A cloud cost savings spectacular (sponsored by CloudFix) (CloudFix)

ok. hi. uh before we kick off this cost saving spectacular, i just wanna do a meta presentation here. we're gonna be acting for you just so you know, so please take whatever we do here with a grain of salt. it's all fun. it's not gonna be serious. the lessons that steph is gonna teach are gonna be very serious. but the way stephen and i are gonna act is gonna be something we've experienced how we screwed up our cloud migrations and i'm sure plenty of you did it as well. so um enjoy the show and um yeah, have some fun.

so let's bring on step. come on applause for her.

ok. you are here today to learn about cost optimization. what you didn't bargain for is that we're doing a play about this. so we decided to mix things up from your typical breakout session because we know how many you go to. we thought it would make this a bit more fun. a bit more interactive. so please like darko said, bear with us.

what we're gonna do is we're gonna have the story of two engineers and their manager and their journey into migration however, they don't know anything about cost optimization pricing or fin ops. so the lessons they learn, hopefully you will learn as well and take away into your organizations.

so let's kick it off. here are two engineers, darko stephen and our manager, raul.

hello, i'm doctor, the developer. i throw code at servers and hope it sticks so far. everything in our data center is stuck for better or for worse.

hello, i am steven the script slinger. i'm very enthusiastic about our cloud transition. i've never used the cloud before, but how different could it be to our old data center?

everyone. i am their manager. and hopefully with some luck, i tell you a lot of luck, we will migrate some of our applications to the cloud and see how it goes. let's try. ok?

so steven darko, we've decided that we're going to migrate all of our clouds, all of our applications to the cloud and we've picked aws. ok? of course, the best. so i want to get right to it. let's get it all done quickly. ok.

so i walk up. hey, steven, what's up? oh, hey, darko, how's it going? oh, good, good. what are you doing, man? well, i'm migrating this rack, this big rack of servers here. i'm migrating this rack of servers to the cloud. ok. so how many servers is that? well, let's see. 1234 per unit, 1234, 44 4 times 44 100 and 68. so you're moving 168 servers to the cloud. yeah, it was so easy. it was like three lines of code. i just wrote a four loop 1 to 1 6800 and 68 servers that printed out right in the cs b file here. so one ip address that matches with the sticker on the, on the server rack. it's great.

so we were, how did you know which serves to launch in, in the cloud? i mean, do they have ib ms in the cloud? well, these were, these were slightly older instances. so i picked, um, i picked an m four and yeah, these were pretty good size instances at the time. they had 32 cores. ok.

so, so we, you launched 100 and 68 servers in the cloud? yeah, that was so much faster than racking those instances, don't you remember? like all those brackets and screws? it was, this took like 30 seconds. the cloud is awesome.

steven, you launched 100 and 68 servers in the cloud. what are they doing now right now? it with nothing but they're ready for our migration next week. i guess they're getting warmed up.

ok. ok. so we have this whole storage system. what about storage? oh, i kind of, what would our storage needs to do? io and so let me see what i did here. hm. yeah, i picked one that looks fast. ioio one, number one io number one. that must be the best. i'm sure there's no way this is going to backfire.

hey guys. oh no. oh no, i am your friendly neighborhood solution architect from aws. my name is steph and i just wanted to welcome you to the cloud. thank you. now, i see you've provisioned quite a lot of instances. are you actually using all this capacity right now? oh, no, steven just launched it just in case we want to be ready and nothing says ready like 168 idol instances. yes.

ok. all right. so to you guys, what the guys have done here is a classic lift and shift of the cloud and they provision stuff just in case like the old school. now, you guys don't want to do that. we wanna make sure that when you do your workloads in the cloud, you create them efficiently.

so when you're doing this, there are three ways to do this. number one right size. we want you to check the cpu the memory, the storage you need for your applications and actually choose an instance type that fits and probably newer than an m four. what's great about this as well is that this is you can change these right sizing. if you do decide to go smaller or bigger when you're in the cloud, you can adjust as needed.

number two, turn things off the same way as when you left your rooms today, you probably turned off the light, you should have turned off the light, you could do the same with the cloud. so turning off instances, turning off databases is very simple. we have two ways in which we recommend doing this. the first is instant scheduler, which is a solution aws created where you all you have to do is tag your resources with the time frame. you want to turn them off and we'll do it for you. so stuff, we don't have to have 168 servers running all the time. no, but how i wanted to lift and shift. stop this. none of that, none of that.

