 Good morning everybody. My name is Pete Chadwick. I'm Director of Product Manager for Cloud and Systems Management at SUSE. And with me today is Kershaw Madoff, Senior Technical Staff Member for IBM. What we'd like to talk to you about is how SUSE and IBM have been working together to pull main frames into cloud environments. So that I'll turn it over to Kershaw to start. Okay, very good. Good morning everybody. So I first wanted to just start about talking about our main frame environment and our latest deliverables that we had. Now the main frame has been around for well over 50 years and what I can tell you is we've been evolving for the last 50 years. You wouldn't expect the technology that you know in this in this type of day and age that has been able to survive that long. And the only reason why we've been able to survive that long is we have continued to improve and evolve ourselves making use of new technology including technology such as OpenStack. Last year our latest versions or our latest servers that we made available were known as Linux one servers and we have two different models for that the Linux one emperor a larger box and then a Linux one rock hopper. For those of you who were in Austin we actually had a Linux one rock hopper down on the show floor. We weren't able to get one here in Barcelona but the intention over here is to show you the fact that these main frames still continue to exist and continue to to provide work workloads to be able to go off and run workloads. This is going a little slower than usual. Okay so from a from a standpoint over here with Linux one we have what we call diagonal scale you know people are familiar with scale out technology people are familiar with scale up technology. Well with our environment over here on Linux one you can have both scale up and scale out. And this is what we call diagonal scaling and this is really important when we're talking about being able to go off and do things like extreme virtualization where you have a hypervisor partition that's built into the firmware to be able to go off and have complete isolation. You can go off and have bare metal 85 85 instances of bare metal instances and then from a virtualization standpoint over here you could have the most densest environment around with respect to virtual machines. We also have a very super elastic system over here being able to go off and combine both horizontal and vertical scaling and being able to non-disruptively add or remove resources to the guest environments underneath. The main frame as I mentioned has been evolving for quite some time. I just wanted to put up an icon chart over here of all the various different technologies that do run on our platform. Some of this newest stuff over here Node.js, PHP, Ruby, these are all technologies that run on our platform over here. In fact at the last OpenStack summit that we had in Austin we did a joint session with Ubuntu where we now have Ubuntu and Juju running on our platform. So again a whole bunch of new technologies that are able to be run on our environments. The big differentiator over here with our environment compared to for example the traditional x86 architecture is the fact that we have a much better data serving capabilities and it has to do with the underlying architecture that we have within our platform. I'm not going to go through the various different charts over here but what I want to just show up is that for various different databases that exist out there whether you're talking about MongoDB or whether you're talking about MariaDB you can have a lot more scale of running those applications on our platform compared to others. You get better throughput compared to you know x86 environments and you have the ability to be able to have improved performance running those workloads on our platform. Okay on our platform over here we have two different two different hypervisors. We have the KVM hypervisor that exists on our platform and we also have the ZVM hypervisor. KVM Israel is really new to our environment that's been available for about a year now and before that our own proprietary hypervisor that we've had is ZVM has existed for 40 plus years. Today's session we're going to focus mostly on the ZVM proprietary hypervisor that we've had over here and that's in fact exactly what SUSE has been providing support for within their SUSE OpenStack Cloud product. The two different hypervisors that I mentioned from a ZVM standpoint our drivers are available out in GitHub. They haven't been accepted yet into the community over here. We'll talk about that a little bit later over here. We're planning on trying to get that accepted into the OpenStack community next year in 2017 and the KVM drivers actually for our environment have already been accepted into the driver. They were part of the deliverables that came out earlier this year and you can also see them over here from the GitHub standpoint. Okay, let's now talk about, I'll turn it back over to you Pete over here to talk about the OpenStack support that they have within the SUSE OpenStack Cloud. So one of the things I think most everybody understands what a distribution is. It's a pre-packaged version of OpenStack that is, it comes with all the drivers, comes with all the necessary capabilities to actually stand up and run OpenStack easily. We believe that that's a much more efficient way for most organizations, especially enterprises, to download and deploy OpenStack as opposed to pulling it directly from trunk or maintaining that. We still use, distributions do use code from upstream and most of them are shipping fully open source tools. They're not, they're not adding in a lot of proprietary things. So in the case of, in the case of SUSE OpenStack Cloud, that's what we do. And if you look at how we've been working with system Z, we've been, as Kershaw says, we've been working with CVM to, to deploy that. Now one of the things we've actually been in the OpenStack business since 2012, which