 So welcome everybody Good that you took the long road to this room We were we provided you with some food. So you needed maybe that for for the way back and So it's good to have you here People today or yesterday or the day before ask me actually two questions and that was not the question about red head, of course But it was about why are you wearing these red head t-shirts? So what's the So what's the red bull t-shirts about so so what is so if you look at formerly one It's really an innovative sports So it's all about being the first one on the finish line. It's all about reliability so there are a lot of analogies that you can use for it and Of course, we are very proud because we have a very good Dutch driver That's Max Verstappen and we are a Dutch company. So we also a little bit proud on that So that's why we use this red bull t-shirts so the second question is is So where can I get those very nice throat bubbles that I got? So if you want to have a second one, you can get it on our boots. It's on a 29 So you can can can get it over there So let's quickly introduce myself. So my name is Eric Kessels I'm working for Fairbanks a company in the Netherlands and we are doing professional services all around open stack On this side is Matthijs van Velu. He is the senior open second consultant in our organization So he knows a lot of stuff about what about open stack how to use it how to implement it all everything around open stack That's what he knows and if you look at our roadmap We are all started in 2011 with open stack and that's worth the early days I think when we started the open stack foundation wasn't even there So when I had my first training in rec space, there were only Three projects for open stack so that was Nova compute Nova network and it was swift. So it was very It was a pretty easy training but if you look at now how many projects are it's huge and and I'm gonna talking to people during the Exhibition a lot of people find it difficult to to see what kind of project do I need what kind of functionality? Do I need where do I need to look at and even I? Working for open stack for all 86 years, but it's not easy to to see what kind of stuff you need for open stack but we went that route and Now we are here and I think all our businesses within on our organization is fully switched to open stack So originally we were of Yammer shop and now we are an organization. That's fully oriented around Around open stack and I think we are very proud in achieving that So what I want to talk about today is about what do we see? Trends of our open stack customers. What do we see? What do they ask? What is going on with them? So what they are asking is fast faster adoption is required Because when you implement open stack and you want to go into use it The time that you put your workloads and open stack needs to be short when the environment's ready and up and running So what we see is that organizations take a long time before they put weight on their open stack Environment because they think it's complex. They think how do a hell can I leverage the open stack platform? So faster adoption is really required for that and how do you do that? Put workloads on the open stack environment very quickly so that the open stack environments gets weight in your organizations It always also convinced your internal organization It always convinced your IT teams that open stack is a super good platform in running workloads on top of so Make sure that the workloads go fast into open stack Another trend that we see is that there are a lot of organizations started with open stack They put workloads on it in the beginning. It was a maybe a test machine and slowly it gets to Acceptance there and then finally somebody said hey, where's my machine? Oh, it was a production machine. So What happens with that? So people are started with open stack that build it They extend the environment and at some stage customer ask for knowledge They ask for as a lace on the open stack environment and that's that request we get often So can you help us with getting the platform into the next level getting it get open stack in a size and make it Maintainable also for the IT organization itself And that's what we provide so we help them with to make them making the right steps To make sure that they meet the as a lace that customers that are using the open stack platform are being met and the third one is The need for developer infrastructure. So developers want to start with developing Code, of course, that's why they they are there and ideas come very quickly And so if you have an idea and you want to build your developers So you will you build your application Then you really need to speed up in building that infrastructure for those developers to to get them going and you want to do it in a structured way so and So that's really important when you want to speed up because that's all about from only one The one who is innovative the one who is the fastest will be the first that reached the finish line and Those topics we will show you that in a in a demonstration. So how can you address these topics? And so we prepared two demonstrations. So one demo is all about Migration, so how do I get my workloads on open stack? And the second demonstration is how can I can I build the developer infrastructure for your developers? So one thing so I will start with with smooth workload migration. So what is the challenges about that? So what you don't want is that you if you want to put weight on your open stack environment You don't want to do everything manually you want to automate the process for