Max Clark: [00:00] Just cut it out. Let me fix my, I wanna fix this headset thing here. I need to get like another clip, like right behind my neck. Figure out how to do that at some point. It’s always like,
William Sellers: to wear.
Max Clark: I used to wear the big, the big muffs, but
William Sellers: some
Max Clark: I just,
William Sellers: these are nice. They’re, they’re, they’re, they sound really good. They’re closed back, but they do get hot after a while, so I need to
Max Clark: Hmm.
William Sellers: some in air monitors and kind of do it that way. um,
Max Clark: You know, there’s, um,
William Sellers: uh,
Max Clark: uh, if you look at the aviation headsets, so like David Clark, they sell a, um, like a fabric cup. Like, like it’s, it’s flexible that goes into the headset and it’s specifically for that to kind of help you from like sweating into your headset when you’re in the plane. So if you’re in like a really big, you know, humid location, all these different things, a
William Sellers: Yeah, I’m gonna
Max Clark: little bit.
William Sellers: at those. Uh, I’m kind of interested in the [01:00] in-ear monitors mainly ’cause I play bass as a hobby, so need to get
Max Clark: Hmm.
William Sellers: some in-ears. Most people are going to like silent stage these days where they don’t have like big wedges and things. So some in-ear
Max Clark: Yeah.
William Sellers: the most comfortable.
Um, you know, and they, they tend to work pretty good.
Max Clark: These are just, you know, I mean, they’re nice. They’re nothing special. I got to a point where I was like, oh, you know, I should, I should look for, you know, and, and you know, you find these things that are like, you know, really nice solutions out there on the market that are usually musician, musician related.
And I’ve thought about doing it and having custom ear molds. Um,
William Sellers: noticed that,
Max Clark: I actually noticed that, uh, it started from like, from like, uh, shooting sports. Um, but then I noticed it with most earplugs, like my left ear canal. Just, it’s large, like, I don’t know how to say it except my left ear canal is just big. [02:00] And um, so I get, you know, you get those like ear, mu ear, you know, like those, the foamies where you have to like swirl ’em and then put ’em in like the normal one of those will not seat and seal on my left side of my ear.
Right. So I have to turn them around and wear them backwards. ’cause then I actually get a seal. Um, and there’s not these like quirks you learn about yourself as time passes. Like,
William Sellers: is as you get older, things start to change. That used to not, I used to, I, you said shooting. I
Max Clark: ugh.
William Sellers: to, I, I used to shoot competition and stuff long time ago, and I started picking it up again and I used the little ones. You roll like this, you put ’em in your ear and they foam up and they close, they close the channel. Um, those don’t work on me anymore for some reason. It’s not that I got a big ear, ear canal, it’s more just that
Max Clark: Oh.
William Sellers: just doesn’t, just doesn’t fit right. So, I don’t know. The challenges
Max Clark: Yeah, they, uh,
William Sellers: Um,
Max Clark: I had, I’ve had a couple of different, um, uh, molds done, you know, where you get like the custom foam molds in your ears. I have one [03:00] set that work phenomenal, but the woman, when she did them, they are deep. And so you get that like in the back of your throat that like taste when you’re like, take them out of your ears.
Like you can, like, you get that like, you know, like twinge, but they are deep and they work. And I’ve had, I’ve had, um, other ones made sense then, and I tried to get some of the, uh, like active noise canceling, you know, with the thing mold. I mean, it’s crazy expensive. And I went, I, they had, they remolded like three times and all molts, all the molts failed.
Like I, you get the box in the mail and you put ’em in.
William Sellers: I was just
Max Clark: And after the third time, I was just like, I, I, I’m, I’m just, whatever. This is just a loss and I just tossed, you know, I just gave up,
William Sellers: you do, you
Max Clark: but mm-hmm.
William Sellers: because I’ve had so many friends talk to me and they’re like, William, I have this constant ring in an ear or two
Max Clark: Oh,
William Sellers: loss and it, it’s, you know, it’s not the worst thing that can ever happen in life.
But if you, especially if you’re into music and things like this, you want to be able to hear well. ’cause it’s important [04:00]
Max Clark: Texas, the, the founder of Texas Roadhouse ended up, um, committing suicide. And it was, he had, um, I guess suffered from tinnitus for a long time. And just so the, the, the report that I read was just, you know, like there was cor, I don’t know if he left a note or what was cor and people just knew it was just correlated to that.
And that, like, that thought of just having that like constant, like ring in your ear, um, I mean, it would, it would make you crazy after a while.
William Sellers: Bat Cool.
Max Clark: Are you in Baton Rouge?
William Sellers: the Baton Rouge, uh, at the Baton Rouge Data Center. Absolutely. Yep.
Max Clark: Awesome.
William Sellers: My
Max Clark: I have, uh, my, my, um, family lineage on one side came through New Orleans, so
William Sellers: is an interesting
Max Clark: we, uh
William Sellers: I was there, uh I was there this weekend with my wife. We were, we we’re attending, uh, this year will be our 30th wedding anniversary. one of the things we’re
Max Clark: Oh, congrats.
