 Okay, so thanks more for your usual brevity and leaving me more than two minutes to get through my stuff And they put me after more that's worse So despite all of the evidence the contrary, I'm not actually Andrew Trossman and Andrew had to head home today Unfortunately, he's actually on an aeroplane. And so I took over from him. And I However still one of Andrew's colleagues who appears in the same organization and I'm going to take you through Andrew's material And I hope to be as quick as possible. Now. I know we don't have much time. I know my computer is not advancing Which is weird And here we go. So that was unexpected. So thank you Andrew Putting me in the spot. I'm not rebooting goodness Ready to go first Well, I get this fixed This is This plug-and-play no signal technology All right, let me try again. All right, so let's hope that this works now this thing is black, but I guess that's fine I Just want to start Where the previous Presentation has left off want to say hello. My name is Frank Schwichtmerk. I'm the technical product owner of IBM Cloud Manager with OpenStack We had a variation of the start before And what I was what I'm up to do now is I want to show you some of the implementations of what we have shown And you may realize some correlation between what we do a chart where in a demo We were discussing a little bit if you should show it to your life or not life And while life has some authenticity Some of the operation take quite a while and the video has the advantage that we can time-trink that to value your time I really invite you to come to our booth And I go through every step that you would like to see live So I will straight after this I will go to the IBM booth and I would be happy to see you and we can talk about it We can stop and drill a little deeper than we can hear if that's okay with you okay so unfortunately at the video and This is IBM Cloud Manager with OpenStack as we said it's a full OpenStack Juno code base that allows you to build a cloud on a premise Of your choice. So not only local premises. We have Customers that have requirements for multi-premise Be it for multi-location be it for scalability from an architecture perspective They're multi. There are many reasons why you may want to have multi-region and a hybrid like approach So the one thing that the team spent quite a bit of time implementing is The ability to serve multi architecture multi hypervisor environments I don't know any other company that presses that to the degree like we do to an example here We have a rest region here, which does OpenStack on x86 We have the first region which is pretty unique. It's a legacy VMware situation, let's me slow that down that allows your coexistence between Traditional VMware deployment that you may want to drive or you used to drive with VCAC or other means And you want to have your entry to OpenStack cloud management You can use this self-service user interface that we add as optional value add and basically get a visual integration Single pane of glass so you see your VMware stuff and you see your open you're emerging OpenStack ecosystem Being combined in the one thing single pane of glass, which I think is pretty powerful In addition we have OpenStack on ZVM so things that used to be not very accessible for the that say younger talents that may not have The skills of mainframe can now use the the the benefits from quality of service the of the big iron Let's say of the of the mainframe systems The east region is quite interesting as well because the east region is quite busy because we have power KVM Which is a very symmetrical stack with lip word and KVM, but on power eight boxes beautiful Wonderful architecture robust We like it very much. It's little Indian based from a technical perspective However, as Mo pointed out, there's a rich set of systems of a record solutions available in the power VM base with AIX and IBM I That our customers would like to introduce with from a system of record to system of engagement kind of solutions and combine the two So the power we see environment that enables that is one of our targets as well and implemented here as well In addition, we are now a third release to deliver OpenStack for hyper V which Offers access to an on power hypervisor to to VM where at a very attractive cost base With very compelling features like shared nothing live migration to just name a few In addition, we have something which I'm not sure if you're familiar with that IBM elastic storage Which is you may have heard the GPFS term before which is very popular in the high performance computing world as a very robust high performance low latency file system and not only you can use that for Glance our easy shared file system to create mobility in KVM But you can use it for Cinder and for Swift as well Extremely scalable. So the Cinder case for example is if the hypervisor needs a volume, it doesn't have to do ice-card. It screams basically just using GPFS based elastic storage In addition, electric storage is so flexible that you can actually create a highly scalable really large cloud In not requiring a storage server. You can utilize local storage And and basically create a very high throughput real good scaling environment So if you're if you have needs like that, you probably should validate if that's true for you as well All right, I Have a horrible cold. I apologize So the the the self service user interface Is giving you access to open stack region and to that v-center and we map some of the concepts So for example the project concept that we offer maps to open stack tenants No surprise here. Well, we add value add here it is a feature that we add here that you can on the region base or on the on the project base you can set Policies like exploration. So you can for example say I want my instances in this project to expire after five days and I know people won't like it. So they probably want to have an extension So the policy allows you to specify the extension period maximum that you allow So let's put that in so five days Let's say we allow 25 days extension because they're friendly and because we know people ignore all the warnings And they don't believe that we mean it. We probably should add two days before we actually wipe the resources So after the the instances get torn down. We don't quite delete the instances yet So so we can still do some recovery and we do have email notification To to have appropriate warning and that is all highly customizable Another thing is just imagine you're hosting an SAP solution You have these very empowering self-service