moving workloads into open stack And that's a that's a process that you want to automate But so as you know because when you are having a workload running in vmware and you want to move it into open stack During the migration process you need to change you need to change the drivers You need to put software in that vm that can run an open stack and if you do that all Manually that takes you a lot of time, especially when you have a for example five six hundred vm's running on vm Or you want to move it the open stack so then you want to automate that process and There is software for it that you can use to speed up the process in that So for for this process we use haystacks That's a software partner from us that that provides a software that runs on an open stack And what I would like to show you is how is that tied into open stack? So how does it look like when you run it in open stack? So that's one point so and how do you set up the migration? So how do you make sure that the workloads start replicating between for example the vm rope? environment into open stack but it also could be an open stack to open stack because We still see customers using liberty for example as their open stack platform and they want to move to the latest version It's really difficult to to upgrade those platforms So so you need to have a way to lift and shift that workload from the older open stack stack platform into the new one So the migration you can use it in different ways And so how has been how how does that work? so if you look at the architecture picture in this case you see on the On the right-hand side you see the source environment So in this case it could be a vm re environment. It could be an open stack environment. It could be and It could be Azure it could be Amazon So it really doesn't matter where the workload is coming from and the process is controlled by the haystack accura That's an appliance that you run inside of open stack and it consists actually of two parts So the top three green parts, that's the appliance that controls the migration and the replication process So that's the controller that's linked into open stack. It's linked to the keystone It's linked to the open stack service that it will use to do the migration stuff and The bottom part of fourth one That's an additional machine that runs an open stack and that's actually does the replication because you can imagine when you have For example 400 vm's you don't want to have the have one machine doing your replication process into open stack You want to scale up in that so that's Why there is an additional machine that does the replication because you want to scale up when you have a lot of workloads running And you want to move them in open stack So on the on the on the right-hand side, that's the target cloud in our demonstration. That's open stack and There will the the migration will process will move to so so actually what happens when when you set up a migration The data first is been replicated to the target environment So the first step that you do is set up the replication so the data Lands on the open stack platform and when that's done and the data isn't sync Then you are ready to do the migration itself So actually there are two basic steps in that whole migration process And if you think about this another interesting use case is also using it for disaster recovery So in this demonstration, I focus on migrating workloads But in the same idea you could use this also for disaster recovery purposes because You replicate data from one platform into open stack and you can run that workloads also within open stack So you could also think about a disaster disaster recovery model for using that So let's see how it looks like Let me first log into the environment. So I do the demonstration into the sandbox environment that we have at our in within our organization and If you want to test it peacefully feel free to contact us because the sandbox is publicly available So if you want to run some Migration test or you want to test the developer infrastructure? Yeah, please come by a little boot. We are a 29 so you can Request for access to the environment and maybe play with it so so within open stack I have a tenant where the Hastax Acura appliance running so as you see here, I have one Hastax VM so that's the appliance that's controlling the migration process And I have my fairings cloud agent. So that's the VM that does actually the data replication so here is the The hastax appliance so I can log in with my credentials and I can have access to the appliance And this is the appliance where you control the migration process with So as you see I have mentioned I clicked on fairings. So that's a customer within the hastax Acura appliance So you can have multiple customers Configured when within the hastax hastax appliance because I can imagine when you have one customer You want to work work run the workload migration into another customer will maybe have other Requirements to do the migration so that in that way you can have one appliance and doing multiple customers workload migrations So the first step that I need to do is to set up the replication between my hastax appliance and between the VM that I want to migrate The steps that I'm doing now. I just want to show you how the steps are So the steps are all also possible to automate But just to give you an idea what's happening and what the steps are doing underneath, but of course the process is also That you can also cross you can also automate the process of course So when I'm doing the replication I first prepare my agent So what happens is that