so you can turn things off. but before you ask, yes, auto scaling, you can scale things in the cloud, you can grow and scale as you need to. so for scaling, i always recommend people have scaled actions just because you have an auto scaling group does not mean that you aren't cloud efficient. ok? you can scale down those workloads in auto scaling groups using scheduled actions in your auto scaling and do this at the weekend or in the evenings. none of you should be working on the weekends. i hope none of you are working on the weekends. so turn off those instances.

third and final stuff, delete things. some of you might have been in the cloud for a while now. you not, might not be as naive as these guys. so when you do this, you can delete stuff you're no longer using. we really recommend that you kind of clear house as often as possible and review what services you actually need. so they only pay for what you need.

ok. hopefully that was helpful. i'm gonna leave you guys to it for a bit longer. it is. we're gonna get started. yeah, i think we could probably tighten things up just, just a, just a bit, just a smidgen, just a smidgen. so, um yeah, guys. oh no.

have you seen this bill? that, that doesn't look too bad. really? oh, steven, what is this? you know, i felt like mickey mouse and fantasia. i was just here. da da da da start some instances here, start some, i guess i got a bit carried away and maybe i forgot about a few, you know, fantasia doesn't end up in a happy ending, you know, that that's a good point. um i am not happy now and so i want you guys to look at this and get it fixed right away. ok. yes. um steven. um this is bad. how bad windows vista bad. um oh, so i have a look. oh, steven, what's, what, what is a metal instance? well, i'm not sure i was browsing the ec2 console and it looked kind of cool, you know, like metal instances, i guess i turned on some music and kind of i, i must have forgot i launched them. i, how are we supposed to kind of figure out how much things cost? i mean, it was fun launching them. but how do we know how much a metal instance costs?

hey, guys. oh, no, she's back. what is this guys? what, how did you choose any of this? um that's a good question. i, like i said, i tried to match what we already had. i, you know, in my mind, i thought we could just recreate exactly what we had over in this rack in the cloud. that was my idea.

ok. what about, what about storage? did you plan any of that we were doing? i, i did io one. he, he has a script here. he launched all the servers using python. so that should be good.

no. yeah, the code. all right, let's, let's go some, go through some ways in which we can improve this. so again, what the guys have done here is when they've moved to the cloud, they didn't know it was their job to think about spend. now they do, we do.

so a couple of ways in which you can make sure you don't have a bill like that is using something like aws pricing calculator. you can input the specs for resources that you need and get an estimate price. this is really important so that you can plan your budgets and you can understand what your scope is. and at least have a plan of what's gonna happen in the cloud.

once you get there, we recommend starting off with a wbis cost explorer. this is a great tool allows you to group filter and get insights into your data and is the fastest and the simplest option and it's free. that's easy. but if you want to go a bit more granular, see a bit more information, we recommend enabling the aws cost and usage report will occur. this gives you resource level information by the hour. so you can see exactly what you're spending we could have. that's a good idea.

now, what's also really important is you guys monitor these costs. so when you're doing that, you can set up budgets in aws in the cost explorer, these allow you to set a number. if you come even close to that number, you will get alerted. whereas anomaly detection, any spikes like your weird metal instances will get alerted. you'll get a notification on that.

ok. right now, this is up to you guys to track this and monitor it. ok? all right. let's do it. let's monitor.

amazing. so, um steven, listen. um amazing things. i have to show you something. remember anyway. oh my god, you're go, you're gonna love this, take this out. um you know how steph told us, showed us the cost optimizer thing. did you know that everything in the cloud is an api call? what do you mean? by that. so everything we do in the cloud, all these instances, even pricing is an api call, which means i think i see where we're going with this. we can build a tool that well, that gives us the proper two instant size. we can take a bunch of metrics ins, instance types and shape the proper instance. we use instance recommendations powered by machine learning, machine learning jv. i maybe, i don't know. um yes, that's great.

do we need for that? well, i remember i took a machine learning course and in, in that course it's only about sample size. so if we had millions of hours and instances across lots of different workloads, we could come up with some pretty good estimates of which workload runs where at the best cost.

