Wirelive: Episode #15

'AWS Cost Optimisation'

Thank you for watching!

If you're interested in the book that was referenced throughout this episode, it is for purchase available here.

Still, want to know more about AWS? Check out more of Wirehives resources here.

Show Notes:

Eli Mansoor | LinkedIn

Rob Harding | LinkedIn

Yair Green | LinkedIn

Mastering AWS Cost Optimization: Real-world technical and operational cost-saving best practices (Second Edition) eBook: Mansoor, Eli, Green, Yair: Amazon.co.uk: Kindle Store

Amazon Web Services (AWS) - Cloud Computing Services

AWS Free Tier (amazon.com)

https://www.wirehive.com/thoughts/cloud-consultant-aws-view

Rob B 0:06
Hello, everybody, welcome to Wirelive Episode 15. We are here today to talk about all things AWS, and specifically, AWS cost optimization.

For those of you thinking, What on earth are those things? Well, you're gonna find out a lot more about those topics in the next half an hour, for those of you that are thinking, yes, finally, I can get a handle on my AWS costs, this is the place for you. We've got a number of amazing guests with us today, including a sort of co-host from the wire hive team, also called Rob, just to make it really confusing, affectionately referred to as young Rob within our team, which I suppose makes me old Rob, which is a little bit depressing. So yeah, we have Rob Harding, one of our AWS Solution Architects from the Wirehive team, who leads, you know, a number of our more challenging areas, projects and works with customers day in day out to architect their solutions. And, and the cost is definitely one of the components that comes up a lot in his day today. And we're also joined by two people with names that I could get completely wrong. But I've, I've made sure that I'm going to get them right. So I'm now going to do a terrible job of getting them right, as we're joined by Eli, and Yair, who are working on their day off kindly,

who are both masters of all things, Amazon, and there might be a book that we're going to talk about that they may or may not have been involved with, which we think you should all buy as well? So yeah, let's, we'll dive straight into the moment and let the guy say hello, we as we get started. But just, I think, to set the scene. AWS is obviously one of the big cloud providers that everybody's using these days to run their infrastructure, whether it's something basic, like a website all the way through to a complex business system or back end application. And the promise of cloud, one of the reasons why it succeeded really and is continuing to dominate was utility billing and this flexibility of cost. So the idea, the kind of basic premise with most cloud services is that you pay for what you use. And in the olden days, which was only five years ago, in technology, and people used to buy servers or rent servers from companies or buy them themselves, and you had to, you had to buy something, or rent something, and it was always on, you paid for it for a month or for a year, whether you were using it or not. And so cloud turned up and provided this amazing toolkit, this rich tapestry of different products and services. And I'm sure he can tell us but AWS last I checked had over 100 different components and services. Now, it's probably way more than that now. And each one of those does something really specific in a really clever way. And you can bolt all these tiny little pieces together. And in theory, pay, in some cases by the second or by the minute for what you use. Now, that all sounds great, and it is great. But one of the things that happen, and this is very common is people adopt these systems. And then they get their bill at the end of the month, you know, a year into the project. And it's not quite high, because they've got all of these tools and systems are amazingly capable. And so you turn them all on, you start using them and without really thinking too much about what the cost impact might be. And you find yourself in this position where you've got this amazing infrastructure that's doing all this cool stuff, but it starts to get quite expensive. And so there are a number of different tools and techniques that you can use to optimize costs. And we tend to see that come up, you know, six to 12 months into a customer's journey with AWS. And so Rob Harding young Rob spends his days helping customers with this challenge and stumbled upon this book that the chaps were were involved in putting together and found it incredibly useful. So he suggested Well, why don't we get them onto a Wirelive? So here we all are, that hopefully sets the scene and I'll pass over to Mr Harding to introduce himself and kick things off.

Rob H 3:56
Brilliant. Thank you, Rob, thank you very much for the invite. lovely to be here and to talk with things my native Yes, it's one of my, my loves in life. And so yeah, I do as cloud solution architect here at wire hive. So I deal with a whole range of customers across a multitude of technologies. And of course, the cost is one of those subjects which is reoccurring and genuinely comes up as one of those pressing matters for all our customers. So it's always a challenge to kind of work out what is right for the customer, how much they want to spend, how do we weigh up all these various services to give the best value for money. And I was on LinkedIn one day, looking through my feed just kind of scrolling through. And then as it so happens, there was someone recommended around a book mastering address cost optimization, this book here, which is kind of like 400 pages of absolute gold for some really insightful information which we're going to go and talk around today. I think for me what was I've read about

You know, read the book many times. And the most important thing for me here is its validation. It's a validation of how complex this topic is, and how much how many different facets and moving pieces there are to the world of cost. And as I say, like, customers have very different needs and wants for cost-saving. So, yeah, I throw this kind of like them to Eli first around kind of what, what brought you to, to wanting to write this book and pay everything down on paper?

Eli Mansoor 5:33
Yeah, thank you, Rob. And so actually, this

takes us four years, back when I was with cloud health, in Israel, leading cloud of business in Israel dealing with supporting customers, AWS customers, or provider costs. And

I have my dear friend, Yair, here is my colleague is back then it was the VP of engineering, and offer my and ad network company. And he was actually managing

billions of dollars of spending on AWS on a daily basis. And when we group together to discuss how to optimize costs, we found out that it's much more than a one-time activity of, you know, purchasing reserved instances back then. And we've thought about it, that it's actually a philosophy, it's a, it's a process, it's an ongoing activity. And there is a lot to know about it. And it actually impacts how you architect your environments on AWS. And it also impacts how you operate them on an ongoing basis. If you just look at it as a one-time activity, yeah, that's not good enough, you actually need to implement best practices. And we sit down together and we thought about the structure, we thought about what are the most important and services that we should focus on. And, you know, a year or year after the pastoring, AWS cost optimization, the first release, was published.

Yair Green 7:21
Yeah, and it was mentioned in here that the K Oh, methodology, cow methodology, which stands for knowledge, architecture and operations, we introduce this methodology in the book. And it gives our readers a well-tested methodology to apply in their own cloud environments. And for each service they use, you can actually get familiar with its pricing structure, get tips and tricks on how to reduce the cost. And to apply this methodology for this specific server. And by understanding this methodology, you can also apply it for other services that are not covered in the book yet. But since it's a generic methodology, you can use it everywhere. And by the way, in any cloud.

Rob B 8:27
Okay, so we won't, we won't speak the competitor's name, obviously, during this call as AWS lovers, the three of you, but good to know that the, you know, the methodology does apply to other services. And Rob, perhaps you could add a little bit of colour on, on where you found, you know, what, you talked about validation? Like, just for our audience, perhaps you could just add a little bit more context on your own relationship with this conversation about cost optimization. So, you know, maybe talk us through like, an example of where that comes up, typically, in your day to day like, when does cost surface in a customer conversation? Is it at the beginning? Is it a year into a relationship? Like, how does it tend to come up?

Rob H 9:15
Absolutely. Um, so I think when I talk first around the validation, and I kind of recommend people to have a look at the AWS networking costs, the various things around data transfer, Cloudflare cloud front, excuse me, Elastic Beanstalk inter-region versus a cross AZ costing. It's a minefield, you guys have done a diagram. I think I just put it up here. It just showed the complexity of how difficult this kind of thing is. So that's where kind of the validation comes from. And trying to work this beast out of a dress cost and get better at it. And where it comes up with customers in the conversation. It's generally we talk around the requirements, we talk around what solution they want some a technology, and then it's okay and how much it costs me. And how can I save money when we're doing Greenfield deployments when it's a new customer as an existing customer that we've worked with before that has an existing infrastructure within AWS is generally one of the earliest conversations we have with our customers. And I know we're going to talk a bit further around like the security versus cost aspects later on. But yeah, the cost is one of those early questions, especially when you're dealing with cross, or board members, they want to keep costs down. Definitely.

