OpenStack architect for almost 10 years putting in my two cents after stumbling across this thread searching for something else.

I think there is a missed opportunity here that isn't being addressed which is: why are companies moving away from OpenStack to begin with. The question that should be asked is not how do we revamp documentation or how to add more contributors but rather what this community can do to draw them back. By failing to attract customers with a platform technology that is easily consumable and understandable, it's akin to brainstorming new flavors of cough syrup while failing to address the underlying sickness.

To be perfectly blunt, my feeling is if the TC cannot solve for or drive towards easing/addressing OpenStack's technical challenges that drive away customers/inhibit adoption overall then I honestly question the value of having a TC in the first place. This should be central to their charter and it's my view so far they've failed in that regard.

As one who gets annoyed with criticism without alternatives, I believe we can resolve the contributor issue with some focus in a few keys areas and none of these suggestions involves trying to attract them directly:
  • Prioritize a simplified installer with minimal inputs that gets a platform online on a simple form factor with minimal dependencies and an abundance of assumptions.
  • Revisit the charter of the TC which needs a serious rework since we should all be highly dissatisfied given their past tendency to maintain the status quo.
  • Revisit what enterprises need and how they're leveraging cloud. If we aren't listening to why folks are leaving we will continue producing increasingly-irrelevant software. Trust me, if we don't solve our problems in this regard, someone else will and OpenStack will be just another could-have-been technology that failed to watch, listen and adapt.
If we solve for why companies are abandoning OpenStack -- both as corporate contributors and as consumers -- getting contributors will just be a by-product of that success. But not before.

//adam

Adam Peacock

Principal Architect
Office: +1-916-794-5706


On Tue, Apr 7, 2020 at 11:20 AM Jeremy Freudberg <jeremyfreudberg@gmail.com> wrote:
On Tue, Apr 7, 2020 at 12:42 PM Rico Lin <rico.lin.guanyu@gmail.com> wrote:
> [...]
> Maybe we should start to think a `one command` installed environment so the user can
> try on for the first time and without learning huge knowledge base to write devstack script.
> Like `openstack install`? :)
> [...]

I think you are describing https://microstack.run/ ?
I never tried it but it looks interesting.

Should the TC maintain a reference document of simple OpenStack
evaluation tools? e.g. packstack, microstack, tripleo-quickstart...?