user interface and let's say the user say I wonder what the lead does and Say it's gone. That would be bad So many of our customers really really appreciate the fact that we have the ability to very Individually select on and off which operations require approval and the approval goes into the project admin and allows some Securing of these empowered use cases So let's talk a little bit about another feature So you see images volume and instances all have a visual federation of all the regions being connected So let me stop here because that's actually quite interesting a x is coming from power VC This is coming from VMware Whatever you see it comes from all sorts of spaces But as a self-service user you don't care you can focus on your work You can focus on your innovation and the the complexity of Implementation is an implementation detail to you that you can safely ignore in many cases If not the cloud admin in the service user interface can expose the things that may matter to you In the way how they configure it So we're trying to do a solid mix here between empowerment and simplification Without dumbing down the users either and the customer can really manage and adjust that So if you look at the instances you see a lot So obviously we will allow scoping this down to the region as well You can scope it down on the project that you're working on you can scope it down on the user So as a user you may be in multiple projects for example, so that doesn't make sense so let's look let's say at the West region for now and we have five tempest test instances running and Let's imagine that we want to instantiate five more so we go into the images tab and and After a while I decide to do that Sorry for that little pause here and And we can basically showcase that with virtually no training really everybody can do this It's a very low toss very low education requirement for the end user So let's go into the tempest image that we have in this region the West region click on deploy and The person who imported this image has made it very easy So everything is basically pre configured and the only thing that you can really set here is a name Which which project it goes in we can overwrite the exploration and we can say how many instances of this image we want So fire and go so now let's say How about horizon We obviously feature and serve a full feature horizon that enables all the open stack features so you can select your Tenant you can select your region which I should have said West region here so that I ran out of time to do that properly We have The host aggregates and we use that and we define let's say four nodes C and four to C and six here And this is important for the next thing. I'm showing you One of the value adds is that we have platform resource scheduler that offers placement policies We talked about that briefly Which is the initial placement and the runtime placement? Initial placement is the what happens if you instantiate an instance the runtime placement is a replacement Basically dynamically the the mix of instances is changing and you have a policy What what the what the scheduler should do about that basically by the means of life migration So if you look into that You see that we have multiple different policies. That's actually quite powerful I've elected striping here, which means evenly balance, but you could do CPU loot balance memory load balance Striping I said packaging policy the Packing and CPU utilization packing would be an example if you have like a large data center and you are Like a emerging cloud and you don't really quite have the utilization yet Maybe you want to pack them all together and in a few hosts and be able to switch off a few hosts to conserve energy Or maybe this is a power cloud and IBM has this nice feature that you can enable hardware licenses as a software and you you may not want to pay that so you basically you scale down the Virtual hardware consumption that you have and just use it on a subset. So you are very flexible here to play with your resources Okay, remember we had five instances running in West region and I instantiated five more We have striping Defined as the policy for for placement. So let's look at the hypervisor view here where we see the instances and over time They get obviously a condenser time here. So I'm honest They better to get provisioned And we expect an even distribution and you will not to be surprised to hear that actually works And we get a 3 3 4 whatever combination out of 10 instances All right, so let's go down on the command line to make some other cases So you see here We have a hypervisor status list where you see the ha status and the maintenance status of our three nodes that we care About right now in our scope Let's say we have a plant maintenance activity Maybe we need to upgrade the microcode of the ethernet adapter or the yearly power maintenance Thing comes along and you need to do power maintenance on the power line So you want to be able to take a note down and you want to do that with the least amount of Impact to your work. So wouldn't it be nice to have a single command to say you are on maintenance and get them all out of the way That's what you have you have a command to do exactly that You say migrate all of that note and put it on maintenance mode. That's all you do done so it enters a maintenance mode you can see that here and You can in a short moment of time. So you see it entering here CN4 And you can see in horizon in the hypervisor view that this actually happens So the the former more instances in CN04 get gradually moved over to the remaining Once I just just mentioned I should mention something The reason why this is happening relatively slowly is because there's a policy in place how many Migrations you allow at a certain amount of time to not create too much load through all this smartness so so so there is a Slowdown Basically intentionally created to basically be less Impacting in terms of of your other aproletic quality of service So don't be surprised by that that it takes like two a few seconds to to to kick that in so at some point you have five five This note now is not have not having any instances You see in a short moment of time that our maintenance status is off on the screen is not all that great But believe me on Which means you can turn off the note and you can perform your maintenance as planned So now let's say tipsy Franks comes along visit the data center And there was this one ethernet cable that he didn't see and he trips over it and gets out of the system Which is a catastrophic