Rob B 10:43
Yeah, okay. Great. And so let's, let's move on a little bit. So, Eli, if we come to you, just maybe for those that aren't, aren't really that familiar with AWS? Could you talk a little bit from your perspective about just in broad strokes, where you think AWS's capabilities are really exciting, and perhaps you could outline some of the virtues of a cloud model in the first place, right? Because there will be people watching this here who still haven't taken that leap? Or maybe you're considering AWS versus other options?

Eli Mansoor 11:15
Yeah, so AWS is a subsidiary of Amazon, the big Amazon, and we're providing demand cloud services in a page, you go in the pay as you go, model, so actually, and customers can leverage over 200 services like compute databases, storage services, IoT, Ai, ml, you name it, and via web services, or via the AWS console. So at the click of a button, actually, you can spin up infrastructure over 25 geographical regions. So actually, we are covering the walk also. So you can then go global in seconds. And we have over a million customers. And those can be banks like Capital One, those can be technology companies like Netflix. And those can be also small startups that just start their operation, and they don't want to spend too much money on Compute capacity that maybe they will use in the future, maybe they will not. And I think the main benefit that we offer is the

the wide range of services that are integrated to one, one to the other, and the ability to go global in seconds. And then need, you know, there is no need to plan for capacity, because you can scale up and down automatically, as per demand. And also, I think, the wide partner Network, for example, like Wirehive, which are supporting our customers in their journey, and to leverage those services and to benefit from so we actually we help our customers innovate faster. And we have partners like yourself, who are supporting our customers to achieve this goal quickly.

Rob B 13:29
It's a great, great answer very compelling set of services, as we've outlined. And then and Yair, coming to you from your perspective. You know, you mentioned the methodology applies to other clouds, but you've really focused on Amazon with your career in technology. Why, from your perspective, do you feel AWS is the right choice for our perspective, project or for somebody considering using it in the first place?

Yair Green 13:53
Yeah, so definitely AWS is my choice. And I managed large scale global environments for different companies such as Fiverr, Matagami. And currently at Site VC.

All running AWS workloads. The main benefits of using AWS are the web services available for us automatically via web services. And globally, as Ellie mentioned, the global footprint, allows us to spin the environment in many different regions around the globe. And this brings a huge value for the organization that they can scale up and down whenever they want, and merge the infrastructure to the load and retiring the need to purchase you know, capacity and advanced and it also speeds up the time to market and it helps you you you can never run out of computers. And also each team can get their own infrastructure that requires required for their own job. And if you know how to set your environment in cost of cost-effective way, you can make the most of each dollar you spend on AWS and become cost-efficient.

Rob B 15:25
Absolutely. And I suppose that's ultimately the premise for why we're all here today. So, so moving, moving on and coming back to you, Rob, let's talk a little bit about this cost optimization topic versus some of the other areas because when, when somebody is considering using AWS or is already using it, there are a number of different factors that people consider and the cost is definitely one of the big pillars. And there's a framework that we go through with a lot of our customers and partners called the well-architected review framework, which is an Amazon standard that white hive is, is part you know, we're entitled to run these audits for our customers, it's something that Rob spends a lot of time doing. And so cost is an area, the security is another, which comes up a lot. And there's this interesting sort of relationship between cost and security. Because in theory, you know, that all about I mean, I'll let you outline it, Rob, but let's, let's get into that. Right. So, so talk through your perspective of the different pillars, and how cost kind of factors into that.

Rob H 16:28
Yeah, brilliant. So yeah, from the well, artists view, we have the five pillars that where we review a workload against so we have reliability, performance, efficiency, operational excellence, cost optimization, and security. And it's interesting to four of the pillars versus because of cost optimization, how you balance that, and what's right for a customer. And I have this belief that if you want to be the most secure, you cannot be the most cost-optimized, because those things aren't combined. Why do I say that? If we think around security, from an application point of view, you may want to put that in a private subnet with a NAT gateway.

