 Alrighty, so we're here for another one of the Newton Designs series interviews and today it's with Steve Dake. So Steve, why don't you go ahead and introduce yourself and tell us about your background in OpenStack. Sure, so my name is Steven Dake and I work at Cisco Systems where I'm bringing containers to OpenStack on behalf of Cisco. I work on COLA as a PtL. Prior to that I worked on and Magnum wrote most of the initial code and recruited the core team and then prior to that I initiated the heat project in OpenStack. So I'm three for three. I'm projects on OpenStack and that's pretty successful. They've all gone through integration of the Big Ten. So I have quite a bit of experience with OpenStack as a result. Yeah, that's great. Gosh, it's wonderful to see the veterans moving on to new technologies and bringing more capabilities into OpenStack. Can you tell us a little bit about COLA? Yeah, so COLA is a tool for managing deployments. A lot of deployments tools just deploy and get you to day zero and then you're on your own. COLA is designed to get you to day zero in OpenStack deployment and then after day zero it's designed to upgrade and reconfigure your cloud now and we have more things planned for the future. But really we're focused on operating a cloud not deploying it. So deploying is the first step. Operating is the whole point of having a cloud in the first place. So that's what COLA is all about. Great. What were some of the hot topics that you and the team were talking about in Austin? COLA as at the current moment has a big gap and the big gap is that it doesn't deploy bare metal. We did talk about we had a big long session about having bare metal as an option and what we're going to do there is use Bifrost. We're going to integrate with Bifrost which is an ironic project. So we have the full from Pixie all the way to your deployed cloud on commodity hardware using commodity open source free software. That's pretty cool. Another hot topic was the idea of around COLA Kubernetes which is making COLA run in a Kubernetes environment. I think that work is realistically one to two years out before it reaches maturity. But it's just getting started. So that was there's a lot of excitement around Kubernetes and open stack integration. So we figured we want to participate in that rather than not participate. There are a lot of other things. One of the big points I'd like to point out is we had this big pain point session where we identified about 15 or 20 pain points that operators are running into and we are definitely going to resolve those pain points a cycle if we can if it's technically feasible in the time we have available. Wow that's a great session to run. So the following on to that are during the Newton cycle are there some specific user needs that you're going to be trying to target bringing the new features to or problems to resolving. Yeah so this is all around manageability and all of our features are designed in the future to make COLA more manageable. For example we will be adding a feature to add OSD. An OSD is a storage device or remove an OSD that's to remove a storage device will be adding stuff to add compute nodes and remove compute nodes and we'll be adding other things that make the cloud more operable. Right now clouds in general if you look at off-the-shelf deployment tools very not operable they're not operable at all. COLA is probably leading in this area because we have full upgrades that are atomic and we also have the ability to reconfigure running cloud from a single source. So those are really two fantastic operable features. I wait for our community to invent more. I think there'll be many more that are invented by our community not to the point to where we blow but to the point to where we serve all of the common needs that people commonly have. That sounds great. So what would you say are your top priorities for new features as you go into Newton? Well our number one priority right now is Ansible 2.0 and that is to move with the rest of the universe on software just like open stack moves every six months, stocker moves every two months, Ansible moves every about once a year. Once they move if you're stuck on an old version then there's no support, there's no maintenance and eventually you can't use the software so because of that we need to move forward with that. Then of course the other areas around all around manageability and making COLA more manageable by solving the pain points by kind of adding features that are necessary to make COLA more usable for operators and then finally we have some kind of future-looking R&D around COLA Kubernetes. When we create the community roadmap for open stack we have a couple of different themes that we use to organizing and group features together. They include scalability, resiliency, manageability, modularity and interoperability. As you look at where your team is going to be focused in the Newton cycle are there a couple of these themes that you think will be more important to the team? Definitely manageability is where we're going to spend our time. As long as we can keep up with the rest of the community that's the other part that we have to start out because we've got a lot of projects to deal with. Yeah there are a lot of projects in the Big 10 these days. There's about 50 or so last I counted and we've only containerized about 20 of them so we've got a lot of work to do. So continuing to expand the reach of COLA across the Big 10 is also an area that you and the team will be looking at? Absolutely yes and what about as we look out beyond Newton to Okada? Are there some things that you can see would be a focus in that cycle? That's more challenging because we run on kind of like a one month roadmap. We actually have done an 18 month roadmap but I think it's all kind of a guess. I would like to see us focus on stability long term, maintaining our z-stream branches long term. So we've committed to releasing a z-stream every 45 days. A z-stream is a maintenance release. So x dot y dot z the z is a z-stream. We were going to release a z-stream for all of our branches which are Liberty and Matak at this moment every 45 days so essentially for a cycle and what that'll do is add a lot of people to pick up bug fixes. So I'd like to focus on the resiliency and then of course those bug fixes and then kind of our future looking R&D work around COLA Kubernetes. Oh great. So I'm glad to see that you're looking at on some of these key themes. The stability is something we always hear from operators. That's great news. Well thanks a lot. I appreciate your time today and your thoughts on where COLA is going and Newton as well as Okada. Look forward to talking with you in the next cycle as well. Great. I'll be happy to speak.