event. So all the instances on that host would be disconnected all right, and Little technology here things so you if you don't understand the details ignore that what I basically do here is I Showcase that you can get onto the note. You can virtually rip out the ethernet and Through the h a policy that we have the platform resource schedule as part of the product realizes that They have to remote restart those instances on remaining nodes in the host aggregate and Let's go through that Here I just show okay. That's the that's the interface We are still okay as a We can define the policies obviously with a remote restart But this let's ignore that detail if you have interest. Let's talk about that at the booth I think that would go so far here, but if you now look at Horizon you see we have the five instances on one note and five instances on other note because we have striping remember So now go to the note. We have a serial over land console We really have five instances running right now, and I if config down the interface Which is the equivalent of ripping the ethernet cable? So notice gone boom I look at Nova and Nova has a little polling going on so at first it doesn't see it a second later It sees whoops my Nova compute node went away and says we're down Now platform resource scheduler comes in and quickly enters a rebuilding stage Meaning applying the h a policy of re of remote restart and you can see here If you look closely that the five instances here get restarted here Mission accomplished you could say Almost so here. I want to be honest with you IBM has a lot of cultural change here, and we're very open with continuous delivery and Telling you what we are doing Gathering your feedback and so on this is good for the catastrophic case. It's excellent So the thing burns up in smoke you have a remote restart your quality of service is higher for intermittent errors We're not quite done yet. So we will at some point have to introduce effect a new and additional feature named Shielding where those instances on that note cannot possibly get back on the network again So so we do not create conflicts So we're not quite done yet on GA time on 4-2, but we will work on that very hard and it will come very soon But just as an example of transparency, we give you a feature that's already useful We say what the limitations are and we're already hinting that we're working on it to improve that just as a transparency proof point Okay, that was the the the short rundown of the demo I have Actually another plug to make If I had a network that would be nice And Since URLs specifically IBM URLs can be horrible and not memorizable. We created a shortcut for you I can't see here. It's a moment This one If you go to this link IBM dot bi Z business IBM cloud manager with OpenStack So Cmwo you get to our home page so to speak where we offer our Product downloads 90 days for free. So you get the ice house release for one now You get the continuous delivery version of the 4-2 release which offers you Juno now Everything I showed you here is done with the preview release Identical code. So if you want to replay that at home and your own premise be our guest It's free for everyone for 90 days free to use. We have other trials if you say, you know I really do not have a premise I don't have much time, but I would like to experiment a little bit with the user interface of this self service I wonder how this approval thing works and so on and so on we do have the flow network and We have a couple of trials. We have a hosted trial for five days You can get for free so you get a hosted instance You can play with the self service user interface which has like a mock backend and we have the other trials Which is basically The in the complete install code of our product which you can apply the fix-packs on and you should To to basically experiment on a premise on the hardware of your choice with the architecture of your choice to get things going And in the same way the very last one I want to Make the point you get the version here, which is going to be our product version But it's basically in our agile development cycle at the end of each development cycle. We've republished Extremely transparent how it's supposed to be One more thing Well, I think we should keep it simple Let's put it this way. I welcome you to come back to the IBM booth and let's talk about Sorry, let's talk about the Details that you may be interested in that we don't have time to cover right now because this is a 10-minute overview There is a lot more to to the open stack possibilities that you can do the architecture You can do and some of the edge of the patterns that may be interesting to you I Want to mention it was said that this is for local premise only Remember, this is open stack. We do offer hybrid with with keystone as well with the base building Blocks of cloud manager was open stack. So you can build your your your distributed multi-region cloud like I've demoed here yourself Any questions? I talked to you in the All right Thank you very much for your yeah Okay, so here's the deal IBM works as a big IBM team The IBM cloud manager folks and the distributed open stack team within IBM Produce with the community open stack The cloud manager team is packaging that up and creates the product and do the does the testing and does a lot of work to Harden that to make it usable for customers The cloud orchestrator folks use the very assets the very same RPMs and most of the chef recipes I believe as well and basically build their product out of that So you have a good compatibility because there's really the same binaries are using and guess what IBM cloud open stack services is built from Same thing. So you you're talking about a family of offerings here Which really fit well together because they're built from the same components What I haven't mentioned which is probably a given that our deployment server is based on chef We do extensive work with the community. There's one sense. I would like to stress we do not fork meaning if we need a change in the community recipe We propose a patch to the stack forge project that covers that cookbook We work with the community till gets gets accepted and pull a new version of the cookbook down So you will not have the situation that's the stuff that we shipping in our product is something which is Dramatically different than the community version because clearly we sourced it from the community version We take great pains in that which is not so much a technical pain But every time we have to legally approve it. So there is