On 11/09/15 22:18, matt wrote:
Hell. There's no clear upgrade path, and no guaranteed matched
functionality just for starters.
Also most enterprise deployments do 3 to 5 year deployment plans.
This ties into how equipment / power / resources are budgeted in the
project plans. They don't work with this mentality of rapid release
cycles.
We assumed early on that the people deploying OpenStack would be more
agile because of the ephemeral nature of cloud. That's not really
what's happening. There are good and bad reasons for that. One good
reason is policy certification. By the time a team has prepped,
built, tested an environment and is moving to production it's already
been an entire release ( or two since most ops refuse to use a fresh
release for stability reasons ). By the time it passes independent
security / qa testing and development workflows for deploying apps to
the environment it's been 3-4 releases or more. But more often than
not the problem is most of the VM workloads aren't good with ephemeral
and mandating downtime on systems is an onerous change control
process. Making the upgrade process for the environment very
difficult and time consuming.
More than that vendors that provide extra ( sometimes necessary )
additions to openstack, such as switch vendors take at least a few
months to test a new release and certify their drivers for
deployment. Most folks aren't even beginning to deploy a fresh
release of openstack EVEN if they wanted to until it's been out for at
least six months. It's not like they can really test pre-rc releases
and expect their tests to mean anything.
There's almost no one riding the wave of new deployments.
Matt - every word above is golden. Well said!
On Mon, Nov 9, 2015 at 3:06 PM, Tom Cameron <tom.came...@rackspace.com
<mailto:tom.came...@rackspace.com>> wrote:
>I would not call that the extreme minority.
>I would say a good percentage of users are on only getting to
Juno now.
The survey seems to indicate lots of people are on Havana,
Icehouse and Juno in production. I would love to see the survey
ask _why_ people are on older versions because for many operators
I suspect they forked when they needed a feature or function that
didn't yet exist, and they're now stuck in a horrible parallel
universe where upstream has not only added the missing feature but
has also massively improved code quality. Meanwhile, they can't
spend the person hours on either porting their work into the new
Big Tent world we live in, or can't bare the thought of having to
throw away their hard earned tech debt. For more on this, see the
myth of the "sunken cost".
If it turns out people really are deploying new clouds with old
versions on purpose because of a perceived stability benefit, then
they aren't reading the release schedule pages close enough to see
that what they're deploying today will be abandoned soon in the
future. In my _personal_ opinion which has nothing to do with
Openstack or my employer, this is really poor operational due
diligence.
If, however, a deployer has been working on a proof of concept for
18-24 months and they're now ready to go live with their cloud
running a release from 18-24 months ago, I have sympathy for them.
The bigger the deployment, the harder this one is to solve which
makes it a prime candidate for the LTS strategy.
Either way, we've lost the original conversation long ago. It
sounds like we all agree that an LTS release strategy suits most
needs but also that it would take a lot of work that hasn't yet
been thought of or started. Maybe there should be a session in
Austin for this topic after blueprints are submitted and
discussed? It would be nice to have the operators and developers
input in a single place, and to get this idea on the radar of all
of the projects.
--
Tom Cameron
________________________________________
From: Maish Saidel-Keesing <mais...@maishsk.com
<mailto:mais...@maishsk.com>>
Sent: Monday, November 9, 2015 14:29
To: Tom Cameron; Jeremy Stanley;
openstack-operators@lists.openstack.org
<mailto:openstack-operators@lists.openstack.org>
Subject: Re: [Openstack-operators] [openstack-dev] [stable][all]
Keeping Juno "alive" for longer.
On 11/09/15 21:01, Tom Cameron wrote:
> From your other thread...
>
>> Or else you're saying you intend to fix the current inability
of our projects to skip intermediate releases entirely during upgrades
> I think without knowing it, that's what most would be
suggesting, yeah. Of course, like you mentioned, the real work is
in how upgrades get refactored to skip intermediate releases (two
or three of them).
>
> DB schema changes can basically be rolled up and kept around for
a while, so that's not too be a problem. Config files OTOH have no
schema or schema validator, so that would require tooling and all
kinds of fun (bug prone) wizardry.
>
> This is all solvable, but it adds complexity for the sake of
what I can only imagine are the extreme minority of users. What do
the user/operator surveys say about the usage of older releases?
What portion of the user base is actually on releases prior to Havana?
I would not call that the extreme minority.
I would say a good percentage of users are on only getting to Juno
now.
>
> --
> Tom Cameron
>
>
> ________________________________________
> From: Jeremy Stanley <fu...@yuggoth.org <mailto:fu...@yuggoth.org>>
> Sent: Monday, November 9, 2015 12:35
> To: openstack-operators@lists.openstack.org
<mailto:openstack-operators@lists.openstack.org>
> Subject: Re: [Openstack-operators] [openstack-dev] [stable][all]
Keeping Juno "alive" for longer.
>
> On 2015-11-09 17:11:35 +0000 (+0000), Tom Cameron wrote:
> [...]
>> I support an LTS release strategy because it will allow more
>> adoption for more sectors by offering that stability everyone's
>> talking about. But, it shouldn't be a super-super long support
>> offering. Maybe steal some of Ubuntu's game and do an LTS every 4
>> releases or so (24 months), but then maybe Openstack only supports
>> them for 24 months time? Again, my concern is that this is free,
>> open source software and you're probably not going to get many
>> community members to volunteer to offer their precious time fixing
>> bugs in a 2-year-old codebase that have been fixed for 18 months
>> in a newer version.
> [...]
>
> Because we want people to be able upgrade their deployments, the
> problem runs deeper than just backporting some fixes to a particular
> branch for longer periods of time. Unfortunately the original poster
> cross-posted this thread to multiple mailing lists so the discussion
> has rapidly bifurcated, but I addressed this particular topic in my
>
http://lists.openstack.org/pipermail/openstack-dev/2015-November/078735.html
> reply.
> --
> Jeremy Stanley
>
> _______________________________________________
> OpenStack-operators mailing list
> OpenStack-operators@lists.openstack.org
<mailto:OpenStack-operators@lists.openstack.org>
>
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
> _______________________________________________
> OpenStack-operators mailing list
> OpenStack-operators@lists.openstack.org
<mailto:OpenStack-operators@lists.openstack.org>
>
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
--
Best Regards,
Maish Saidel-Keesing
_______________________________________________
OpenStack-operators mailing list
OpenStack-operators@lists.openstack.org
<mailto:OpenStack-operators@lists.openstack.org>
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
_______________________________________________
OpenStack-operators mailing list
OpenStack-operators@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
--
Best Regards,
Maish Saidel-Keesing
_______________________________________________
OpenStack-operators mailing list
OpenStack-operators@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators