Ben de Mora - The Chaos Is To Be Embraced
Ben de Mora may be the newest staff member at the FinOps Foundation, but he is absolutely not new to the Foundation at all. Ben has been a FinOps Instructor for years and he shares his experiences teaching the FinOps Certified Practitioner course. The conversation goes in a number of different directions - architecture, sustainability, estimated cloud costs and banjo. As Ben reminds us, the chaos is to be embraced.
06:52 - FinOps Certified Practitioner Course
11:08 - FinOps & Architecture
18:52 - Implications of preemptible compute and sustainability
24:57 - Cost estimates while launching infrastructure
Be sure to follow FinOpsPod to catch the latest episodes.
Learn more about the FinOps Foundation at FinOps.org.
[00:00:13] Stacy: Hey, this is Stacy case[:
[00:00:18] Stacy: And this is FinOpsPod.[:
[00:00:28] Stacy: That's a great question. Cause I think we got a lot of feedback from the last one, and I think what I like most about this is we're touching on FinOps but it's not just that it's the people that have done FinOps that are doing FinOps and who they are as a person.
It's showing that nice balance of this is FinOps, but Hey, here's the journey of the person's doing FinOps and also some really cool, fun, personal things of those people as well.[:
[00:01:08] Stacy: Exactly.[:
[00:01:13] Stacy: Exactly. I was just gonna say that it's building the community and learning about the community. So I love this. I love this and I hope that we continue to have these amazing interviews.
Hey, Noel. I'm so excited to talk to you today, I'm even more excited to hear who you interviewed for this edition of FinOpsPod. Who was it?[:
[00:01:40] Stacy: I've actually worked with Ben. Pretty much since I started with the foundation because he's been an instructor and he was my go to to lead some of the FinOps certified practitioner training.[:
[00:02:01] Stacy: Yeah there's a lot, I don't know about Ben but I'm excited to listen to this, especially when he starts talking about how he got into the FinOps foundation helping out with the instruction of it, but also where did he come from? And, and how did he actually get into FinOps in general?[:
[00:02:31] Stacy: A little birdie told me There's kind of something special at the very end of the interview.[:
[00:02:58] Stacy: That's awesome. I can't wait to listen to it and hear the interview Ben, that you and Joe did.
Before we move on to that interview for everybody to listen to, I just wanna say one thing we do have FinOps X coming up. This is our first in person conference. This is gonna be in Austin, Texas, June 20th to the 21st registration will be open by the time this podcast goes live and you can go to X.FinOps.Org to register and learn more about it.
All right. So let's on the interview.
[00:03:52] Ben: Wow.[:
[00:04:00] Ben: 200 people, actually[:
[00:04:04] Joe: So first question if you train 200 people, was it an hour each[:
[00:04:28] Noel: I actually helped you on one of those, may or may not remember that.[:
[00:04:35] Noel: I didn't do well on it. I discovered that while I love talking, finops, I'm perhaps not the best trainer in the world. And I remember listening to your stories and when you were talking, like you were such a natural and someone who really enjoyed this space as well powerful stuff,[:
[00:04:57] Noel: I don't agree with you there because I looked at your profiles that looked you up online and stuff. And you, you coach under eights and under nines, rugby[:
[00:05:08] Noel: used to yeah.[:
Which he overcame through rugby. As he, as he was was growing up Coaching his, him and his team. It was just the most wonderful thing to do. It was great fun. And then my younger child, Joseph got old enough to start playing rugby. Joe asked me to coach his team instead.
So I switched and I was coaching the little like P four, they call it in Scotland, which is like age like seven, eight. And just coaching a bunch of kids at that age to play a game as structured and as, as disciplined as rugby is kind of like trying to herd cats, you have to, you have to just abandon all hope of any actual play and just embrace the chaos and get them to run in the right direction whilst holding the ball.