yes. well, so. ok. ok. good, good, good. we need a lot of data for that, right. yes. millions a month of, we'll get that eventually we'll get there. so. ok, we need to build a spec, we need to build an api. yes. you know, we have to do api infrastructure infrastructure. oh add machine. we did learn machine learning. add that to the list. we need to learn machine learning. of course, we need to uh oh my god, all these things, you know what we can, we can name it. i, i think i've got something, the ec2 auto compute server a i machine or eki sam. what do you think? ysy sam? yes, i love it. um shall we start working on it?

you know, raul's been bugging us to get v two of his. but this would be a great side project. and how long do you think it will take us?

steph told us, cost optimization is more important. so we should spend 100% of our time, 40 hours a week on this and we should get done with yaky sam in like a couple of months. if this, this investment of time will pay off eventually.

exactly. exactly. and it'll be fun. it's gonna be fun.

hey, guys. oh, she's back seth. i couldn't help but overhear your conversation and i hate to break it to you. this tool already exists it, yes. where is it?

well, we have something called the aws compute optimizer tool better better. yeah, i think so.

so what this tool does is it doesn't even just look easy to

It looks at all scaling groups, lambda licenses and s and provides you with right sizing recommendations straight away. You don't have to do any of the work. And how much does this tool cost? Well, it's actually free. The first version gives you free. I do recommend using the free one first. It gives you a 14 day look back period on a step, step step.

There's an announcement on sunday. It's 32 days. Now, why do they do this to me? Change my thing that does it work on, on all instance types or just yeah, and everything we even cover gp.

That reminds me darker, those gp u instances we know we can launch a cluster of them. Stephen. Stephen don't even think about it. I have seen those gpu prices. You're not going there. We'll say we'll say we'll say gp for later. Amazing. Ok. So this is amazing. So how do we, how do we get started? Is there? What's beyond the free version?

It's, it's brace yourselves. It's 0000366 cents an hour. Ok. I don't think yak sam can beat that. I don't think. Ok, maybe we get rid of yak sam. So while they get rid of that, why don't i just show you what you can actually do? So we're going to do a quick demo of computer optimizer. We used it before. Don't start with us. I want to see you.

So in my account, this is compute optimizer. Nice free tool for you guys to get started. So what i always like to do is showcase how i would go about finding an optimization opportunity in my account. So if i go in here, i can see immediately from the dashboard, i can save a whole $73 in my account. Uh and if i scroll down, i can see different recommendations you guys were focusing on ec2. So let's have a look at those.

Now, when i do this. It will not only give me instances that are over provisioned, it will also show me ones which are under provisioned. So you might need to right size the other way as well as optimize. But let's have a look at one that is under provision. So if i click on the instance, it will immediately give me a recommendation of where i can change what instance type i can change to, to save some money at the moment, we have some with a g. If people don't know what that is, we're going to come back to it later. But maybe i don't care about these gs. Let's get rid of those. You know, where did the the data for that instance recommendation come from? How did it know you beat me to it?

So if i scroll down here, it actually provides you with cloud watch data. Now, what is really important about can we optimize it is when i was a customer? And what we hear from customers is when you provide a recommendation, you want to see why we gave you that recommendation. So the good thing about this is with cloud watch data, you can see a lot more information and you can even see my current and the option in here. So how will it actually affect my utilization? If i go back to the top, i can see lots of different types of instances that i can try using and what the price impact will actually be as well as things like the performance and the risk involved and the specs.

So giving you all this information will enable you to get started quicker. Does this only support two or can i do like rd in other instances, we can do, how about we do some lambda functions? Ok. Look at that. So if we go back to our dashboard and i scroll down to, i don't know what lambda functions are but doesn't matter. Ok? We'll come back to la. So when we do this lambda, i always like to showcase lambda because people think it's very cheap. If you don't, what lambda is, we're going to cover that in a second. But it's really good a way to recommend the sizing and the configuration because often you can add a little bit more storage to it and leave me more cost efficient to do that.

So this is all well and good. This is all this information in here. Very handy. You can have this in every single linked account and all you need to do is enable this at the organization or at the linked account level and everybody will get it. It will work for majority of our regions, always check your regions, but it's very easy and simple to get started. However, i do know from experience of being a developer and an essay that having to log into this tool when you have all loads of different dashboards has been quite annoying.

So i wanted to highlight something that maybe who didn't know about application manager. So if i go on ssm and i go on application manager, which you can find on the left hand side of this panel, what this will do is it actually connects to my cloud formation templates and with a tag, i can actually get cost data based on these. So i know one that i'm gonna focus on which is my stressed instance. And if i pop in here, i can see a bunch of different information, i can see monitoring insights and i can see costs.