From a security point of view, that means the application is isolated and does not have public accessibility. Alternatively, if I was less secure, I could put that in a public subnet with direct public accessibility. And so where does the cost aspect come into that? Well, if I've used a NAT gateway, my costs would generally be higher, because I would need to account for a NAT gateway charge. And also the data transfer going through a NAT gateway, where I would have a different set of costs and a smaller set of costs if it was in a public subnet and being accessed directly. So it's kind of how do we balance these things? What's, how do we look at these pillars and security, especially versus cost? And what's right for our customers?

Rob B 18:01
And I know that I know that I really feel strongly about this one because we talked about it in the setup. So, so early. This is where you get to tell everybody that security can have no compromise. But what's your perspective on this from any sort of cost versus security debate?

Eli Mansoor 18:16
I can say what I tell my customers, so I tell my customers when they onboard to the cloud, that they need to have a security person to make to ensure that first security is the top priority. So there needs to be either an expert within the team or a report there to support them with this activity. And once we ensure that we are secured, there, let's deal with cost optimisation of the services that are being used. Of course, if you are using a firewall or Nat gateway or any other AWS service, it has a cost. But the good news is that every service you use has the way to you have the way to optimize it or to use it cost-efficiently. And to ensure that you are using it you are being secured. But you do it cost-efficiently.

Rob B 19:16
And Yair. Is there anything you would add? I mean, I think you know this, how do you tackle this, like there's this concept, you've been responsible for running some, some very large workloads, security, obviously, there's like, you have to get to a minimum standard, that is good enough, but then there's always more you can do and those things tend to have cost implications in some cases, quite big ones. How do you wrestle with that in your own mind and when you're talking with others who use the services?

Yair Green 19:44
Yeah, so in my job, I'm responsible for both the security and the cost aspects of our environments. And security is always our top priority, as it's crucial to secure our platform and the customer's data.

The interfaces. So of course, there are costs involved in using AWS security services, such as the guard duty, guard to and others. But as always with AWS, there are ways to optimize these costs. Sometimes specific configurations are bad, both security and cost-wise. And a great example is what Rob Harding says, about using external interfaces within AWS or a C two interfaces.

That both, on one hand, its security might be a security vulnerability. And on the other hand, it's cost inefficiency. So yeah, definitely. These areas are the top priority of for us.

Rob B 20:54
And then, and going back to you, Mr Harding. Any other sort of observations from doing these? Well, architecture reviews, time and time again, you know, with regards to this cost question, I mean, you know, security, obviously, is one of the pillars How do you find cost interacts with the other pillars?

Rob H 21:10
Yeah, so I did some analysis of our all the well, Arctic ruse that we've done over a six month period to kind of work out, what do our customers really want. And it turned out that security and reliability were kind of most commonly requested, these are priorities and cost actually sat in like, third place below them. So still quite an important pillar. But you know, as the gentleman Bradley says, you know, the security-first mindset was definitely there. And I always get told, you know, I want the most secure, I want the most reliable service, and I do so but I'm not going to write you a blank check for you to do that. So it is always a difficult one, to understand how much you can kind of get away with and how much funds you have available to work with some of these services.

So, you know, there are some smarter things that we've looked to do for some of our customers in terms of

turning things off that aren't needed and scaling down overnight, having smaller instance footprints, you know, right-sizing is a big piece of what we do. So take easy to use RDS databases, elastic cache, all these kinds of services, where you have an instance type.

When our customers start out, they start with, you know, I think I need this amount of CPU, this amount of RAM. And then how, how often do you go back and evaluate that? Because they're constantly changes and your application performance today versus six months or a year's time could change drastically whereby a smaller instance footprint is all you need? So yeah, we are constantly battling these, these other pillars against the cost for our customers.

