 Hello and welcome to another OpenShift Commons briefing. This one is on cost management and Sergio Acon Cardianis from our Red Hat Product Management Group is going to walk through and give us a preview of some new SAS capabilities that are part of Red Hat OpenShift. And I'm going to let Sergio introduce himself and his topic and we will have live Q&A at the end. And as always, I will post this on YouTube along with a link to the slides. So please, Sergio, take it away. Thank you very much, Diane. So I'm Sergio Acon Cardianis, I'm the product manager for this new SAS capability that is part of OpenShift since version four to three. And I'm going to tell you a little more about why we're doing that and I will deliver a short demo afterwards so you can see it in real life. Of course, this includes some broad map and it's a technology preview. So just bear in mind that when you're saying this. So one of the reasons for this is that understanding whether the cost of a container in the cloud is really complicated. There's a lot of things that move all the time and that's the beauty of it. When you have containers, everything is more dynamic that you are used. And so you get basically a misalignment between what you are seeing, the resources that you are using, the way you are using, and your business. When it was a VM that supported a single application, that has been fading away into a place where you're basically running thousands of containers in different machines that appear and disappear. So that's a complete paradigm shift. And that means you need new tools and you need to make sure that those tools are working. And cost management is an application that is basically helping you fix those problems. And I would say this is important not only for the operation or the IT manager that wants to know how things are working and just basically need to know more about what is going on, but also to the business side of things. So just understanding how much something in particular is costing me, it's more and more being important for the business. It's easy. We talk a lot about VM sprawl and about people creating accounts in the cloud providers and using them. But it's more than that. It's being able to accurate report your financial numbers. It's being able to set up a forecast and a budget in an easy way that makes you really happy with the result and allows you to take business decision on time. So we're creating an application service that is part of the OpenShift container platform since version 4.3 and is going to allow, first, see your cost, see your cost in the cloud, see your cost in OpenShift, understand what is going there. Number one is map your business to your infrastructure, to your OpenShift elements, to your cost, and then understanding how you're using your infrastructure, how you're using your containers properly. And the last part of this is getting the information that you really need to communicate that effectively to the people that need to know. So in order to generate your show back, charge back data, and influence how your customers, internal customers, external customers, are going to be seeing and using your infrastructure, your platform. So this is the new service, cost management for OpenShift container platform that is right now in Tech Preview and will allow you to see, to have that visibility in your cluster, both on-premise and in the cloud, and allow some mapping, some modeling of your charges, and also allows you to have an interface between your cost in the cloud and your customer are using them. There's one big thing you need to understand that's already some capabilities of this in OpenShift. So if you install operator metering, many of those things can already be done. So you can have reports, you can schedule them, you can see the basic data because OpenShift includes Prometheus out of the box, and of course you have optional integration to Amazon and you can plug that into any BI tool. So if you really just wanted to have a report, a weighted report on the usage using Prometheus data, you already have the capabilities in operator metering and it's already extendable using an SQL query language. So that's not what we want to do with cost management. We actually use that information to provide additional value at it. So the first thing is we do cost management for any OpenShift, for most of one, anything that you have, it's basically you can add it, provide that you have an OpenShift for environment. We enhance the information using the metrics, of course, but also metadata coming from both OpenShift and in the cloud. We provide a full graphical interface on API, so you can use it in a better way. It's easier to use with dedicating a lot of resources to the UX piece of this. And of course we don't just report, we just don't do reporting. We also allow you to model the cost in different ways and start using labels and tags to do some business mapping. Just to give you some example, so the first one will be, well, you're running OpenShift and you're running that in many clusters. You have your production clusters, you have your QE clusters, and they are all running in the cloud, or some of them are running in the cloud and others in on-premise. So the first thing you need, if you want to make sure that your costs are under control, it's a way of looking at all that information in a single place. So we have customers with more than 200, 300 accounts in Azure or in Azure. We have customers internally that have up to 2,000 Amazon accounts. Just being able to make sense of all of that is a challenge. But it can be the same with OZP. You have 10, 15 OZP clusters. How are you going to know the number of resources that you are using? And if really the deployment, it's being used in the way that you intended. And also, if you're really getting the TCO that you're expecting from that. There's a piece there too, that is the link between these two layers, how OZP clusters are deploying in Amazon or Azure, how that is going to affect your costs. So if you're using a database in Amazon that is not inside OpG, but outside, like many customers do, how that is affecting the cost of your service, not only that inside your cluster. And of course, that's basically going to map really quickly to how much is my project cost to me, how much is my application cost to me, and how can I generate a report that is going to allow me to see what services are running, how much is cost to me. And just do as soon as I can make the right decision, the right business decision that we have. So let me go quickly to a demo. And I'm going to show you a demo of the application. So you can see here, this is cost management. It's part of cloud.redhat.com. Coventry is in beta, but it's going to be GA soon. Our schedule is to be released as GA in Summit. So part of the cloud.redhat.com services, you will see cost management. And any OpenShift 4.0.3 customer or plus will be able to use it as part of the subscription. So it's included with your OpenShift 4.0.3 plus subscription. So what are you going to be seeing here? The first thing, it's, of course, you need to manage your infrastructure. Being Azure, being Amazon, we plan