It's it's great fun.[:
[00:06:23] Ben: I'm an ex player as well. I was a forward. I was I was a number seven, a flanker. And yeah, the referee is sir, on a rugby pitch. That's the way that works.[:
[00:06:36] Ben: All right. were the quick[:
[00:06:42] Ben: Yeah, I was the idiot. That was basically my job on the pitch.[:
[00:06:51] Ben: Joe, I'll teach you soon, mate.[:
[00:06:57] Ben: Hmm.[:
[00:07:01] Ben: Ooh. eight year old rugby players.[:
[00:07:07] Ben: Less opinionated and the chaos has to be embraced no. I'm kidding. It varies every time. So I've, I've taught quite a few of these practitioner courses now. And I don't think any two have been the same. I don't think any two have had the same kinds of conversations. Things evolve all the time.
Things are evolving with FinOps as a discipline with the experiences people are having with the maturity that people are coming to cloud with and sort of getting to that point with concernes are changing as well. So the conversations are different in every single course I've taught so I've taught so far the kinds of people, the representation that you get coming to the course is different every single time.
So you might have a couple of CEOs and CFOs on one course. Right? And we've had, I think we had one course with like five CEOs and it was weird. The conversations were fascinating and we've given like another course. Like 12 of the people in the course are engineers. Right. And they have a deeply technical conversation about like what global automation and management and all this cool stuff.
So I learn loads every time I run that course. It's for me, I'm kind of keeping up a lot of the time. Right. I'm trying to figure out trying to get one, one page ahead in the manual before the folks that are on the call and trying to keep ahead of what they're talking about, which is fun.
[00:08:22] Ben: Yeah, I really don't know. I started the course. It was like people introduce themselves and I'm making a little note, like CEO one, CO two engineer, finance person, project manager. Cool. And kind of going down that list. Right. But this is, this is kind of what is, when you, you know, coaching rugby, everybody goes there to play rugby and they're all fairly good idea of what that's going to be.
I'm going to carry a ball and then knock people over. It's going to be great. We get covered in mud. Brilliant. Go to a finops course. And some people come to that with a goal of understanding how to do all the usage, optimization stuff and all the engineering, clever, clever, awesome stuff. Some people come to it with this one single goal of how do I make my bill less other people come to it, and like, I want to realize what the actual value of this is and how do I get there? Right. So there's loads of different priorities and loads of different sort of focuses from one single conversation. And, and that does lead to the similar level of chaos to a bunch of eight year olds on a rugby pitch with a ball sometimes.
But yeah, you kind of have to encourage the chaos a little bit and then direct it, shape it.[:
[00:09:35] Noel: So I remember it being, and I did a couple of them. I think I did. I probably did some of the first online. So back in 2020, when they first went online and it was always an interesting mix of people because it was a very new discipline then, and I never, I don't think we ever had any CEO type people on it.
We definitely had some, I guess, what I would call people who are responsible for portfolio of applications who didn't have an in-depth technical knowledge. they were just trying to understand what was going on underneath them. There was a couple of very finance driven people. I remember being in some of my, some of the ones I did at the very start very, very fine finance people, pure accountants, where they wanted to purchase reservations all up front. So they knew that they own them not convertible so they could get tax credits or write them off for tax and do all mad things. And that sense of like, just multiple opportunities to say if they were the kind of people that were there and some of the finops stuff didn't bother them because their whole logic on saving money was, Hey, you have a budget, don't go over it.[:
How do you know what changes are going to occur in terms of engineering, roadmaps, whatever. Or your customer demand levels or the cloud service provider itself. Maybe they're going to change that pricing strategy, this is it's much more fluid than, than the traditional it model.[:
[00:11:58] Ben: Both right. Bit of both. I mean, you, you always going to get the architect who wants to use every single complicated, shiny toy in the box. Right? And I came from a data center engineering world where I've encountered projects where there's been an architect. I had to put something like 360 separate pieces of cat six copper between two racks, because that was the level of interconnect between the devices in the racks that this architect had designed. And he'd done it with like 25 different types of solution of system in each rack. And it was insane. It was the most complicated thing in the world, but, and there was a far simpler way of doing what they were trying to do effectively. It was a VPN endpoint. Right? But it doesn't need that much, much complexity, but that had it because he wanted to play with all the toys that were there and available to them.