I prepare an agent and that agent I install it on the VM that I want to migrate So that's my first step that I'm going to do So in this case I can select what I would what do I want so in this case you see three options So one is the visa options So what happens there is that you don't put the agent inside of the workload, but you put the agent Inside of the hypervisor within VMware So then you can speed up the migration process instead of migrating one by one you migrate One hypervisor within VMware in one go So that's why you have the fees for integration to to make the process for migration workloads even easier So the second two are very obvious So I have Windows workloads and I've Linux workloads So that's something what you can choose of and those are the agents that I that you install on the VMs So I selected the Linux and then I select The group where I want to use the agent for and then I can download it. So it will download a RPM package in this case So one thing I would like to show you is that In this case you can also select which Linux flavor you want so I can also select the sender with one So what happens then is it will download an RPM package on my laptop in this case And that's the package that I can use to put on the VM that I want So the process is running. So what I do it just I Just just to show you I use a filezilla so I can migrate or move the agent I can copy it on the VM that I want to migrate to Just wait it's cupping and over So what I'm not if if the agent are there I can install the agent on the machine. So Let's flip to the machine and reconnect it So there's my agent. So I put it on the on the machine. So now I can install it. It's super duper easy So what happens now it installs the agent So what you see in the installation process I do not specify anything about the IP address of my appliance or anything Configuration metrics I don't specify because I prepare that when I designed the agent So the agent is running now and now the agent automatically connects to the haystakes accura appliance to say hey I'm ready to to go. So if I'm flipping back to the to the haystakes appliance I See that there are two machines in a discovered state So now the machine is discovered so the agent has connected to the haystakes appliance and saying hey I'm ready. So tell me what to do. So the first step what you are going to do then is do the migration Oh, do the replication. So that's the first step that are you are going to do? So I start my replication and now that the replication process is going to start so what happens now in OpenStack, but that's the of course the stuff that that's happening on the background is when you look at the volumes in your haystakes appliance, it will create a volume in OpenStack and This volume will be used to replicate the data through the VM that's running in the haystakes tenant and When the replication is finished then the volume is ready to use for the migration process So then you can make into the next step and saying okay data is replicated also based on the speed of their connection between the infrastructures of course and then you can say okay now I'm ready to go and now I need to set up number migration So what you can do with the migration is also setting up a Application group so in my demonstration is super easy because I just do one I just do one VM, but of course when you want to migrate an application Maybe you have 10 or 15 VMs that are combined as the application so within the Migration tool you can set up a migration plan. So the plan is actually saying okay now I'm going to do the migration and this is the steps that I'm going to do So let's see how that looks like you see that that the VM is migrating. So it's yeah It's going through the internet connection. So it's not super fast, but I already prepared of course a VM that's already been replicated So what you can do for making the migration process easy there is a migration plan So the migration plan can for example say My IP range is X and when I'm doing the migration the IP range is going to change It also can be the same because I can imagine when you have more like a legacy application that you don't want to change the IP addresses and The cool thing about opens tag is that you can run that application with exactly the same IP configuration Within your tenant because that's a closed network So you can fairly easily put your application in the tenant you can test it you can see if it's working Maybe you do the replication on another time you do it three times and when you are feeling well and then think okay Now it's ready to go Then you couldn't can do the actual migration and the good thing is that you can run the application within the tenant So you can you can prevent that maybe This migration will come on to the network and saying hey, I have the same application on the other side What do I need to do so you can isolate the testing of the application on open stack to let's go back So the migration plan in this case only one machine So I cannot easily add additional machines to the migration plan because it's an application. I can add it So I have a fairly easy Graphical interface to set up the migration plan. So that's fairly easy So let's see what happens when I'm doing that So what happens when I'm doing the migration just to explain it is when the Migration plans actually happens. It will create a network in open stack. It will create a router in open stack It will create a virtual machine and open stack to make sure that the that the machine is hooked up to the Volume that is going to use so all those steps are automated from