some work involved All right Sorry for talking so long Hello, okay So what I wanted to do is pick up off what Frank was talking about with CM the CMO product Which was more around you know what you do on your on-premise you take your sought the CMO software the distribution And you put it on to your own premise what I'm gonna be talking to you about is New offering we just announced last week. It's called IBM cloud open-stack services My name is Ashish Patel. I'm the founding product manager for this offering So what I want to do is I want to spend the next 10 minutes and I want to make sure I leave time for Scott here To complete our hybrid cloud story here But I'm gonna give you a quick short demo and I'm gonna start out by just telling you about what is this new offering Okay, and so the reason this picks up off of where Frank left is like I mentioned This is on-premise what Frank was talking about this is the situation where you know You're as a business you're considering well I need to get an OPEX model and one thing that I think resonated very well for me at the The keynote two days ago was when there was a panel of people from different companies And they were talking about their pain points around how to manage open-stack how to have the right skills in their organization to actually manage it and Deploy it and scale it and operate it the whole point of this offering is that it's off-premise Hosted with software any of our software data centers around the world and IBM will actually manage it for you So now in this situation as a business what you are doing is you're using the power of open-stack you get the dashboard You get the APIs, but you don't have to worry about managing and scaling it and operating it, right? You will now become more focused on your applications and the workloads You're trying to put on top of the cloud you become more focused on reaching out to your customers You can also now you know you have an option for something that's off-premise And if you are already using for example some of the technologies that Frank talked about on your premise You get interoperability between them because hey, you know the API's and the interface are standard amongst both of those So you can start connecting some of your assets that are maybe more need to be more highly secure on your premise So interoperability is one piece the the operations management's another piece In and now you can focus on your applications the other thing is around speed and agility you hear this comment all the time You know we want to get faster and faster and faster with this offering because it's a service offering And we sell it on a monthly subscription basis You can tell us anytime during the month I want more servers or I want more storage and we just add it into your existing capacity pool And it'll show up into your dashboard and a dashboard that you're already familiar with So what that's what I want to do and take you through this This particular offering a couple things. I haven't mentioned about it yet You see the details at the bottom here on the landing page here is that you know, this is purely a private cloud This is another really big piece of value add that we think resonates very well with our enterprise customers Not only are you having the option to choose from our 16 global data centers across the world with our wide footprint? Because it's hosted at software, but all of the hardware is dedicated to you You don't have to worry about your nice neighbors. You get predictable performance You know exactly the hardware spec that's underneath full transparency there Security is now, you know completely under your realm. So you really get to benefit from a lot of that The last thing I want to touch on is just the management piece So I said this was managed by IBM I'm going to show you in a second what is managed But I just want to emphasize that this is backed by a 99.95 percent SLA And let me actually just show you what what that means and what that covers So in addition to the SLA it also includes 24 by 7 support So this here is just a quick little architectural diagram just to kind of put in context What's behind the scene of what I'm about to show you so what you can see here is the in the overall system There's really four components here There's a management system which is in the dark blue at the top and what you get in the service So that you pay on a monthly subscription is you get three way highly available open-stack controllers This is again fully managed by IBM You don't have to worry about anything in it and this actually runs the same distribution that Frank was talking about So the same consistent experience that you have on your premise now you have it off-premise Okay, you also have a secure way of connecting. There's a VPN and firewall devices and in fact there's they're actually diverse So we actually segregate out where your customer data traffic will come from versus where IBM will manage the environment from Enterprises consistently ask us for diverse data paths and this is this is our way of providing that And so lastly, you know these three pieces here are fully managed by IBM The other if you look through across each of these services here, there's compute nodes We have a block storage service and an object storage service again Remember all the hardware here is dedicated to you you pick as much as you want And and IBM will met will manage the hardware and in this case a compute will actually also manage the hypervisor That sits on top of it in this case it's a KVM hypervisor and the SLA covers everything that IBM manages so it's not just you know These dark blue management system environments It's also the hypervisor on the compute compute and nodes for example as well as our block storage is backed by It's a just a bunch of disks, but it's running seph as well as object source also a JBODs and it's it's running a Swift there So IBM will also manage all that piece. You can then just focus on on your applications above So that's essentially the layout around the architecture What I'm gonna do is show you this website Just so you have some reference after you leave this this session And I'm happy to talk to you further if you hit this URL open dot IBM cloud calm It'll take you to our landing page This is publicly accessible on the internet and all the everything I'm about to show you from here on is on the internet Okay, so I just want to be very transparent about that so The way that you connect into this environment like I said there was a secure set of VPN and firewall services, right? So