Ok. Wow, i love first of all stressed instance, my favorite type of instance. Amazing. So, so you can actually have a look. So is the computer optimizer only looking at compute resources and cloud formation template or all this will cover all of the cost in this section. But when it comes to optimization something i wanted to highlight. Whereas over here on the right hand side, if i click on cost, i can see my spend.

So why i really wanted to draw into this is an easy two or a lambda function is just one part of a cloud formation system sometimes. And so you want to see the whole thing, you don't want to have to go into all different places. So do everything a cloud formation creates darker now that i'm thinking about it. I use gen a i to write one of our cloud formation templates. I don't remember what instance size it shows. So this is a perfect way to detect those kinds of things. Ok, perfect. All right. Thank you.

Should we get going with this thing? Let's do it. Awesome. Awesome. We're off to optimize one month later. Hey, steven. Um, what's going on? I mean, the thing is, i feel like after the compute optimizer after dashboards and shutting, shutting off your 168 servers, um, we, i feel like we're fully optimized. What, what's, what's there to do? I'm feeling the same way. I mean, we've got cost explorer now. We can see how much things are going to cost in advance. That's kind of cool. That's amazing. Yeah, we don't have to be scared of the bill. We, we know auto scaling now. So scale stuff we don't use and compute optimizer, optimized stuff. We don't need, we need to optimize, we're using the right instances at the right. I still kind of think esam would have been cool. We put the back burner. So what, now, what's, i mean, what, what are you gonna do for the rest of your life?

All right. Well, i've got this, this problem that's been bugging me. Ok. We've got that api end point. Ok. Now, sometimes it runs like 23 times an hour. Sometimes it is 1000 times an hour. Now, i want to pick the right instance. I don't want to over provision. I don't want to make that mistake again and you don't want to shut down it. No, but i don't know what the right size for this could possibly be. So, you need to figure that one out. So that's, that's one thing, you know, i don't know, do we plan for the worst case for the average case? That one is really bothering me. What about you? What are you up to?

I need to do what? Everybody's favorite job is patching databases. Yes, i have a, you know, those sql classes i have. Yeah, i need to go patch them this week. I need to go make some changes to reboots, all those fun things. There's a lot of stuff there. It's still running on the cloud. Mind you. Right. It's not an easy two instance. It's wonderful, but there's so many things to do. It's such an unmanageable service. Yeah, that's a good point. And speaking of services, that was that thing on our to do list that raul was telling us about. We need, we need search, we need search for our app and i was looking at a couple different options and we've got, there's all kinds of different services we could use, but whatever one we pick, we're going to have to manage it for the foreseeable future. Right. That's the service we have to manage and it is such a, i'll put it down. It is such a heavy lift to manage this and it feels like such undifferentiated work. I use this term a lot now. Undifferentiated heavy lift. Yes, absolutely. It's totally normal english. Uh yes, but i've been using that term a lot. This is what this feels like.

Well, so i, i, yeah, we, we need to, we need to get manage databases, deploy this big old search thingy, which is not going to be fun and we need to create these things. You talk about something that sporadic function function you call it, right? How do we, what do we do? I wonder is there something that exists that can help us optimize that?

Oh, hey, guys. Oh, no. So listening to your conversation eavesdropping again, steven, that sporadic workload. Why don't you try something called lambda? And it must be the jet lag. I'm sorry, isn't seattle in the same time zone as vegas but it's further from the ocean. Sorry. That made sense in my head. Ok. Ok. For sporadic workloads, ones that do run on a chrome job or low complex code, check out lambda use serverless. You don't actually need to use an ec2 for every single thing. Same for you. You don't need to run twos with your database. You can use that managed service. So check out things like rdf, your managed services allow aws to do all that boring patching and management of it. We can take care of it. Ok? Can i have a server database as well? Yeah, there you go. Look at that man. It seems like less of that lifting to do anymore. You're gonna put your back out of all this heavy lifting.

So and finally opensearch just wanted to comment on that. You can use opensearch for that search plan and a another managed service that will be easy for you to use. The best part of all of this is you can use graviton for this based on arm. Have you guys heard of graviton? Oh, yeah, it's the wonderful little r mc pu right. Yeah, i don't even know if i've ever used an r mc pu before. Look at your laptop, man. Look at your phone. Yeah. Sm one. Thank you.