Right? And you still get that now because if fact, in fact worse now, because cloud is. You've got so many different products that you can choose from on the shelf now that you're kind of spoiled for choice and for an architect to go, wow, I'm going to play with all these things because I've been given a remit to build a big thing.
I'm going to build a big thing with all the bells and whistles which can be useful, but it can distract from the actual point, which is what are you doing? What are you here for? What are you doing with this system? Right? You're trying to produce a service for a customer that is performance, resilient scalable, and cost-effective. And are you able to prove that it's cost effective? Do you really need all of these things? Right? So yeah, you get the architects that want to play with all the shiny things, but your also get the architects who understand that this ultimately is a business that needs to operate. This is a, this is a business delivery piece.
And as a result, we have to basically justify the benefit of every dollar we spend on this infrastructure, because we don't own any of it. It's all, it's all consumption-based stuff. So we can assuming this thing. Does it, does it give us anything to consume that thing or is it just a, just in case or a nice to have, is it a cherry on the top that we don't really need? Nobody will notice if it's not there. It's that different mindset we see both at the moment.[:
[00:14:25] Noel: I have really got to know how can he put that in as an architecture? My head says no problem. I even talked to people who are going take into account or EDP and savings plans when you're putting how much it's going to cost.[:
[00:14:42] Noel: But how would you rate that as part of your architecture?[:
[00:14:51] Joe: Right. Well, and so this is where our team, my team started becoming more architecture based is because we would always say, well, what is the financial impact of this solution you've delivered? And they'd be like, here's what it is. Here's what it is based on our common rate that we, that we get for our company.
And they'd be like, well, no, no, no, this is overstated. And I'm like, well, it's all an estimate at this point, because nothing been built. Like, no, no, you're not doing the three-year upfront. And I'm like, no, no, discounting methods shouldn't be part of what drives the architecture, I remember one, one solution was just seemed wildly, wildly expensive.
This was being built in AWS. So it was all ec2 based. I'm like, well, this, you could just replace this entire corner of the architecture with the Athena service. And then everyone looks at the finops person. Okay. Oh, and, and, and architects giving you a dirty look and you just start drawing and that's how the finops starts playing into the architectural world.
I know we're seeing more. . . I review the membership applications that the FinOps foundation, so I'm really happy to see more architects .People with the title architects coming through. So I hope they start headed heading towards your training. .Ben.[:
[00:16:04] Noel: Can I go back to something? You said a few minutes ago, Ben, you, you were an engineer. I think you said in an, a massive data center.
And , the question I want to ask you is today are you now poacher turned gamekeeper as in FinOps practitioner or FinOps trainer or not.[:
[00:16:26] Noel: Going more so from building, creating, making stuff. Right? As an engineer, right? And now you're the FinOps guy who is, is going back to these engineers, going actually try and do it differently don't do it the way you've always done it, you know?[:
The consolidation of all the stuff into these virtual private clouds and that kind of thing. It's kind of been a thread through my whole career, has been, how do we maximize the benefit? How do we maximize usage of this resource that we've got in the data center? You can't really easily resize a server without a dremmel and then you get told off for doing it. But you know, If you, if you've got a, an amount of compute, you want to make sure that very little of that compute is just sat there doing nothing, because ultimately it's occupying space and power in the data center and it's an investment.
I don't feel like I've gone the other direction. There's just a continuation and evolution of what I've always been doing. And actually the fact that I'm now able to do this with incredible amounts of data analytics, with cloud. That I could never do before and we can do it in a way that allows us to be much more efficient than we will ever be before.
Really, I think is kind of an awesome progression. So it's, to me it's a natural, natural flow.[:
[00:18:48] Ben: Okay.[:
[00:18:51] Ben: Yeah.[:
I have my opinion interested in yours.[:
There isn't that direct link. The moment that doesn't seem to be a direct connection between I've reduced my costs by half to I know that I have reduced my energy consumption by half or whatever it's going to be because at the moment that can, that's kind of a black box. We don't really know what the vendors are doing in terms of idle resource.
