Kubecost helps firms monitor, optimize their Kubernetes and cloud spend

00:00 
Hi everybody. Welcome to DEMO, the show where companies come in and they show us their latest products and platforms. Today, I'm joined by Andrew Dawson. He is a solutions engineer at Kubecost. Welcome to the show, Andrew.
 
00:09
Thanks a lot, Keith.
 
00:10
So what are you going to show us today? What is Kubecost?
 
00:13
Kubecost is actually a finops, or a financial operations platform that companies self-deploy within their own cloud accounts to track their Kubernetes and their cloud spend, essentially to understand what they're spending across AWS, Azure, Google, or on premise Kubernetes, as well as other cloud services, and then potentially save money by optimizing those cloud services.
 
00:35
Is cloud spending out of control over the last couple of years, and that's why this was developed?
 
00:40
Yeah, absolutely. I mean, for the last 10 years, it's been growing exponentially, and people have been using finops tools. But with the introduction of Kubernetes a few years ago, it kind of introduced a new challenge for finops. So Kubecost actually came out as an open source tool that was downloaded by developers all over the world. It eventually became an actual business and a real enterprise product. And today, it's a fully fledged enterprise finops tool that's used by a ton of the Fortune, 500 companies. And we're growing so helping people track a ton of spend across all their cloud resources.
 
01:13
Now, who is this designed for? Is it just a company that's using Kubernetes, and I'm assuming a lot of companies have cloud, obviously, but it's not highly specific, just for Kubernetes. It can be used by more than just an engineer. Correct?
 
01:28
Yes, absolutely. So we see a lot of developers to finops, to finance pros using the tool. It did start with the developers. So ideally, you're going to have all those people collaborating to understand, when we launch new infrastructure, what does it cost? And if there's software running on it, is it profitable? Are we running a lot of wasted infrastructure where we're turning stuff on and it's just not shutting it off? Yeah, so those are the kind of things we can discover can really be helpful for small to medium businesses as well. Where a $2,000 a month cloud spend is a ton, if they can reduce that to $1,000, it can really help them out, right? And then the big enterprise companies we're potentially saving hundreds of thousands to millions by shutting a lot of the stuff off.
 
02:06
So if they didn't have this platform, what would a company do to try to track these costs? Would there just be like a big stack of paper somewhere, and with spreadsheets and things like that?
 
02:16
Yeah, they're using the internal tools from the cloud providers, but a lot of times it's a lot of different logins. You kind of have data that gets lost in translation from one system to another. So Kubecost is just trying to put it all in one place for you come in and look across all of my environments and just see, what are we spending? What's the potential savings? And actions on top of that.
 
02:35
All right, let's jump into the demo and show me some of the key features.
 
02:38
Absolutely. So what we're looking at here is the Kubecost platform I've deployed, and I've actually set up a bunch of different clusters that are checking in. This is self-deployed, remember? So it's within my own accounts. It's not being exposed to anything over the internet. Companies like the security of that, because they want to make sure all the stuff just stays in house secure, right? So here we can show an AWS cluster, an Azure cluster, Google cluster, all checking into the same Kube costs. We're actually tallying up those costs and giving you the total Kubernetes costs across all those platforms. In our case, we're looking at the last seven days over $1,000 in Kubernetes costs. But then we also have total costs here, so that's plugging in the cloud accounts and the cloud billing to get not only the Kubernetes specific stuff, but the stuff like EC two servers, storage that's being used potentially by these applications, but isn't actually on the Kubernetes cluster. So we allow you to track all of that in one place. And you can see my total cost is a lot bigger than my Kubernetes cost. So I really want to get a handle on both right and on the flip side in regards to tracking, we have this Kubernetes and total cost, but then we also have optimization and efficiency. So that's taking a look at what you have and then really seeing, is it efficient? You know, 14.8 that really means I might have some areas I could size things down. that
 
03:55
That seems low, right?
 
03:58
Yeah, you don't want it. You want a higher number for efficiency, right? Yep, you want a higher number. That means that you have less waste. And that's what we're trying to identify and give you actions around.
 
