Rexray: Planning for REX-Ray in a foundation

[clintkitson]

Hello REX-Ray Community and Collaborators!

Lots of news and great opportunities in the container eco-system. With the introduction of the Container Storage Interface (CSI), the community has cemented the role of storage and volumes with applications. This enables storage platforms and tools to focus and create trusted integrations to universally support persistent applications.

REX-Ray provided important early validation to the CSI specification by releasing initial support in 0.10. The upcoming REX-Ray 0.12 release will include CSI 0.1 support for all existing storage drivers and any new native CSI drivers.

With the future looking very positive for CSI, we feel like there is no time better than now for the community to come together and collaborate on a great tool and framework for CSI plugin implementations. More specifically it is clear to us that providing a solid user experience and a base set of features/functionality is important for CSI adoption and all drivers. REX-Ray can fill this need.

To increase collaboration, the {code} team is planning on contributing the REX-Ray project to a non-profit foundation in 2018. Thumbs up!

For this we could use the support of our community to be vocal about a couple of key things.

For our users: Are you using REX-Ray or have history with it? Can you provide a reference to these foundations? If so reply to this thread or ping me on Slack!
For the storage eco-system: Will you collaborate on REX-Ray if it has open governance and is in a foundation?

[j-griffith]

I think this would be a fantastic and extremely beneficial move, not only for vendors and developers but more importantly the Container eco-system and it's adopters. Currently things are extremely fragmented, with folks off doing their own thing. The ability to have a Foundation governed and community driven framework with common goals means that we can focus our efforts. I'm a big believer in the power of Open Source Communities, and I think it would be fantastic if we could build one around CSI. Yes, REX-Ray has always been Open Source and welcomed PR's and contributions, but building a community around a project associated (no matter how loosely) with a Vendor is always a bit of a challenge.

Anyway, to me this seems like a fantastic step forward; looking forward to a diverse and talented community and contributing to it's success.

[cygnus8595]

As long as you keep the mascot I'm good with it. :)

I think it's a great step and it'll more likely to ensure the future of the project. One thing I've noticed about the container orchestration space is that a lot of good projects fall away if they aren't widely adopted.

[gourao]

I think REX-Ray would be a good addition to the storage landscape. As a CSI implementation, it provides an open source solution for multiple storage backends to the community.

[chakri-nelluri]

+1. It's a great move. Will reduce the fragmentation. All in for it.

[yasker]

REX-Ray has achieved unparalleled success in building a container storage framework for various storage vendors since the early days of Docker. It has reduced the effort to write a storage driver for the different plugin systems of container platforms, especially with the support of CSI added in. We've integrated REX-Ray with Rancher since 2016 and got great feedback. We think it's a great step forward to contributing REX-Ray to a non-profit foundation and continuing building up from there. We support the move 100%.

[akutz]

REX-Ray? Never heard of it.

[h0tbird]

+1 I am using REX-Ray in one project. I support any initiative to keep it active and healthy.

[benh]

+1 moving this into the foundation!

[brainstorm]

As an outsider this move sounds a bit confusing:

https://twitter.com/braincode/status/984624499025854465

What are the actual actions or philosophy behind "To increase collaboration, the {code} team is planning on contributing the REX-Ray project to a non-profit foundation in 2018"?:

1. stopping the current development flow and refactoring most of the codebase to accommodate CSI spec better?
2. Continue with the current codebase mostly untouched and solve the growing issue list?
3. How will the companies involved in RR (VMWare et al) contribute to RR under this new model?

Just trying to get a sense on where this project is headed and whether our particular org can invest efforts on solving our current issue and having it as a solid foundation for our infrastructure ;)

/cc @reisingerf