Rob B 22:55
And then a question that's coming from the audience, which we can maybe open up to, to anyone who had an opinion on it, there's, there's a kind of growing interest in the world in the environmental impact of technology, which I think is great, you know, personally, it's a cause I'm very, very closely aligned with some of my other ventures. I think that the sort of, you know, the other advantage of right-sizing and being cost-efficient, is you're not wasting resources which has a carbon footprint, right. And I think that it's a really interesting component to this conversation, which is starting to become more relevant. And maybe we just get everyone's opinion on this one by one. I mean, why don't we start with you, Ellie, and then we'll come, we'll come around to the others. Is this something that you're seeing come up more is? Could you talk to this subject a little bit about the kind of, you know, the environmental side of cost optimization? What's your perspective on this topic?

Eli Mansoor 23:50
Yeah, I think that's something interesting about AWS, that when it's bringing innovation to the market, we do our best to provide our customers with the innovation and the better performance, and actually also for lower costs. So actually, if you sometimes if there is a new service coming or a new model coming usually it costs more. In AWS case, it's the opposite in most cases, but in most cases, the bag familiar with so we are trying to be cost-efficient is as well as resource-efficient. So you will see that for example, we introduced a new process or reason recently the Graviton processor that is based on the arm. So you will see that it's providing you as a customer better performance, but for lower costs. So I believe also behind the scene. It also requires I have less power and less, you know, resources to to to enable it. Yeah.

Rob B 25:08
Yeah. And a great example, you know, as technology moves on in theory, it becomes more effective, more efficient. And hopefully, we will benefit from that, as you say. And, and yeah, what about from your perspective? Is this something that you see coming up or something that you're starting to consider with the workloads that you run?

Yair Green 25:28
Yeah, definitely, yes, I can say for my experience, over the year, I was a consultant for many different high tech companies, both here in our local environment, industry, but also in Europe and the United States. And I can definitely say that over the last few years, I do hear more and more questions about how the environment affects the global environment and global warming and stuff like that. When we plan, we, of course, plan for job security, and to be cost-efficient, and then to be also environmental friendly.

Rob B 26:17
Good stuff. And I you know, I do think that that's the sort of silent benefit of being cost-optimized anyway, as we've talked about, and, and back to you, Mr Harding, something you've bumped into much or has come up so far with our conversations.

Rob H 26:30
Yeah, I think one of the conversations I had with a customer became part of the year, that they actually specifically looked to AWS because they were doing their own research. And they're going from on-premise to a cloud provider, and from their research, basically looked at EBS because of the environmental impact that they had, they had looked into so it's definitely something on the forefront of my customer's mind sighs speak to

Rob B 26:56
good, it's good to hear them then. So let's move to the kind of what's next. I think we've sort of, you know, without getting really into the weeds, which we'd love to do with anybody watching one on one, if they're interested, we're going to spare you all the kind of really deep dive stuff. But I think it's just good to talk about what happens next, you know, what's the future of this topic? Do you know, does the AWS platform eventually get to a level of sophistication where it can automatically cost optimize? Maybe? Or, you know, does machine learning get so capable that it can make such sort of swift and dynamic recommendations that all you have to do is accept them and you'll automatically get there or not? You know, the promise of technology is that in theory, it gets simpler to use over time. And I suppose I'll just open it up to the very dangerous game of predictions. And let's see what you all think. And Eli, where's it going? What do you think comes next in this world of cost-optimized cloud services?

Eli Mansoor 27:55
So I think the good the good news about it is that AWS is taking this topic very seriously. And you can see the number of new services and capabilities that AWS introduced with the cost management tool that is getting better and better on a daily basis with tons of recommendations about how to optimize a customer's costs, it can be recommendations for right-sizing like Ron mentioned recommendation for reserved instances and sane expense recommendations. And, and many more was setting and budgets, anomaly detection, and so on. So there is a lot that AWS customers can benefit from just looking at the AWS Cost Management Console. And I also think that they need to the interesting thing I mentioned earlier, the Graviton instances, which which are providing up to 40% price-performance improvement over the x86 instances. So if you compare the six g family instance generation, those are much more cost-efficient compared to the fifth generation of instances. And it's quite easy in some cases to migrate, from one instance to another, for example, elastic cache, Elastic Search and many managed services are built in a way that they can leverage these Graviton instances. So you can actually copy all that spin down. One one cluster spin up the gravitron powered one and you will benefit from any very interesting you know, very compelling cost advantage.