So, graz on is a great service that aws created the processor and it gives you an easy two of 40% price performance. So 20% of that is cheaper than intel, like for like and 20% improved performance. I've seen so many customers, especially on managed service, be able to downsize their resources just by moving to graviton and all you need to do is change your instant type you saw earlier, i got rid of all those gs. You can add that back in and try graviton. So simple, so easy cost effective. Honestly, from today, if you guys can go and do one thing, i would move your manage services to graviton and save instant amounts of money.

Seems like we've got a lot to do. I think. I think so as well. So a lot so steven uh let's go ahead and set up manage services, serve us and do this type of thing. Graviton four is launched by the way, just so, you know, amazing. Ok. I guess i need to show you something. So i was looking at this tool called trusted advisor. I thought we had like 5, 10 things on our to do list 27 here. You know what if there was a tool? I mean, how are we going to get all this done on our own? I have no idea. There's only two of us man and i need to work on that yam version five. I don't want to work on this man. What if there was a tool that could get that done for us? What if there was a like a roomba for your or not like a robotic vacuum that could wander around, pick up all your idle instances, all your idle resources. What if there's a tool that could do this all for us be like, hey guys who could use cloud fix? Ok?

Now you didn't expect that you would come to a breakout session hosted by cloud fix and not get a demo, right guys, a demo for demo, please. Ok. So cloud fix is a tool that will solve all of these problems that you have. Ok. It's really simple. When you log into cloud fix, you see a very, very simple screen. We want you to focus on three simple things. The first one is potential savings, which is that big dial up front. When you run a scan with cloud fix on your accounts, it goes and looks at all the different places where there is waste and it will collect up all the potential savings over there and display it in that one dashboard. The next two dials are your metrics. It tells you how much of the savings you have acted on and approved. So please make sure you've got all of them done. And please remember that all of these potential savings are analyzed. So they accrue month over month, week, over week. So the last one actually tells you how much you've realized and that number will keep going up month, over month, week, over week, that's what you actually put in the bank

Ok? Now, once you've got that, you can see that the next section is really simple as well. We made it really easy for you to consume this data. The recommendations are all categorized as easy, medium and advanced.

The easy ones are the really trivial ones where you don't have to think about it. They are completely non disruptive. There are changes that will just, you know, get you the savings. It doesn't cause any disruption in your service. It doesn't need restart of instances, even ARNs on the resources will not change. So you can close your eyes and just turn them on. They will start saving your money right away. Ok?

The medium ones, you might want to do them once in, you know, in at a time where you're watching it because it might need an instance restart and the advanced ones require you to do this during a maintenance window because it might cause disruption. So we had made really easy to make decisions and act on it immediately. Ok?

So let's dive into some of the easy ones. You get a list of all the recommendations right there in the dashboard and all you have to do is click on the recommendations. You go to the available to execute tab and all the lists, all the instances are right there. You can select all and then hit, execute.

Well, I've got a question for you. Right. So I'm, I'm starting to wrap my head around all this cloud stuff. I was talking to Steph and she said, don't just give out the root credentials to your AWS account.

Yeah, it's a good advice. How do, how does this make changes into my AWS account without if I don't give you the root credentials? And I don't want to give you those.

That's absolutely right. You should never ever give out your root credentials or your AWS keys to the tools. So to make resource changes. So when you hit execute, CloudFix takes you to Change Manager. AWS Change Manager is the AWS recommended way of changing resources within your accounts. The reason to use Change Manager is that all the change requests that come into Change Manager have audit trails, you can audit it, you can see who made the changes.

Most importantly, you are in charge of approvals within Change Manager, not the tool. The only permissions that CloudFix has is to take change requests and push them to Change Manager. You have to log in to Change Manager and then approve them. So it makes sure that you're following AWS best practices about how you make changes to your infrastructure. And it's really that simple and Change Manager can be automated. So when you hit the proof, an automatic rook can make those changes.

That's going to make things a lot easier. Exactly. No more. Click it to click, click, click. Absolutely.

Well, here's what I saw on the easy list or there was a, there's something about instance resizing. Now we just learned about Compute Optimizer. So, are we reinventing the wheel here? Are you doing instance resizing and Compute Opti how does that work?