what you can see here is that on my on my environment here? Actually, I'll just show you the actual Connection information. I'm just VPNed into my environment and This is this is we what we do is when you get the service we issue a bunch of certificates SSL VPN certificates you can use open VPN on any of the platforms and just connect in like I have here once you do that you could hit the landing page and You can go to our dashboard So this is our dashboard. I'm just gonna zoom it up a little bit so you can you can see it a little bit better But when you're not connected into the environment, you know, you try to connect the dashboard It won't work because you can't connect over the internet. You need to have the VPN connected So when I connect into the dashboard here, I can choose to embed it into the screen or I can just pop up a new window I typically like to pop up a new window What you'll get here is you'll get the horizon dashboard that you're already used to it's just skinned with IBM information Our logos and our colors and things like that and Just wait for that to load And then okay, so once you're in here You'll and you'll know that you're in the service because it'll clearly say IBM cloud open stack services at the top You see an overview of you know the typical thing that you're already familiar with you know What's your capacity? What's available to you and so on and so forth what I want to emphasize here is That not only in this situation in this service Do we take the IBM distribution of open stack? But we also have made a number of enhancements on top of that that are only available via the service Okay, so that's what I want to kind of go through so go through right now Let me I'm just gonna I have some Preloaded tabs here with the same access. I just did it. Well, it was logging in here so what I'm going to show you here is Around our rules one of the things we do is we try to give our customers really good defaults So the two that you just saw me check off here This is what one of the things that we consider good defaults because one of the experiences we've when we were talking with developers We were originally Formulating the product was get open stacks great But when I install it I have to do a lot of configuration things that I would expect that to be there by default So when if you're familiar with open stack you you set it up You get an admin account to it But that administrator can do anything that they want within the open stack environment So what we've done and Frank, you know, this is also picking up what Frank was talking about projects and tenants What we do is we give you one cloud admin Role so you can add as many users you want in here, but this cloud admin You know, they think of them as your central IT group who's managing your cloud Okay, then there's project admins. So just like Frank showed each tenant You can associate different project administrators there as well but what you can also do in here when when I if I was to actually go in and show you the policies is You can actually control how much of your overall capacity in this hosted environment is actually available for a particular project So it kind of helps with capacity of planning, right? If you have an environment Let's say that can have 500 VMs You may want to give a project who's only needs a hundred VMs access to only consume that much amount of capacity, right? So you can you can control your administration here So we think this is again like I'm saying a good default you can get started just right out of the box once you get access to the environment Another thing that we've done is I'm just gonna flip to this next tab is is around flavors So one thing that resonated very well with the with the cut We surveyed like 500 workloads of all different types of industries when we were coming up with this But what one of the things that we constantly heard is I don't understand what tiny means like as an operator You know if you're familiar with open stack, you might right? But when you're working with a lot of variety of workloads in a larger company Tiny small medium extra learn doesn't really resonate that well with the workload that they're ultimately trying to board on to the environment So what we've done is we created two classes of out-of-the-box flavors that are loaded in here by default What you see up here from big machine down to the high memory and high CPU in the middle here These are what we call our specialized workloads And they're really easy to remember by the name because the first number is basically the number of eCPUs The second number is the amount of RAM right so these are just out-of-the-box defaults And then we've also created a second category of just standard and what we call standard and super but these are common workloads So if you're you know a company that runs a lot of VMs, you know for your web workload You're probably gonna use something like this if you maybe have something more specialized where you're gonna run your workloads You're probably gonna pick something at the top set of the list But the point here is as good defaults you get out of the box ready to go so we've done a lot of configuration there Last I want to just kind of show you what we've done in terms of images So in our model here, you can BYOI or BYOL Okay, let me explain what that means so be bring your own image is the first piece So you might see a lot of you know cryptic names at the top here So these are just other users are in our environment who have brought their own image in and this is typically you know They in their enterprise, you know now they're thinking of adopting a private cloud that's hosted somewhere else So they have imported their images maybe from their existing KVM or their Zen environments or Their VM where environments just using standard cloud in it and we give you the full documentation on how to do this But you can import your own image or what we give you again good defaults out of the box from that you see listed here Is we just give you the image we give you a vanilla image that IBM manages and maintains and patches keeps up to date And you get the image you can provision your VM using the image and bring your own license in to activate it and off You go okay, so you don't have you know again think about the model where okay I was gonna go set up my own old input stack I have to get all my own images and set that all up And I'm taking a lot of time to do all that so so that's been one thing These are the three major things that I like to talk about when I talk about the good the good