is when we shipped our first release, imaginatively called SUSE OpenStack Cloud 1. But back then, realistically, you could have, you know, prior to, this was based upon Essex, OpenStack Essex. Prior to that, realistically, if you didn't want to use, if you wanted to use a hypervisor, use KVM. We worked within the community to get Zen support. So this is kind of what the, a distribution looked like back then. You had Nova, you had Glance, you had Swift. Networking was built into Nova. So what was, you know, what became Cinder, what became Neutron were actually sitting inside the Nova code base. And then you had a pretty, pretty simplified set of choices for what you ran on. You know, this is kind of what it looks like now. This is, even this is not complete, but we pulled networking out, we pulled block support out. We have a bunch of kind of microservice type things on the right that we're pulling in. We've got new APIs in terms of heat and, and salometer. But the key point is, we now got a wide range of hypervisors to choose from. You can use N, you can use KVM, you can use Hyper-V, you can use ESX through, through V Center, and you can use EVM. And as, as Kershaw pointed out, you can use EKVM. We currently support, we at SUSE, support all of these in a single cloud environment. So if you want to have a mix of open source hypervisors and proprietary hypervisors, you can easily do that. And from your user's perspective, you've got a single, a single dashboard that you can, that you can manage all that environment from. So let me turn it back over to the Kershaw to talk a little bit about what the cloud manager appliance is. Sure. Okay, so with our ZVM hypervisor over here, we have something built into that, known as the cloud manager appliance. And essentially, the cloud manager appliance is an OpenStack controller and compute node that we ship as part of the hypervisor itself. This means that the same exact drivers that are, you know, part of the SUSE OpenStack cloud-sick stuff is also shipped inside of ZVM over here, if somebody wants to be able to go off and use the stuff that comes directly with the hypervisor itself. Under the covers over here, we have some, we have some technology over here inside of ZVM. We use XCAT technology over here to drive some of the underlying systems management API calls that we have within a ZVM environment itself. But this cloud manager appliance that we have as part of ZVM can very well be configured in multiple different modes. It can be configured in a pure controller mode environment, where the OpenStack controller and the compute nodes are running natively on the hardware itself, on the hypervisor itself. Or in the case of the SUSE OpenStack cloud stuff over here, it can be configured in minimum mode, where essentially we're just providing the underlying XCAT capability over here for SUSE OpenStack cloud to be able to go off and take advantage of. And given having said that, let me turn it back over to you, Pete, to talk about how you take advantage of that. Sure. So these are the capabilities that are available in CMA today based upon liberty. There's obviously some enhancements that are coming as Newton rolls out. But the key point is it's got support for some of the latest features. It runs on the most current versions of Red Hat Linux and SUSE Linux. So where SUSE has been in, as far as OpenStack goes, as I mentioned, we had our first release of SUSE OpenStack cloud in 2012. We actually joined the OpenStack community as a corporate contributor in 2011 and then participated actively in the creation of the OpenStack Foundation. Alan Clark, who was our director of community engagement, is the chairman of the OpenStack Foundation board. And then we've had a series of releases since then. We actually just announced yesterday SUSE OpenStack cloud seven, which is our Newton based release, which will be available at the end of this year, early 2017. And as I mentioned, we have what we call mixed hypervisor support. So you have, we kind of grayed out the control plane. So you can deploy a control plane currently on x86. So we're, you know, back to Kershaw's comments, we are running an MN mode. We're managing, we're managing to the ZVM environment. But we can use all of these different environments within a single cloud. So if you have an existing environment with system Z, or you're looking at pulling in Z into your environment to take advantage of the diagonal scalability that Kershaw was talking about, you can do that and still have it coexist with either existing VMware workloads, or as you're looking at moving into a KVM environment. Some of the key features of SUSE OpenStack cloud, you know, it's built on Newton. So we take advantage of a lot of the exciting enhancements to Newton. If you want to learn more about that, I remember the product working group, we have some sessions tomorrow that talks about, dives into a little more detail about the OpenStack roadmap and what we've been delivering in Newton and on. We do support the most recent version of Susan X Enterprise Server. One of the things that we've really been working on is non-disruptive upgrade. So now be upgrade from one release of SUSE OpenStack cloud to the next without taking down any running workloads. This is clearly something that we heard from customers as being a big inhibitor of being able to migrate to take advantage of newer and newer capabilities within OpenStack. We're also adding container support so you can now deliver containers of service. We're also, one of the things that we think we pioneered was deploying the control plane in a highly available state with SUSE OpenStack cloud six. We extended that to compute nodes so you could automatically recover compute nodes if there was a failure in your infrastructure. We're now extending that up to actually managing the virtual machines themselves. And then lastly, we're enhancing with SUSE OpenStack cloud seven. We're actually providing both managed from as well