Yam. I think you got to bury ys for a bit. So interviews. Computer Optimizer does an amazing job of being very conservative in their recommendations. They will make sure that whatever recommendations come up are ones that are safe to execute. And so CloudFix does not re invent the wheel. It takes the recommendations that come from Computer Optimizer. And the amazing thing from the Sunday announcement is that you can now customize what recommendations Computer Optimizer produces.

You can go to the recommendations in Computer Optimizer. You can select what headroom you want in your instances to always be available. Once you configure that the recommendations that come from Computer Optimizer will then be used by CloudFix to actually implement the transition. So think of these as finders and fixers Compute Optimizer is actually the finder. It finds the recommendations and the easy way to implement that fix is CloudFix.

I get it. So Computer Optimizer is doing the recommendations that is then brought to us. So we just click. Ok. And then it does, it, it'll do our instance sizes. That's basically right. It's really simple. All right, I've got another question for you. I'm looking at these realized savings. These are realized savings based on what we would have spent on AWS. So this is almost like money out of their bottom line. AWS must be furious at us.

I would vehemently disagree but Steph you want to jump in. Yeah, I'll come back in. So customer obsession, you probably hear us talk about this a lot and we do really care about you guys and we want to make sure that the reason we built tools that can be optimizer and trust advisor is to make sure that you are creating the most efficient workloads in the cloud. So we wanna help, we wanna make sure businesses thrive and make sure you optimize as much as possible.

Ok. Well, I think I have one more. Oh, here's the one I was really meaning to ask. So another thing you told us about was we want to look into serverless will CloudFix. Just rewrite our app into serverless for us. That would be cool.

CloudFix doesn't solve all of your problems. That's where you come in. You guys should be spending your time working on changing code, taking your applications from monoliths to serve and doing all of that hard work. That's hard to do. So spend your time doing that. Don't spend your time on all the mundane fixes that have to be done continuously. With the free time we have, we can do that makes sense. We have lots more free time now.

Ok. Well, how often should someone run this? Is this a once and done or how does this work?

Ok. So please remember that your cost optimization effort is nonstop. There are multiple variables. Number one is that your developers are constantly upgrading code, pushing things into production. There are new resources being, being created all the time. So you're gonna have to stay on top of it continuously. So your anti synapse operation or your cloud optimization cost optimization exercise is a nonstop exercise. And so is CloudFix? CloudFix runs continuously.

The second variable of course is that AWS is innovating all the time. The new features, new services are being created all the time. Just in the last 72 hours, we've had 150 feature announcements, which is pretty remarkable and you have to stay on top of all of that. What was optimal? Three months ago is probably not the most efficient way to do things today. So you have to stay on top of it and that's why you need automation to do this at scale.

Well, we can't scale just the two of us. So automation makes a lot of sense. How do people get started?

Ok. Getting started with CloudFix is really simple and we'll just bring this up on the screen so you can visit us at cloudfix.com. Really simple to get started. It's a self service page. The assessment, you can get started in less than five minutes. It is incredibly simple. Once you have the assessment, the assessment is absolutely free. Ok. We want all of you to know how much waste there is in your account, make it really simple for you to understand that. And once you decide that you want to act on it, you can sign up in the marketplace and get started. So really, really simple to get started. You can also visit us at booth 1529 on the expo floor. There's an amazing team over there and they'll walk you through what you can do to get started.

One final question. Have you ever scanned an account? And you run all your finders and fixtures on it at the very end. It says account perfect. Totally optimal. Nothing to do here. Go home. See thumbs up.

Definitely a pipe dream. I have never seen that happen for a simple reason, like I just said, that might happen the day. AWS stops innovating and I don't see that happening in any time in the future. So as long as AWS is innovating, there's always going to be new stuff coming up, there's always going to be a more efficient way of doing things. And so you need to stay on top of it. There's always going to be things to fix, got it. So it's a moving target and that why CloudFixing help keep up with that moving target. We're always going to be there to add more finders and fixers help keep you optimal because what is optimal now AWS are going to innovate and then it might not be optimal a few months from now. Exactly. Got it. Thank you.

Well, amazing, I think. Yeah, go check a call at the expo floor. What else do we have? Yes. Let me, let me give me that. I wanna, I wanna take off my developer darko hat for a second here and talk to you all a bit now, a couple of things we learned today, right?