one single point and this from the Acura he sticks Appliance, so let's do the replication itself. So let's see what happens. So I Run the replication so I can start sorry. I run the migration and then I run the migration So let's go through it. I selected plan. So that's a demo plan. I prepared. Let's give it a name so It's saying it's it's showing you the machines that are inside of my migration plan and Then I say run migration So now the migration starts and now it creates when opens tag It will create the the machine that you want to migrate So let's see what happens there So first of all it needs to prepare the volume. So let's see if that's happening So I go to volume So that's the first step that needs to do because it doesn't use the volume that's doing the replication because that will Of course that will impact the volume that is doing the replication selves You want to have a new volume that you are going to use for your application? So it takes a while because the internet connection is not super fast So in that way you can fairly easily do the migration Into opens tag. So that's the way in really automating the process I will show you the screen later on because it takes a while because the VM is has been building up and In that way you can fairly easily migrate workloads in Into open stack and put weight on that open stack environment because if you take too long to put really weight in your open stack environment People also lose their attention within the organization or managers think it takes too long So in this case you can very very easily do the migration into stack and speed up the process in using that that environment so when the within the Within haystacks you also have the event view viewer of course, so then you can see all the events Stuff is going to happen. You see that on the top the migration process has started So it's running on the background now and now it's preparing the VM to to be prepared so so this is the demo for the haystack's appliance so When I'm finishing or refinishing the demonstration you can ask question So I would like to move forward with the next demo and then after that you can ask questions about About the migration process and Nick is here from haystack So if you want to ask him questions He is available also for answering that because he is the founder of the organization and he built the application for this team Okay, so let's move to the next to the next demo to flip it over so So the next demo is all about developer infrastructure. So if you want your developers start working so So what do we need so one one thing that you need a desktop or you need a laptop that they can use for running their developer tools and of course love developers ask for a Kubernetes cluster or something like that to build their applications So we as fairbanks we focus on the infrastructure layer So we're not focusing all on the tooling that you put on top of Kubernetes or that kind of stuff is not where we all where we're not experienced in and that's not our goal So we provide the infrastructure for the for the developers. So we are on the the bottom level of it So one thing is fun How does it infrastructure look like because all the advantage of open stack? It would be great if you could also use that that advantages to for example to put in desktops For developers in open stack because you can exit this from any place where you are You can you have influence on the performance that you want to give to desktops Do you want to have fast desktop? Do you want desktop with a lot of data? so running those desktops in open stack can really can really help in Defining the right desktop Layout that the desktop that the developer needs to use and another thing is when you want to Because when the developers going to start and they want to for example use cuban 80s It would be very cool if you could just deploy a cuban 80s in one go and not thinking too much about How the cuban 80s cluster needs to be confer configured all the steps that you need need to do to make the cuban 80s Cluster running so make those steps easier will speed up the process for the developers to get going with their work So but that's the goal of course because you don't want to wait for four months before they can start programming so and The infrastructure helps you to speed up the process. So for the cuban 80s installation. We use juju because We are a partner from canonical and we use juju and mass already for Five years because it also is a super cool to deploy open stack with that's not the talk for today But our experience in using mass in juju to deploy Open stack environments It's also very useful when you want to deploy a cuban 80s cluster because that's It's also an infrastructure that you want to deploy in one go So we use juju for that and juju extra look Juju is actually looking like a little bit the same as the haystack supplies because you run a juju controller Inside of your tenant. So that's a machine that controls the deployment process And when you trigger the process then you can say, okay Let's deploy a cuban 80s cluster and because of the that the juju controller is running inside a tenant It will deploy the cuban 80s cluster in the tenant where the control is sitting in And because the control of course is tied into open stack. It's attached to the to the APIs So that's why the juju controller is that said that's the way how you can use that because it's sitting inside of It's inside of the tenant So how's it going to look that how does that look to make sure that the screen is refreshed So this is my juju controller sitting an open stack