as managed to for CVM environments. And just as, you know, there's a lot of projects in OpenStack. This is just a quick overview of what we supported with SUSE OpenStack cloud six. And you can see we had kind of some, you know, decent coverage. We pick up a lot more. We also have a lot more things that we're doing in technology preview, which means that you can download the packages. We'll help you get them set up and configured. But we're not necessarily sure they're ready for full production environment yet. So in terms of how do you, how do you use EVM? You run CMA in minimum mode. We can support the network configurations that are named in the IBM documentation. You will be able to run either from X86 or directly on CVM for the control plane and but, but use, use EVM as a, as a compute note. There's a lot of documentation on how do you get CVM set up. And we would recommend anybody going down that path to start with that. Because you have to do a lot of configuration of the Z environment before you can install SUSE OpenStack cloud six. So that was kind of wrap up. Oh, sure. So I do want to mention, you know, from a customer standpoint, what we're seeing is, you know, most customers are moving to OpenStack environments initially on their X86 environments. For those customers who happen to have mainframes also as part of their infrastructure, they're looking at how can I move the mainframe also into this OpenStack environment. And when you have this environment of both X86 and the mainframe wanting to be able to often manage that with a single OpenStack distribution, that's where the SUSE OpenStack cloud product positions itself. You can go off and use that single product to be able to go off and manage both your X86 from an OpenStack standpoint, and also the mainframe environment to be able to go off and do both those. Again, Horizon UI will be able to go off and do that. Now our intention over here is to continue to upgrade the OpenStack capabilities that we continue to ship inside of the CVM. Lockstep with our partners over here as they continue to move up in OpenStack releases. So our cloud manager appliance that's currently based on Liberty is going to get upgraded to Newton. That's going to be happening sometime early next year. And, you know, as SUSE OpenStack Cloud 7, that's also based on Newton comes out, you know, both will be able to go off and participate together in those types of environments. So from a summary standpoint over here, the support that we have as part of the CVM is shipped as part of an integrated appliance with the community drivers. We do continue to plan on enhancing those or upgrading those as the new OpenStack releases become available. And we do continue to plan on working with partners such as SUSE over here who provide support from a CVM capability standpoint. So it's going to wrap up? Yeah, I guess that's it. Yeah, there's some hidden slides over here. So one of the things over here that maybe an ask that I have of all of you over here is the fact that many of you might be existing SUSE customers to take a look at the SUSE OpenStack cloud product itself to see how that could very well be used potentially with your mainframe environment. And then the other ask that I have of you is, you know, as I mentioned earlier, the CVM drivers right now are currently out in GitHub. They're not into the community yet. And, you know, we're trying to get that accepted into the community next year. We'd like to, you know, recruit or ask for any help that you may be able to provide us in trying to get those drivers accepted as part of the community over here. We're getting our CI environment all set up. We have all the necessary infrastructure in place to be able to go up and prove that we can become part of the community over here. But the fact of the matter is, you know, we see partners such as SUSE taking advantage and already downloading those drivers and incorporating in those products over here. So we do see viability from that standpoint and anything that you can do to be able to help us get those drivers accepted as part of the community next year, that would be, you know, one of the asks that we have of you. So that, are there any questions or comments from the audience? Are they looking to just get to maintain utilization and just continue to say, sir, as they have formal new technology? Are they looking to do something new with what was that for the mainframes? Well, we're seeing actually, you know, people who have mainframes today in their shop are doing it mostly for workload consolidation. You know, they have Oracle databases today that are being spread across their x86 servers. They want to bring those databases on to the mainframe environment to be able to have workload consolidation, consolidation of licenses, et cetera. So that's an existing workload that already exists. The move to OpenStack is more from a service delivery standpoint. How can I go off and make that service that I'm providing easier to my IT staff, easier for my end users for them to be able to go off and do that? Because, you know, today without OpenStack, the end user having the capability to be able to go off and start a virtual machine, Linux guest environment, for example, and then, you know, running an Oracle database inside there is mostly done by homegrown scripts. The move to OpenStack is to be able to help automate quite a lot of that, standardize a lot of that and remove the need for the IT staff to maintain those homegrown scripting environments. And that's pretty much what we see as well, is it's a combination of, you know, it really is about how do I provide a more flexible cell service interface into a Z environment so they can either do development test of existing applications or creating brand new applications. So it's pretty parallel to what we see on the x86 side as well. Good. Any other questions? Anything else? Okay. Well, thanks everybody for your attention and have a great rest of the summit. Thank you.