Defaults that we give you around images flavors and the rules just out of the box One of one of the other major things that I just want to touch on this is probably our next biggest Enhancement other than you know taking software hardware and packaging it up It is a service and then giving you open stack and managing it is What we've done around our network okay, so what we've actually done is we've taken our software to find networking technology And we've integrated it into the open stack That's running on this environment So you don't get this if you're gonna go on to your own premise or if you're gonna you know For example take an open stack distribution you have to get the hardware You have to get go to an SDN partner get their software put it on top You have to integrate it with the network fabric that you have in your data center Test it all do all the driver work that you have to do there you have to do a lot of work there right So the whole point here is that out of the box You get our software to find networking and we've done all the integration and testing for you in a cloud environment So as you scale up and you add more and more nodes Whether it's compute or storage it just builds on top of that right so the whole point here is that you can create now you know Different lines of network whether you want to have virtual machines that are accessible on the public network Or what we have done is created a network here That's called SL private which is if you're already consuming other software services You can connect into the private network here and keep that separate from your other workloads You can also think of you know as an enterprise each of these network lines being either lines of businesses Projects or you know a tier in your application could be the front end the middle tier or the back end So however you want to carve it up and our software defined networking can actually allow you to scale up to thousands and thousands of these These networks so and you know, this is all out of the box experience. So That's what I want to mention with software defined networking The last thing I want to talk about is and just mention I'm gonna let Scott talk more about this in more detail Is the example that he's gonna show is is more around our hybrid cloud. So Well, I'll let him talk about it in more detail like I said But I just want to show you that the instances that are running here You're gonna see this name called trade application. What we've actually done here is we've actually Use this hosted environment this hosted service as essentially a target of resource So from your premise you can actually drive this you this hosted environment You know because it has API you can drive it from your on-premise environment So it just becomes an extension of your data center now and you can deploy applications into that environment I'm gonna let Scott take more detail in that but just to prove just to show and prove that when he's gonna show you I'm deploying onto this thing called I cause this is the actual environment and these are the actual VMs that got deployed Okay So let me just kind of wrap up a couple more points here is that again? Everything's publicly available. This is our full documentation page We actually have a dedicated team that helps us write a lot of the tutorials So if you want to import your own image it tells you exactly what commands you have to run to run cloud in it and Package up your VM and where to upload it and all these types of details. So all of our documentation is listed on this page We also have a support page as well So with this service you get 24 by 7 support by a telephone and ticket And we're obviously enhancing to eventually bring things like live chat and and other forms of support as well And lastly, you know, we just announced this offering and it's listed on the IBM marketplace So if I actually click the buy now button, I'll show you just to prove it It actually just goes to the marketplace page And and you you could you could buy it from here. So again, this is a monthly service subscription And you can scale you can expand and shrink the environment based on your workload needs And this is another way to consume our open-stack service or open-stack technology But now in a service form, right? So when you look at everything and you look at your environment and you look at what I have on-premise and now what I can Have off-premise, you know anywhere in the world with our global footprint. This is the option that you would have there With that I'm gonna wrap up and I'll just pass it on to Scott here To talk about trying again. Yeah, one week later Thank you. So sorry about the hiccup earlier on. Let's see if we can get this working this time. I've rebooted so Hopefully there we go Let me jump straight in here. I know we've got to go quite quickly and I'm gonna talk about the IBM cloud orchestrator It's Just like the other technologies we've heard about it's built on open-stack and uses open-stack extensively I'm going to give you just a quick overview of the the architecture here And I'm very high-level then I'm going to jump into demos. I've got two demos And we've only got a very short period of time So I'm going to go very very fast But hopefully if you have questions, please contact me offline or come down to the booth and They'll get a hold of me So we've got you can think of the architects. I usually describe it in being three layers We have these kind of infrastructure as a service layer. That's open-stack I'm not going to talk to you about that because explaining open-stack to this audience would be Futile waste of all of our energy, you know this stuff as well as I do if not better We will be to get to our private clouds through open-stack We have an extension which lets us get also to the public clouds as you can see Amazon software for now More to come Moving up into the next layer. Now, you know, this open-stack allows us to provision and configure compute resource network resource storage and so on and so forth You know a lot of our customers are a large enterprise customers that are deploying large business applications And that business applications don't run on like one VM or a couple of VMs I mean basically you can describe the business applications as being you know a topology of a number of different nodes You probably have some web servers and web application servers database servers. They're all connecting to each other They have to talk to each other. They have to scale out perhaps have to be clustered for resilience high availability so we need to