Um um see your spend right. We you you may get excited by doing things in the cloud. Remember everything is an API call. It's so fast. He did it with a CSV file. It's so fast. You want to do things you wanna just build and build because it's so easy. It's basically limitless but make sure to see your spend, understand your cost, use Cost Optimizer, check out your plot formation template. It is gonna be too pricey or not. Remember everything in a cloud can go up and down. It's not like buying a server and just well spending the money on it if you don't need it, shut it down. If you don't need it, delete it. If you don't need it, pause it, you can be elastic in your applications. If you have a spiky workload, the cloud is the perfect place for you.

The cloud is I don't want to say perfect. It's far from perfect. The way we give you the Lego bricks to build your infrastructure is amazing. However, there are so many tools out there, both first party, third party and open source tools that can help you build things. So from things like Cost Opti Cost Explorer and Compute, Optimize to the wonderful folks from CloudFix, use the tools at your disposal to make your life easier because hey, you don't want that bill we just pulled out, right?

Um finally, we all use the cloud. We've all ran into issues. We all have experiences like this. Check out the community, check out the community builders, talk to me, your friendly neighborhood developer advocate, talk to Steph and help let us help you make things better. Check out your local the FinOps community as well. It's a big old thing and of course check out your local AWS meetups. It's a great place to learn things and ask questions.

And finally we here that gentleman, the most amazing staff and myself have a couple of things to talk to you about. I run a live show if you want to see more of this, more of this bald head and the serious jokes. I do the crazy, crazy shirts and all that great stuff. You can check out my show on Twitch. It happens every Thursday at 9 a.m. Pacific on Twitch dot tv slash AWS. It's called Build On Live. It's basically I build tinker break cry laugh when I try to build stuff on the cloud. It's great. Join us. It's, it's like a real show. You can come join the chat, make fun of me suggest things. It's amazing. What do you guys have?

So we have a live stream called AWS Made Easy. It's every Tuesday at 9 a.m. and we've also started a Friday uh generative AI hackathon series. One of the things we've been experimenting with is taking all of the transcripts from our previous episodes using AWS Bedrock to build a Raul bot rat. So we can query the Raul bot on cost optimization questions and it'll give you, give you the answers. We've been building that up over the last couple of weeks, you can follow along.

Um now that Bedrock knowledge base is out, we want to use that the AWS Made Easy livestream. It's really three main purposes. We have interviews with AWS folks and other 80 people in the community who are using AWS. We bring them on our show. We just have a lot of fun talking shop. We answer people's questions. So if you, you know, people can write in have questions, we have a good discussion. I want, we want it to be like the Car Talk. If you remember that show, the Car Talk of, of AWS, just have a good time, have a good discussion. So we have guests, we have interviews and then yeah, we really enjoy, have a good time sharing our content and we also rate and review all the AWS announcements. So we, we rate these announcements on a scale of one cloud to five. Try to parse you know, this fire hose of announcements that's coming and figure out, you know, what are the best ones to, to have a look at what is going to affect you from a cost perspective.

We've been doing early morning live streams at 6 a.m. so if you're up early due to the jet lag, please tune in at 6 a.m. We're doing keynote coverage and finally Thursday at 4 p.m. we have a great panel of guests to kind of wrap up reInvent. We have Jeff Barr who he's going to give us a an overview of what reInvent was like. We have Rick from Compute Optimizer. We have Lorenzo Winfrey from EC2 Spots talking about Auto Scaling groups, all the neat things with, with Spot. And then we have Yuri Pardo from the Cloud Cloud Intelligence Dashboards. So we have a great cost optimization focus panel. So if you're gonna go crash in your hotel room around four o'clock Thursday, turn us on or just watch it later, follow us on AWS Made Easy or you can look at the CloudFix LinkedIn or YouTube Twitch. Wherever live content is we're there.

And finally, my one, the Keys to AWS Optimization, a show, we share stories, concepts and solutions to help you unlock your optimization opportunities. AWS. It's on every Thursday as well. Twitch dot tv slash AWS just before dark. I see the whole thing.

And the reason we wanted to, to mention these is because this has obviously been a very silly and thank you for bearing with us production of how to optimize, but we do care about helping you optimize my job is to help you guys save money and these resources will really help you do that and they're on every week and we always find different tips, whether it ranges from any of the things that we've said today all the way from using Cost Explorer down to migrating or optimizing the service. So we recommend checking those out.

Finally. Thank you again. Thank you so much. Thank you so much for watching. We hope you've enjoyed it. Feel free to message us LinkedIn.