to send it in the future. But right now, this is the two clouds we offer. And the first thing you're going to be seeing, it's all the information from all your accounts. And one of the things, most of the time, what we find is that the top accounts, the top services are going to represent, well, Pareto, 80%, 90% of your cost are going to be a limited number of services, a limited number of accounts. So the first things we show you, it's those most important account services of region, those parameters that allows you to see what is going on. The way we do this, we plug into your bill. So in Amazon, it's the core files that provides information about your invoice. And in Amazon, we do something similar. And we're going to show that together and compare that with last month. Reason why one of the first things you want to see when you're seeing this is that the number of resources that you're using compared to last month, the cost of your accounts is going to have a profile that is easily identifiable. If there's something wrong, you're going to be really easy understanding why this change of slope, whether it's forced because you created a new service, you launched a new application after you generated a new cluster, or just basically because there's something wrong and you need to investigate further. So one of the things is let's make it easier to see what is going on in all your accounts. And then in order to understand this graphic that is comparing the cost on last month to the cost up to date, we're going to be showing the components. What is your charges in virtual machines, in storage, network, or databases? For Amazon, it's actually the same. And this is going to compare again this month against the previous month and it's going to show you your top three services, your top three accounts. The other thing that you can see here is that if you have one parent account and many children's accounts, we're going to ingest all that information and show it that together. So if you have like, we have three, five, 11 parent accounts and then several hundreds children account below those, you're going to see all of them here. You're going to be quickly identifying what is going on. And then as you can see, we also are adding information about accounts and regions so you can easily compare services like computer storage on network and databases. So I was using as an example before, what would you have this month? And as you can see, we decommissioned a lot of VMs a few days ago. So the charges are going really slow right now. So that's good, but that's not why this tool has been built. This is just basically something we need to do to manage your infrastructure. The important thing here is to understand how that affects OpenShift. So for that, we are doing two things. The first thing is identifying which parts of your infrastructure are being used for OpenShift. And we do that in two ways. The first way is that your OpenShift notes are running in Amazon or Azure and Prometheus operator metering knows that. This unique ID using Amazon, using it in Azure, that identify which VM is running which note. So we can basically map the notes in OpenShift to the VMs in Amazon and we can do the math. And the other thing is we use tagging. And there's a lot to say about tagging. I cannot extend myself too much on that. But if you tag appropriately things in OpenShift, you use the proper labels and you tag things in Amazon in Azure, we basically will use those labels and those tags to identify that that RDS service or that EBS is being used by your project or by your cluster. So we can show that information together. That's good, but that's not going to support every single use case because you have clusters that are not running in Amazon or are not running in Azure. And you have pieces of your software maintenance part of that that is not being reflected in your bill. There's some analysts saying that only between 25 and 50% of your real cost at that of the infrastructure, being the 50% to 70% remaining, are every single thing, software maintenance, support, backup that raises all the things that you need to be running so your application is working. So we allow you to create cost models. And what is a cost model? So a cost model is allowing you to generate a model of your cost. The way we do that is to fall. So you can add that for Amazon and for OpenShift, but you're basically going to be able to select CPU, memory, storage, or note. And then a selection of how you're going to measure that so you can put a price. So you can put a price per note per month or you can put a price per CPU usage per hour. Just put a price on it. And once you have done that, as we understand that that's not the way to model all your charges, that's going to allow you to model the charges when you know the cost of your CPU, when you know the cost of your memory, or you know the storage cost. The other thing you can do is, as I tell you, put a price on the note and shortly in your project and your cluster. The other thing is, yeah, as well, we have identified that Amazon is 50% of your charges. So you can put 100% markup on top of your Amazon charges. And that will create a full model that you can use to get closer to your real cost. And that's actually, this is not a billing tool, so we are not trying to be exactly accurate. Well, you can if you go yourself and you really have a real number for that. But that's good enough to have a better view on your costs and then take the right decision, because it helps to take the proper decision on what is going on. Let me change that. I'm not going to recall that. Let me go back to that information. So we talk about OpenShift. We talk about the clouds. And we were showing only the top three services or projects or cluster that were more important to you. But that's not enough. You need to have information about every single piece of your infrastructure. So what we are doing is allowing you to go deeper into that. And that's where you see the details part of this. So you're going to be able to select Cluster, Note, and Project. And you're going to see all the information group by that information. Every single piece of that, not only the top three. So you have a list here with every single project that this cluster or any cluster that you have added to cost management has created. We're still comparing this month to last month because that's important information to take business decision. Then you're going to see how your costs are. This allows you to see a lot of information about your tasks, the limits, the capability of your cluster compared to your project. And there's always that view that we have before on time. Right now we have the same, but specifically for these projects. You can filter all the time. So you can define that you only want to see the project that has weather on it. Or you can basically identify a cluster and then find only those projects that are running in one of these clusters. So you have a better view of this actually where you are looking now for. Once you have identified that, you just select the information and you can export it daily or monthly. Or you can access that through the