Rob B 30:01
So it sounds like your prediction is that customers will start to migrate workloads onto these newer, more efficient services. And along with all of the automatic recommendations that are happening, that that simple availability of newer, more efficient hardware really under the hood is one of the big drivers that you think we'll see, you know, cost improvement over the coming years. That's, that's great. That's a really, really interesting insight.

And Yair, from your perspective, and anything else? What's next? What, you know, he can go short term, he can go long term, however wonderful, and where do you want to be like what's coming next?

Yair Green 30:36
Yeah, I believe that the next thing is around accelerating the awareness, for cost optimization across the various teams within the organization, including the development teams, to ensure they develop, you know, the right code that is light on light, in a resources perspective, like CPU and memory. And also I believe, we'll see more automated flows implemented to ensure ongoing optimization and governance, like automated cost testing, as part of the development cycle and stuff like that.

Rob B 31:17
Very interesting, yes, sort of building into unit testing. And thinking of thinking about it as part of that DevOps workflow, I think, is a really interesting future. And, and those tools and systems are starting to get better out there to make that more accessible for development teams generally, which is great to see. And yeah, Rob, what's your perspective?

Rob H 31:37
Yeah, just on the DevOps thing, I think a big shout out to a tool called infra cost, which is kind of doing this kind of thing as part of your deployment sort of structure. If you haven't looked at it yet, definitely check it out.

I think what I seeing is, for an organization that has some level of maturity on cloud already, if they're doing anything Greenfield start to look at the more containerized or serverless offerings, because there's, there's a huge price change and cost change between a workload that's running serverless, or containerize, versus the more traditional EC to aspect. And, yeah, I think we're gonna see a lot more of organizations re-architecting, how they build their applications and go more microservice.

Rob B 32:23
Interesting. So So I suppose your prediction is that as organizations get into that refresh cycle, or on building the new and the next thing, taking on that microservices, or maybe, you know, lambda, or that sort of asynchronous function based architecture, and you think that has a huge cost efficiency and is something that that you will see more of in the future? Absolutely. Yeah. Yeah. Interesting. Yeah. And I know, we've had a couple of customers engaged in those types of projects over the last six months, and they've been very, very effective. So yeah. And I said, Well, I agree with all of the gentlemen here, unsurprisingly, as they're experts in the field. But and I think

my observation is that, if you look at, I think Slack, the tool makes this the most evident. If you're the admin of a Slack workspace, you get an email about once a month, that says, We noticed these five or six accounts on your slack instance, that you're paying for, didn't get used in the last two weeks. So we've temporarily disabled them for you, and we refunded you this much on your belt. And as a customer like that is such an amazing experience. And it builds incredible loyalty and slag have obviously worked out that in the long term, they will be more successful as a business and make more profit by giving back the fees for those services that people aren't using. And that's so unusual, as in, you know, in the recurring revenue business model. So many businesses rely upon, you know, underused subscriptions, dead capacity, really, and which is not in anybody's interest, other than the company and certainly doesn't take the box from the environmental perspective, as we've talked about, when you have, you know, cycles being wasted on on on useless workloads. You know, I think, my hope, and I think it's getting, it's going in that direction, you know, to realize comments about the, you know, the improvements in Azure as cost manager and those tools and recommendations like it feels to me like that's where we're going, we're going to get to a point where AWS will dynamically send you an email once a month that says, hey, hey, admin of this particular account, we spotted that these 15 instances weren't right-sized. So we've just dynamically changed it for you. And if they do need more resources it'll move them back up again, completely, transparently, you'll never see the difference. And we've reduced your bill by this much, right. I just feel to me like, eventually, that's where it's going to end up. Now. That's the very glossed over technical version because there's always detail to how these things can be done and then you get into the hole. Well, what if I don't want it to do that and I need to be able to disable it. And of course, there's lots of detail but it's

