operations", how bugs get addressed, documentation,
operator experience etc etc. Only you can decide which is
appropriate for the constraints and requirements of your
deployment.<br>
</p>
<div>My reply will obviously be biassed, as
I am a big contributor to openstack-ansible. My observation is
that the operators that gain the most out of any of these tools
are the ones who engage with the community around those tools,
primarily in the case of openstack-ansible that would be through
our IRC channel, weekly meetings and bug reports on Launchpad. You
will gain insight and be able to leverage the knowledge of other
operators who in some cases have literally written the book on
various aspects of OpenStack. Trying to fight though every
decision or problem on your own is the worst way to approach using
any of these community driven tools.</div>
<div><br>
</div>
<div>If you instead want a more "shrink
wrap" approach to an installer, or more formal support, then it
would be wise to look at the product oriented offerings from the
large vendors.</div>
<div><br>
</div>
<div>Both openstack-ansible and
kolla-ansible will expect you to make a good number of decisions
about the specifics of your deployment, for example storage,
networking and security concerns. Both would also expect you to
gain sufficient knowledge about how OpenStack itself works to be
able to make good use of the customization opportunities that both
provide. This is really the unique selling point of the community
tooling, you get a very high degree of customization potential but
that can come at the cost of some complexity.</div>
<div><br>
</div>
<div>As you are already using
openstack-ansible I would suggest that you continue, but as I've
already said I have an existing interest here and I really don't
want to start a tooling debate. Join us in IRC in
#openstack-ansible and discuss any pain points you have. This way
we can hopefully help you out, or address specific issues in the
code - you may have discovered legitimate bugs or a use case that
is not straightforward to fulfill. This is how all of the
community tools get improved and evolved over time.<br>
</div>
<div><br>
</div>
<div>On one specific point I would recommend
that you move entirely to Debian 11 as Xena will be the last
release that openstack-ansible supports Buster.</div>
<div><br>
</div>
<div>I'm not sure there is a fool-proof
installer really. Installing the code is one thing, being
effective though upgrades and applying bugfixes to a production
environment is a different and a more important concern in the
long term. Both openstack-ansible and kolla-ansible offer
"all-in-one" deployments which are intended as "should-just-work"
demonstrators of how things fit together and for lightweight
testing. Scaling those out to larger deployments is where the real
work is, and neither tool sets out to be particularly prescriptive
about some parts of how you build your environment.<br>
</div>
<div><br>
</div>
<div>Hopefully this is helpful,<br>
Jonathan.<br>
</div>
<div><br>
</div>
<div>On 09/06/2022 15:58, Dave Hall wrote:<br>
</div>
<blockquote type="cite">
<div dir="ltr">
<div dir="ltr">
<div>Hello,</div>
<div><br>
</div>
<div>My question is about <span>OpenStack</span>-<span>Ansible</span> <span>vs</span>. <span>Kolla</span>-<span>Ansible</span>.
While I am sensitive to the effort that has been put into
both of these projects, what I really need right now is the
most fool-proof way to deploy and manage a small production
cluster for academic instructional use.</div>
<div><br>
</div>
<div>(BTW, I do understand that there are other differences
between Kolla and regular openstack.)</div>
<div><br>
</div>
<div>I have been working for a couple months with <span>OpenStack</span>-<span>Ansible</span> to
deploy a small (3-node) Xena test cluster on VirtualBox VMs
in preparation for a larger deployment on real hardware - 6
to 10 nodes. My VirtualBox deployment has been:</div>
<div><br>
</div>
<div>
<blockquote style="margin:0px 0px 0px 40px;border:none;padding:0px">
<div>Debian 11 deployment, Debian 10 infra, compute, and
storage nodes</div>
</blockquote>
</div>
<div><br>
</div>
<div>It has been slow going, at least partially due to some
issues and limitations with VirtualBox (and Vagrant early
on). However, deploying a test cluster on VMs still seems
preferable to just diving into deployment on real hardware
and going through multiple scrubs/reinstalls.</div>
<div><br>
</div>
<div>I've recently seen more posts in the list about <span>Kolla</span>-<span>Ansible</span>.
So, as a 'beginner', should I shift and look at <span>Kolla</span>-<span>Ansible</span>,
or should I stay the course and continue with <span>Openstack</span>-<span>Ansible</span>?
What are the pros and cons of each?</div>
<div><br>
</div>
<div>For that matter, is there some other deployment mechanism
that would be more fool-proof for a first-timer? Although
I'm more familiar with Ansible than the other tools (Chef,
Puppet, etc.) I'm most interested in how to get a cluster up
and running regardless of the underlying tools.</div>
<div><br>
</div>
<div>Thanks.</div>
<div><br>
</div>
<div>-Dave</div>
<div><br>
</div>
<div>
<div dir="ltr">
<div dir="ltr">
<div>--</div>
<div>Dave Hall<br>
Binghamton University<br>
<a href="mailto:kdhall@binghamton.edu" target="_blank">kdhall@binghamton.edu</a></div>
<div><br>
</div>
</div>
</div>
</div>
</div>
</div>
</blockquote>
</div>
</blockquote></div>
</div></blockquote></div></body></html>