find a way to be able to represent a business application in order that our end user can actually just go in and Provision that you know with one click or you know with a couple of clicks Without having to do all of the time together of the different pieces from the individual elements from down here And I'm going to show you a demo that then describes, you know, how the patterns work and how that relates to the hybrid cloud stuff That we just talked about Going even you know one level further up though I mean even if you can you can deploy your business application and have it all configured automatically, you know with a couple of clicks of the button With you know some of our large enterprise customers, you know, that's only one small piece of deploying a new service I mean you still have to do change control. We still have to get the approvals We have to deploy the application. Of course, we have to do asset management. Maybe update a cmdb We have to monitor it We have to make sure it's compliant for the license and security point of view We have to do backup and restore so the business process part here the it business process Includes the deployment of the compute infrastructure, but you know, that's it's much more much more complicated than that So we provide also this layer which allows you to automate the process end-to-end So you're not just saving time and resource here or here you're getting across a whole ito process or whatever process you happen to use So let me just describe the demo really quick and then we'll jump into it We've got three different data centers We've got one on-premise here and in North Carolina one on-premise run by cloud manager with OpenStack that Frank spoke about earlier And we've got the software hosted site here and what we're going to do is As we said earlier, we're going to deploy an application day-trader For some reason the example applications always happen to be stock buying applications and so on But we're going to deploy it in that hybrid mode So we're going to put the the application server cluster here hosted On our hosted site and we're going to put the database cluster on an on-premise site Okay, this is a demo is actually recorded as well. I didn't I wasn't able to we didn't want to risk you know working given that the The Wi-Fi here was kind of flaky yesterday, so we recorded it quickly So let's see that now I'm having resolution problems. I hope that doesn't get in the way of See the demo. Okay, so that's just a picture. We just saw it and now we're going to dive into the product itself usual login Don't know why we don't skip this thing in our demos The login piece isn't necessarily that interesting. So here we go So this is our son IBM cloud orchestrator as you can this is the the the dashboard the initial page that the user can see There's got some metrics about the stuff, you know, they've the resources that users actually used at the moment so a number of VMs what state they're in how much memory how much in the Storage and CPU etc. We have a number of different regions as you saw We have different open stacks and we can actually go in and change and have a look You see the numbers changing down here as you go in and look at each individual region because the usage will be We would different in each of those regions clearly So now we're going to go up and in actually Get the service. So this is a self service catalog. There's a number of menu items We want to deploy some cloud services and specifically we want to deploy the application So click it in that that brings us to the to this page Okay, we're associating that an instance name We need to have to know how to call the thing if you look here We can see the environment. It's maybe difficult to see that ask you where you want to have your database server Whether you want to on-premise or off-premise and the same with the application server And the application server is on software as we saw earlier in the diagram The database server is on-premise on a private cloud So we requested that the deployment of this This application let's go and see what's going on so have a look at the status of the instances that we've created and As you can see up here, we've got the little I call and tell us that the database service still deploying But you know, we have some pieces of the application that have already been deployed So clicking on that we can go and see what's What's going on there in the application part of the deployment But information about the application and the number of virtual machines that being deployed we see this too We have some information about the usage Let's go down tells you Version numbers all this the usual stuff you'd expect to see the copied IP address so we can go and actually log on to the application And then demonstrate that it's actually working If we do now, this is probably the least interesting part So I'll actually skip ahead as soon as we go on just because it's an interesting application But it's probably not the most relevant piece of the demo for you guys so This demo clicks around here a little bit and shows you how you can actually use application. Let me just jump forward a little bit there we go so talked about the The patterns what did patterns actually look like I mean we actually deployed a pattern, but you haven't seen what it looks like and This just show you very very briefly because we could spend hours here But very very briefly what it would look like to do to define to build an application pattern using using the cloud orchestrator product I'm still missing some of the screen, which is unfortunate But like I said before so we've got a different number of different nodes in the topology here We have three different nodes and as you can see he's dragging drop and you move them about connect them together and so on and so forth each of the nodes has a scaling policy now scaling policy means if this is a cluster then you know when certain When the workload gets beyond a certain level I want to scale it out increase the number of nodes in the cluster when the workload falls back down I can reduce it if you when this dialogue disappears if you look up at the top left-hand corner here There's a number one to ten I mean you can start with one node and scale up to ten minimum one scale up to ten depending on the Workload that we see so CPU memory and so on and so forth one other thing I'd like to mention is the Stuff we're going to see on the left-hand side now We need to put automation onto these nodes So when we add a new node that the certain automation actions that we