My gut tells me that that's probably not happening that much because of course you have to have capacity standing by in case you'll have a burst in demand, but I'd like to think there's a way around being able to sort of sensibly scale down, have a buffer zone and scale down power usage.[:
If can they reduce the capacity, will that impact the spot market?[:
Right. But then that implies that there are in fact systems sat there doing nothing, which is a carbon footprint problem. Right? So , does the attractiveness of spot mean that we are effectively encouraging waste by having those resources spending to provide spot. Is that a good thing? I mean, I get, I get why the vendors are doing it because they're incentivizing usage of their systems that would otherwise be earning them absolutely nothing. But what would be nice is instead of doing that, they would say, we've got this compute, this. Doing nothing and consuming energy. Let's bring it down and reduce the footprint. That would be a nice message.
I know that spot isn't hugely used by many people. There are some people who are using quite a bit of spot and they've got a very good architecture and sort of automation around doing that. But most, most people that I've seen most footprints that I've seen are like two, 3% spot. Right. And they're doing it because they're doing the dipping a toe.[:
And most, most engineers is they give me a two-minute warning for anything and I can react to it. And it's, it is, I think it's encouraging people to get used more spot because it is. A massive, massive savings .[:
[00:23:18] Joe: Let me plug a working group here. That's spinning up in the finops foundation. Cause as of, as of this recording date, which is March 3rd, when we're recording this, we are spinning up our sustainability[:
[00:23:30] Joe: It started off as a working group, but we're actually going to turn it into a special interest group, meaning that it's going to continue living and continue working on this.[:
[00:23:40] Noel: I'm going to, I'm going to ask you another question before he tries to wrap us up you said you did your finops certification. Finops Practitioner certification very early on.[:
[00:23:50] Noel: When did you certify.[:
[00:23:58] Noel: The first online course.[:
That's the thing I've never actually taught a FinOps class in person. I've never taught a FinOps first class, like to see the entire human being.[:
[00:24:17] Ben: it's going to be fun and I'm really looking forward to, I hope that we get to do a bunch of in-person training this year.[:
[00:24:25] Ben: Definitely the finops the finops Scotland community, the meetup group for finops Scotland is is going to start waking up. It's been a bit quiet over the last little while because mainly cause COVID and all that kind of thing. We've got a whole bunch of really awesome people up here. But I'm looking forward to getting together with, and eating too much pizza and having chats.[:
[00:24:54] Ben: Definitely.[:
I'm going to ask you my last question, Ben. This is a question or a segment, I want to start calling, "The answer can't be savings plans for RDS", right?[:
[00:25:10] Noel: Okay. If you could make one FinOps related change in the cloud service providers, anyone you like are all of them,[:
[00:25:19] Noel: what would it be?[:
Here we go. Right. Go hit, apply. And I sold them lots and lots of things scroll down and Pope the page for 10 minutes and everything's built. I'd like it to just come back with something that says what you've just built is going to cost this much per hour.[:
[00:26:08] Ben: now there is a third party tool right now that starting to do that, can't name them, but can we name them?[:
[00:26:17] Ben: an open source one. Okay.[:
[00:26:21] Ben: Infracost. That's the one. Yeah, very cool. I have no idea how accurate that is. I've not used it, tested it at all, but I love the idea. I think that is brilliant. And I think that the impact on having that feedback just as part of the entire cloud build experience is, uh, would be really powerful.[:
[00:26:46] Ben: Oh yeah![:
So, I mean, it's a, it's a great first start. It's it's, it's pretty cool presentation. So you can go, anyone listen to it and intrigued by what Ben's talking about. Cause it is very intriguing. Go check it out on our YouTube site, the New York city meetup.[:
[00:27:23] Joe: watch me accidentally put it up above, but it's going to be somewhere around the thing in the show notes.[:
[00:27:30] Joe: Hey, you mentioned the banjo that's on your wall. And you told me a really great story. It is your father's banjo.[:
[00:27:42] Joe: How many banjos do you have?[:
I was in a shop to buy some strings. And I saw this other one that was just ridiculously beautiful. Walnut wood on it and it was really lovely. So I was like, okay, I need that. That's very low. Very lovely. I'll buy that.[:
[00:28:02] Ben: My dad's more than more than anything else. It sounds nicer.[:
[00:28:07] Ben: What[:
[00:28:11] Noel: We need intro and outro music yeah.[:
[00:28:14] Ben: Let me, let me record something so that I can have 35 different attempts to playing something that doesn't have any mistakes in it for you.[:
[00:28:21] Ben: I'm sure you will. I'll tell you what I will. I will record you as something I'll record a little bit of Cripple Creek for you. And it's gonna be Cripple Creek and[:
[00:28:30] Ben: you can sing Cripple Creek.