04:03
So if someone, if someone was, like, looking at their report and they said, Oh, well, I want to figure out how to optimize, like, you could still use the platform to do that, right?
 
04:12
Absolutely. So really, it's about tracking. And originally, when Kubecost came out, there wasn't an easy way to just break down inside the cluster by things like namespace. These are automatically namespaces are on your cluster no matter what, so they're easy kind of folders to report by. So just allowing you to slice and dice by the different groupings like namespace or the owners of these different resources or the products that you're running, and then not only seeing the total cost for all these things, but if you notice this column right here, efficiency, each one of these line items has an efficiency score, whether it's really low and a high price, meaning I potentially have some right sizing I can go in and perform, or high efficiency, meaning that the usage is a lot higher than what's being set, which can be its own problem. So identifying. The costs, but then also how much waste is on this product or on this owner's team of infrastructure. Okay, another really good thing that we want to just quickly show is the ability to show cost per provider. So this is a report that a lot of companies don't have an easy way to show. Just showing in my last month, what did we spend total on AWS, Google and Azure? And then within that, how much of that went to Kubernetes? OK, so really, just allowing these people to understand what they're spending and how much of it is going towards this Kubernetes engine, and then
 
05:35
What's the red and the green part over there? Like, is that an analysis that you're doing?
 
05:39
Yeah, so that's our trends API, and it's basically showing compared to the previous time period, what happened, whether we got a lot more costly, which we did in GCP and Azure, or we actually saved some money, which we did in Amazon. So just helping you track that. And all of this is available through API for developers as well all these percentage trends and other pieces of data. And
 
05:59
You mentioned this is self installed. So how long typically would it take to install this for a company looking to save some money?
 
06:07
So you can install our free version very easily, probably within 10 minutes. If you're used to installing Helm Charts, which most Kubernetes users are okay, and then our enterprise setup would probably take about an hour. But we do offer a 30-day free trial. We really do empower our customers to kind of set everything up without having to talk to us. First, they can really get a full enterprise set up in a trial mode without coming to us. And then, usually they set it all up come to us, and then we can help them interpret the data.
 
06:32
Was there anything else that you want? Was there anything else you wanted to show?
 
06:36
So another cool thing that we've been working on is network tracking. So understanding the network costs across different environments in these Kubernetes and cloud environments, a lot of customers and clients just want to be able to see, across Azure, across Amazon, where are there network traffic connections on my Kubernetes cluster, and where actually is the money going? So if I look at my last 30 days of network costs, I can in cube costs quickly see that my integration test cluster has sent out 246 GB received over 2000 costs above $30 so it's a small demo environment, but sometimes we'll see thousands of dollars in one connection from somewhere, right?
 
07:16
And then the user of this would be able to figure out how to save some money based on, on those, those that data, right?
 
07:23
Yeah, we’re working on recommendations around this, where instead of going to the internet, you might be able to go to a private endpoint or something like that. But we're trying to get it visible first, so the companies can understand where the cost is. Some companies have maybe 40% of network spend. So we're trying to address the compute, the memory, the GPU and the network for all these companies.
 
07:43
And you just recently announced that you had 10 million installs of the product, right?
 
07:47
Yeah, it's been exponential, which is crazy. It's really great to see. So we're hoping more people download it and really kick the tires on the free version. We have an always free version under 250 cores that people, small to medium businesses can always use, and then also the enterprise version for those who need it, and you pretty much know when you need
 
08:05
Great stuff. Andrew, again, now, where can people go for more information? Because I know there's a lot more features than just what we're showing here.
 
08:14
Yeah, absolutely. So github.com/kubecost, you'll find our Helm Charts and a bunch of our documentation. Also Kubecost.com where you can really see a bunch of different information, use cases, case studies. So you'll find us if you go to our website. We also have a YouTube channel, so check that out as well.
 
08:30
All right. Thanks for the demo, Andrew. Cool.
 
08:31
Thank you very much, Keith.
 
08:32
That's all the time we have for today's episode. Be sure to like the video, subscribe to the channel. Add any thoughts you have below. Join us every month for new episodes of DEMO. I'm Keith Shaw, thanks for watching.