need to take now These actions are basically over here as assets. These are scripts. It could be chef or whatever your favorite Scripting languages. We just drag them onto the node So now when we deploy an extra instance of this node or we take it away depending on policy Then you know this automation will run and we can just add dragging dropping like that. We can add You know different assets different automation Assets on to each individual node. We do supply a bunch of those assets out of the box and we also have a An ecosystem that we resupply Assets on this website, which is the cloud marketplace. We have scripts We have the patterns themselves and we even have some of the higher level business process stuff They will talk about very very shortly and we also want our partners to do that We have a lot of assets that have been built by by partners and you know, hopefully Also with our customers in the future The idea here is to build up think about like an app store that you would access from your phone But you know, this is for enterprise the cloud deployments So that was the first piece. Remember, there was the three pieces in the architecture You know had to open stack with the pattern stuff and the hybrid stuff and we also the higher level business process IT business process automation. I want to show you what we're doing with the third layer now This is a slightly different tack. So we're trying to make it as easy as possible for our customers to be able to Define automation like I said before it's not just about provisioning computer resource We also have to have to manage that So the application developers just want to get their machines and get get to work on them But you know, the operations guys want to be able to manage them and keep them secure and make sure it's compliant and so on But we have to make it easy for those guys to do it And we have to make it You have to do it in such a way that we don't slow down the developers. So this is What we're doing in the moment. We're building wizards to allow the We're building wizards to allow the Administrators the people to build the service offerings to be able to inject management capabilities into the automation without getting in the way of the Application developers this we're actually building a new service here. So we're giving it a name It's it's a service to deploy managed middleware We're giving a name the description the usual stuff then you can choose the category now depending on the category of Middleware application or whatever you want to build the choices the menu choices in the wizards and the dialogues in the next Pains the next few panels actually change So we're trying to remove all the clutter and just making you know providing the Choices in the decision making only two things which are relevant for that specific goal We're choosing the type of middleware in this situation. This case. We've got a DB to for Linux. We love DB to And it was in the deployment options gold silver bronze and that could mean whatever you want memory size CPU size, whatever you like and there's a little check box here Which we didn't check if I check this then that choice would be hidden from the user when they actually use the offering In this case we've left as a default We're choosing the management capabilities that we're going to inject automatically We've driven back up in restore monitoring We've chosen ability to add additional volumes today to the database server Which is going to be provisioned in this case you've chosen not to use security more compliance. Maybe that's because it's a development system This is the backup in restore configuration We have to add some some parameters in here in this case is the location of the agent Which we're going to automatically install when we provision a system and by the IP address so that when it comes up It's activated it connects automatically to the to the storage backup system and just works out of the box We have to we define a few options the frequency of the backups and you hope we're actually going to backup And I just get set up for the for the user and that gets configured automatically when we install our agent Monitoring similarly we put in the location of the monitoring agent the IP address of the server to connect and just work out of the box so this is Additional volumes and this is a possibility to add an additional volume to the server again I've not clicked the hide the configuration option So this will become a default for the for the user and he can change it if he wants down here. We have slightly more technical Operations and we're actually checking this box So that you know the administrator puts these in and doesn't allow the user to change them later that way won't do any damage by choosing the wrong option or Take a more expensive expensive option option and you know what he's allowed to do or whatever So that's it created. I mean really in less than five minutes. We've created a brand new Offering in the service catalog. We've chosen the management options that we want to set it up. So they just automatically configured And so that the user now we'll see just in a second. We'll go and see what the end user point of view with the vision Of this is we're now logging on is the end user We've got a boring login panel again Which we'll get through real quick. Hopefully And we're almost done So Judith lugs in She's going off now to the self-service catalog. She's going to look for the db2 option that we just created And it's in a search field, which is about here But you can't see that it is we found the db2 Entry this is what we just created. It wasn't there before She's going to get some options She has to choose an instance name because we need to know what to call it here We have the flavor and remember the gold silver and bronze We chose silver as a default, but we chose not to hide it so the user can change it Judith's got some money this this month. So she's chosen gold And she wanted to do the better option and this is there the volumes what size do you want? We chose two We didn't check that option. She's bumping it up to six But she doesn't see any of the other options whatsoever They've all been hidden because they've been pre-compiled by the administrator and it was chosen to hide them That's it. I know it was extremely quick, but only ten minutes and Like I say, please get in touch with me or Andrew Trossman Or come come Dennis easy the booth and folks will find me if you want some questions I'll be here until Friday morning Or grab me now, whatever you prefer Thank you very much