Right to see that cloud services are maturing to this point because I remember right at the beginning when AWS arrived on the market, we were all still putting servers in data centres. And that was kind of this was always the like, the panacea, this was the promise of what utility build cloud services would ultimately get to one day was this truth, like, a dynamic elastic cost centre that flexes based on capacity requirements and demand. And, you know, the word the sort of global side of it came up a lot in the discussion today about being able to put workloads where your users are? And so yeah, I think it's fantastic to see that this stuff is really maturing now. And also that resources, like the amazing but you guys have put together are available to make it easy and accessible for people to kind of get their heads around that. And on that note, if anybody watching this or listening later on as a podcast would like a copy of your book, where would they find it? What's the place? You would like them to go buy it?

Eli Mansoor 35:59
Yeah. amazon.com

Rob B 36:04
What a surprise!

Eli Mansoor 36:06
Go clock go global in seconds. So we published the book on amazon.com. And everyone around the world can purchase it. And we have readers actually from around the globe. And there is anyone from South Korea listening so soon.

Actually, these days the book is being translated to South Korea. And later this year, we'll have a South Korean version of it.

Rob B 36:38
Fantastic, and, and Rob, I know Rob Hardings, very interested to know, are you going to write a third edition? Let's go.

Eli Mansoor 36:45
So the Korean one is the third edition

we will leverage this to also publish the release in English.

Rob B 36:59
Fantastic. And, and so wrapping up guys, any final thoughts we'll go around or just anything, and anything you'd like to add or any material you'd like to recommend to people? And yeah, anything else to add, really, that we've not covered today?

Yair Green 37:14
There's a lot to cover, around cost optimization. But I'm very happy to be here with you guys. And

I believe that you know, it's a huge subject that we can discuss for hours. And I'm sure our listeners, time is expensive and important. So we'll leave it as it is for now.

Eli Mansoor 37:41
Eli, anything else to add? Yeah, again, first, thank you for inviting us and also for the listeners to listen to the good news about all of these topics is that there are good content out there organized way and also AWS partners, and to like the tools that AWS provides, and the partners to support our customers. And if they need help, they can do it with the support of partners. If not, they can do it by themselves. But it's a good point that we are, we are in a good position these days.

Rob B 38:20
Yeah, totally agree that the ecosystems really mature it hasn't. And there's so much great resource and help out there now. And, at Wirehive, one of those amazing resources that we have available as Mr Harding, who can now wrap up, Rob, anything else to add?

Rob H 38:37
Thank you, Ravi, thank you, everybody, for joining really, really enjoyed this session. I think kind of one of my final thoughts to say is for people that are looking to do cost optimization, I think the most success I found with my customers is when they get granular and look at each individual component specifically and kind of how can I maximize the most out of this resource, whether it's right-sizing that thing, whether it's you it's utilization rate, whether it's how much data is being used, you know, there's many facets to each component, you can look at the high level, the big number, which you know, it could be a bit shocking to start with, but in order to cut away that number as they go granular look at each component. And that's where you'll find the wins and save your cost.

Rob B 39:23
great insight. There's one for free. If you want another one, do get in touch. And joking aside, thank you so much for listening, or watching along live with us today. Thanks to our amazing guests. Thanks to Eli and Yair for joining us on their weekend, which is all Friday. And please do buy a copy of the book. It's great to read, it's very reasonably priced. And you'll get a lot of value from it if you're running any workloads in AWS or helping customers with their workloads as we do. And if you'd like to go a bit deeper than that and you'd like any support, we're always happy to help as I'm sure the guys would be if you reach out to them on LinkedIn.

So yeah, thanks again. That was Wirelive Episode 15. and see you next time for another exciting episode.