API, same information that you have in here. So you can basically automate all of this. You go to the cloud. It's going to be basically the same, but with the component that makes sense when we're talking about the cloud, not OpenShift itself. So you're going to be services on regions. You're going to be able to select account services on region. I was telling you before, this allows you to map business to the technical piece. So one of the ways of doing this is identifying the tasks. So we ingest the tasks. In Amazon, you basically configure what tasks are relevant. In OpenShift, we ingest all the tasks that are created. And surely, you will be able to filter them out. And you can basically get the same report we're showing based on that. So yeah, we have free apps running in our Amazon web services environment or accounts. Analytics, a catalog, and a cost management application. So you can see them. And also, we are showing you whatever is not tagged. And that's basically mean we need to make a lot of work to tag properly the information because most of the information come from accounts that are not using the tag key app. Still, you can filter as you could do before. And you can see a view inside of services account region and your historical data of your cost, specifically for this application. Let me finish this showing you where the documentation is found. So if you go to OpenShift 4.3 on docs.redhat.com, you will find all the documentation that allows you to configure and start working with cost management today. And I will go back to my presentation and show you a couple of slides more just to wrap up. OK. So just to recap how we do that, we basically get information from operator metering. We get into OpenShift, use operator metering, and put that information, the report that you store into the cloud. So we can use our SaaS offering to show you everything you have seen today. We do that for OpenShift 4. We do that for Azure. We do that for Amazon. We get that information. We put that into the cost model. So we modify the cost, our new things, ad markups. So you get closer to the real cost modeling and you can use that information to take your decision. And then we distributed the cost using tax so you can have your business level application cost or you can have your project, your node cost, and you understand this actually. What is the real cost of using your application, running your project independently or whether your application is running in one or several clusters, several projects? And then we use that information to provide you the dashboard and the API. So you can use that information further to generate your reports, to identify trends on just take your decisions. This is an open source project, has been since the beginning. So if you want to collaborate, you want to add new features, you want to support another thing, you just basically need to have information in another way. Just go to jithub.com, project Goku, and you will find all the code and all the information that we have. We on Red Hat, we offer that as a service. So you have it available with OpenShift. But if you want to go and just contribute, we will be really happy to get your input on your help with Project Goku. And I will say lastly, really quickly, of course this is roadmap. One of the things you're not seeing here is the roadmap. We are working to add new sources, ECP, IBM Cloud, and add new capabilities. In the short term, in the next months, you will see new capabilities for the business user, for casting and budgeting in most of them. Some capabilities we need that multi-currency or better integration to OpenShift clusters or your experience is better. And then the other great thing we are trying to achieve and we are working with the Center of Excellence for Artificial Intelligence in Red Hat is the infrastructure optimization. Being able to define the best way of deploying your infrastructure, how to use reserve instances, how to size appropriately your environment so you get the best for your money. And we will continue. There's a lot of things around cost management that we wanna be doing, like supporting what-if scenarios or adding 12 months to the view that you saw before. And of course, many in changes to the user interface and to the API so it's easier to use. How do you get to start it? Well, it's available today. You just go to beta in cloud.redhat.com and you will find us there. You just need to have an OpenShift SQ. So whatever, if you have an OpenShift SQ, you will be there and it will be available for you to use. You need to be an admin or admin in cloud.redhat.com to create new sources and to configure things. We don't want anybody to see everything. So we have our back and you can define what your users are really seeing. In the tool, by default, is nothing, but you need to add yourself. And you just go to getting started with cost management guide and that's gonna allow you to see how to do properly tagging and how to create cost models that allows you to distribute the cost more accurately and understand better what is going on. For any questions, cost management.redhat.com and we will receive an email. We are really happy to work with customers. So if you really want to contribute more with ideas, feedback or just code, just let us know and we will try to help you as much as we can. I will say that's the end. So perhaps I can accept now some questions. Well, thank you very much, Sergio. This has been really good. I'm sort of blown away a little bit by the visibility that you're getting for people on costs here, it's great. And I'm really looking forward to some of the AI optimization stuff and hopefully we can have you back when those features surface as well. I don't see any questions right now in the chat, but there were... I have one or two, if you have one time. Yeah. Sergio, I may have missed it, but when you talk about the cost models, can you aggregate your projects? I mean, typically when we have a multi-tenant cluster, there's a set of projects that are part of one cost center and another set of projects part of another cost center. And being able to basically aggregate all the costs up to those cost centers is often where the customer starts in not per project, but per cost center. Is that something that we can do already or is that on the roadmap? You just basically need to tag your project with your cost center tag. And then you will have a report per cost center. So when I was showing the application tag and how you choose them, you just need to go to OpenShift, go to your project and label, use a label on it saying this is cost center X. And then in the API, in the interface, you will be able to group by cost center. Perfect, thank you. That was a really, I thought I had some couple of follow-up to that, but that answers it, thank you. So Sergio, thank you for finally coming around and giving this presentation. We've been waiting for it for a while and we totally appreciate you taking the time today. So thank you very much. Thank you, Diane, that's, I was really wanting to do this and yeah, we couldn't. And now I hope you like it and I hope it's useful for any customer. Most definitely.