It's a graphical interface and I only show you that because then I can show you a bit how it works Think we together never use it because we always use the command line That's typical how Linux Linux engineers will act But that's more a way in showing how you can use juju how to deploy a cuban 80s cluster so what you do you build a your controller and On the controller you log into a model So the nice thing about using a controller in combination with the model is that you can have one model That's doing the cuban 80s installation But you can also have a second model that will install another type of application And what we also can use is we use for running the developer desktop infrastructure I'm going to talk about we also deploy that with juju so to build the Desktop infrastructure for the developers we use juju to install that infrastructure in the open stack environment So let's do that. So one way in doing it. I can just select plus so I come in to the Start to into the charm store where all the configurations are that I want to deploy with juju And and here you see there are different flavors in installing a cuban 80s cluster People who maybe know Canonical and know they are cuban 80s distribution. They won't recognize these and what you can do is just select the flavor that you want So so let's let I will select the cuban 80 score And I will add it to the model So what happens here is that it will add the model to my controller. So that's step one So now it's saying okay. I want to build my Cuban 80s cluster and now I'm gonna ready to deploy it the The bundle that you use is configured and saying if I want to have the cuban 80s cluster How many workers do I get? How many how is my flannel network has been built up? So all that configuration work is done within the bundle that you employ deploy So now I can deploy it and I commit and then the cuban 80s cluster will be deployed within the tenant right Controller is running. I didn't commit it. So I can just refresh it then The configuration is gone because I would just want to show you how you can deploy the desktop infrastructure So we use Leo stream for that. So that's a VDI solution that runs in OpenStack in our use case. We show you how you can Deploy Linux desktops in OpenStack But you also can use it as a VDI solution for Windows desktop or whatever desktop you want with different kind of protocols But the cool thing that is really nicely integrated into OpenStack so let's Find it out So in this case, you see that there are two components So one is the broker and one is the gateway. So I just let me select the broker And I add it to the model then I say deploy and Then when I actually do deploy it will install the Leo stream broker within the tenant who that I've configured I was in the wrong model. So I need to redo it. That's running Okay, so now I've deployed it. I selected the model that was already there So it creates a new model to put the application inside. So now my Leo stream broker is deployed So now I just need to wait before my VM is ready in OpenStack to use the the Leo stream broker And but we already prepared that for for you So we're not gonna wait before the Leo stream broker is has been deployed. So we're gonna Flip over to the Leo stream demonstration We would like to give you the word to Matthijs and he's gonna take you through the the Leo stream to do the into the Leo stream demonstration Hello, I'm Matthijs Felu working for Almost four years now for Fairbanks and I will try and show you how How the instance is for layer stream for layer stream work so What Eric just did was deploying the the instance Charm and we already have those those items over here and Layer stream gateways also deployed. That's also a separate to do charm and and both work together to Create the infrastructure for the whole layer stream interface for the remote desktops to connect to And when you log in into the layer stream environment, you are You log in into the system Where you can For in for here you need to add a center so And layer stream works with different kind of environments. We choose in this case, of course for open sec And and we connected to our Fairbanks sandbox environment We given where the API is for the keystone given all the configurations like the domain the project the username the password and and how often it needs to refresh the information it sees it can get from the OpenStack project if You have configured the center then you can configure the gateway In this case the gateway was installed in the center as environment You just give in what it's remote. I feel need it is It's configured to what is internal IP is so it can connect to the broker because the broker is the part where All the data is stored so the user accounts it stores where the LDAP configuration is if you know connected to your LDAP or AD environment So that's where all the data is you can just remove the gateway and install a new one or install multiple gateways To have a load balanced environment So if we're now going to look in the configuration of the desktops, I created a special configuration Ubuntu Bionic desktops And if we look at this configuration You can select on What the name is going to be on where on which center it needs to be installed You can have a lot of different configuration attributes changed here You can say how much? desktops need to be running at the same time or at least started You can configure How much it needs to start in the beginning how when it needs to stop configuring? Which center it needs to be? What name the instances are going to be within the open-second environment? And