There you go.[:
[00:28:33] Ben: I've not read your blog, but I feel like I know you too well already.[:
Do you play any other instruments,[:
[00:28:46] Joe: Which are you better at guitar or banjo?[:
[00:28:55] Ben: I, my, I inherited it from my dad when my dad passed away last November. And he was a professional banjo player for most of his rather bizarre roguish life. So I mean, if you want to hear, so here's another plug for, if you want to hear my dad play banjo listen to the song, New Shoes by Eric Bibb and the opening banjo and my dad co-wrote that song.
And the opening banjo is my dad playing. So there you go.[:
[00:29:35] Ben: Yeah, I will I'll record something in a bit, cause I've only been playing a couple of months, so it'll be thousands of mistakes. So I'll get something don't screw up.[:
You your piano was really good. And it was like piano. She's playing trombone. It's a music joke. It's a volume of, I[:
[00:30:15] Joe: a good thing. I'll[:
[00:30:18] Noel: That's should definitely stay in the silence should stay.[:
[00:30:25] Joe: All right, Ben, I feel, I feel like after this conversation, I've, , I've been in a rugby match.[:
[00:30:33] Joe: We've been, we've been all over the place and it's just so natural and fun. But, that that's FinOps right. Encompasses so much.[:
You need your backs, you need your wingers, you need your your scrum players. You need all these different positions on the team, And you need all these different pers positions in the finops capability in order for it to be, to be a really rounded and successful effort. So yeah, the chaos does need to be embraced in order for that to be successful.[:
[00:31:42] Ben: The banjo always banjo, Yes. Don't forget the banjo domain is a very special part of finops, but really is only reached at the very final stages of maturity development. That's right.[:
That was Ben Demora plucking the old banjo playing a song called "Worried Man" for us. Thank you so much, Ben. And thank you. The completionist listener, listening to the entire podcast. You're not stopping halfway. You're listening to the entire thing You are a true FinOptinaut. And you get to hear me talking over our theme song, cheese puffs and caviar. Pretty soon that jazz flutes gonna pipe right in, I'm going to talk right over it. And that is a great way to end the show.
What an interesting conversation that was. Ben gave some really great insight into what the FinOpsPod certified practitioner course is like. the conversation went all over the place. Architecture, sustainability. Banjo. Maybe I'll give banjo a try. I doubt it.
Let me plug FinOpsX, our in-person conference coming up June 20th through 21st in Austin, Texas. Go check it out. X.FinOps.Org to learn more about registering and let us know if you're interested in doing a presentation at the conference, sharing some of your FinOps skills and your FinOps knowledge with your peers.
I love to have you there. I'll attend. I will attend your session. You can count on it. Unless it's in conflict with somebody else. Who's also listening to this . We'll work it out. We'll figure things out. So thank you for listening. Thank you, Ben De Mora for being a great guest. Thanks to Noel Crowley, my co interviewer and Stacy case for kicking us off. Thank you.
Can't wait to do this all over again next time, on FinOpsPod.