the biggest thing here is you can choose which Which image you want so I prepared a Ubuntu desktop image in my case. I just installed it within a KVM on my own laptop Installed all the components needed for that and then I Uploaded that image into the open-second environment So that the image is in glance Once I did that after that you can change the image update that image create a snapshot of that image and then you can change Which version of the image you want to deploy it for your customers So if you have a new updated version with all the security patches installed and also the new Developer tools or whatever you want you can just change it here and say okay now remove all the desktops That are already there and then you all your users are having the new Updated instances available, so I have here a 4.0 Which was my previous one I updated some Some packages and I selected the 4.1 You can create special flavors in this case. There is an a layer stream flavor if you want for for your graphical designers you can create a flavor which has access to To a GPU which is in your open-second environment so that they have graphical power to to work on You configure the network On which network it needs needs to be you can select your security groups you created in this case We have a desktop and an SSH configuration created the desktop has the RDP ports open the SSH of course the SSH ports and When all this is configured You need to create a protocol plan and we have several protocol plans over here We have an HTML5 Apart, which is very nice to have because this way you can log in from anywhere in the world Even in internet cafe and log in into your system Without having to need any special tool But if you are on your phone or your tablets or you have your own Laptop and you want to use an RDP connection. You can just Have an RDP connection connect to it open your desktop and work on that And they are stream has a few options for power plans So if you don't have that much resources available on your hypervisor, you can tell they used to okay when somebody logs out, please shut down or power off that system and Keep it off until that user wants to log in again. So you can So when the user logs in it will send a signal to that desktop it will start the desktop again and it will Be available for the user to log into and the same goes for release plans like for instance if you want to have something in a Internet access in a library so where everybody can log in and You can create a plan that when somebody has used that system. It logs out It will remove that desktop or it referred to a previous snapshot so that every user information you had is gone The internet history is gone as in people can work on it again with a clear state of desktop so when When we want to log in this is the hml5 part. I just Login into the system. I click connect on the desktop It's going to connect the lay of stream users the guacamole environment and I have here my desktop With several developer tools and which studio power shell visual studio code by charm And what also is nice if somebody wants to watch a video or They want to even have a conference call because it also supports audio input you can just Start Within your remote desktop and If we want to show you the connection Through an RDP session, we will just somebody gets an RDP file and that works on phones on desktops or tablets and you will have remote desktop connection available and The remote desktop connection is available This one first one went through a floating IP so it has a public IP address But it's also possible that that connection goes through the layer stream gateway So you just need to have one public IP address where the gateway is on and it will Connect through that gateway Without the need of any floating IP is all all the connections are inside of your tenant and nothing goes outside So no direct connections your if there is some strange port open on your desktop you can Nobody can access that So and I will also show you an SSH session So you can create servers for instance if you want where somebody can connect to and then they can just connect They can connect with an SSH session within your browser It's relatively safe because you connect first into your system as the user Which has access to your to the desktop or to that system and then you also need to log in again Through the SSH session, of course, so yeah, this is a very nice way to Make sure that all the developers have the the right tools always updated and being able to To have the correct tools and speed up so if you have a new a new colleague You say here's your desktop you can start right away instead of having him to install his laptop all over again and Installing the tools they need you have everything available for you and all those desktops you can see here running are within the layer stream project so that's a That's a bit the case of the layer stream environment, I Don't know if you want to Show something I want to wrap it up Because I think our time is gone. I think we could have spent some more time and chatting about it But our time is limited, of course So if you have further questions, we are available here. We are we have our boot on the marketplace So it's a boot 8 a 29 So please come there because I see that the new commas are coming in so I really thank you for your attention that you was coming here and Yeah, peacefully feel free if you have questions. We are here or on the boot and I would say have a safe travel home When you are going today or tomorrow you are going on. Thank you very much for your attention and see you soon