William Sellers: going to see, uh, [05:00] Paul McCartney, uh, in New Orleans.
He’s coming in October. um, you know, we did the whole, we’re doing the whole like pre sound check, the meet and greet, all that kind of stuff. And, um, we went
Max Clark: Oh, so fun.
William Sellers: to kind of scope out the hotel we’re staying and just kind of see what’s around there. And, uh, it’s a, it’s an awesome city, you know, it’s, it’s definitely good food, good music. Most people are really friendly and, and amenable, so.
Max Clark: I, I think your original comment of New Orleans is an interesting city, is probably the correct way of saying it. You know, like. It’s, uh, my, um, we did a cousin’s reunion trip thing and we ended up in New Orleans and I have, I don’t know how or why we ended up there, but it was, it was pretty funny. ’cause you know, people could, you kind of come and go as you please, you know, and then there would be like one crew that’d be out.
So like, I can remember coming back and crashing out in a hotel and waking up like midnight, you know, 11 or midnight after this nap. And they’re like texting people, where you at? Oh, we’re at this bar down the street. And you just kinda like stumbled [06:00] down.
William Sellers: it
Max Clark: The funniest part about it to me though, and my, my youngest brother like, knew all about this subculture, but it was Pirate Con was going on in New Orleans.
And, and so, you know, it’s just like what you think, like, people that like dress up and like to role play cosplay as pirates, right? So think, like, think like dodgeball or the guys like, thinks he’s a pirate. Like, like that actually exists and there’s a lot of them. And so my brother knew about this ’cause he had friends that were in pirate bands that were playing.
In New Orleans that weekend, like full on, like,
William Sellers: cool.
Max Clark: you’re like, this is, you’re like that, that’s, you’re like, okay, okay. I’m, I’m into it. Okay.
William Sellers: you know,
Max Clark: Anyways.
William Sellers: it’s a fun place to be. You know, bourbon Street can be a little bit rough a little bit, um, you know, to the uninitiated can be kind of disturbing at times, but it, it’s all, I think it’s all done in good fun and, um, you know, it, it’s a, it’s a, it’s a definitely a good place to visit and a lot of people enjoy living there. [07:00]
Max Clark: Yeah, it’s, uh, it was a, John Goodman I think was just selling his house in New Orleans. He had a really nice place, and I didn’t realize this, but he was, that was the usual, like, I gotta get outta la raise my kids in a SANE environment, you know, kind of deal.
William Sellers: who actually live in New Orleans. Um, John Goodman I think was, is is there or was there, Lenny Kravitz, uh, had a home here. I think he still has a home here. Uh, Nick Cage, Nicholas Cage. So there’s a lot of
Max Clark: Mm-hmm.
William Sellers: still love the New Orleans vibe. And, and, uh, go down into, you know, you get away from the quarter you go into like, you know, further down Canal Street or Magazine Street and it’s just, it’s amazing, you know, down by the city park and stuff.
And you see all these beautiful homes and really good place to be.
Max Clark: Um, I’m gonna, let me, lemme come up with a, with a, actually, I, we’ll, we’ll do it live. I feel like the O’Reilly sometimes when I’m doing these things, I’m [08:00] trying to come up with like a really good, like, like intelligent, like intro, but we’ll just kind of, we’ll just kinda go into it. Um, by the way, if anything, um, don’t worry.
We’ll edit afterwards. It’ll come out, you know, if there’s, it’s not, it’s not, it’s not like ambush journalism here. So, um, I think I should tell you is if you look right in the middle of your screen, so it’s probably actually towards
William Sellers: um,
Max Clark: here, I guess. Um, you see a record, there’s a,
William Sellers: So this
Max Clark: there blue indicator. So this thing does
William Sellers: recording. Oh,
Max Clark: local recording and then a progressive upload.
So we’re getting like a low res back and forth with us. So your only requirement, other than talking a lot is, um, at the end, I’m gonna stop the recording and we have to wait until it says a hundred percent. It’ll probably be, it’ll be pretty close most of the time. Um, but that’s, you know, so if it, if you lose me lose like, like video gets weird or whatever, don’t worry about it.
It’s all, all the recording both sides is, is local. On the, on the
William Sellers: jack. I’m on actual hard wire because
Max Clark: wonderful
William Sellers: is funky, so. [09:00]
Max Clark: that, that is the, yes. Wifi gets really funky. Uh, um, so I’m gonna try to do this without starting with a, so, ’cause that’s my little,
William Sellers: I had
Max Clark: I I,
William Sellers: a
Max Clark: you know, I had Toastmasters in second grade and, and the ums and pauses they beat out of me. And I didn’t realize it until I started doing this. And I had a, an editor tell me that, I said so a lot.
William Sellers: though
Max Clark: I had to coax it out of her though. William, I, you sent me a few
William Sellers: of
Max Clark: things to.
William Sellers: about and really
Max Clark: Talk about and really get into, and there’s a common theme across all of them. And the, and this energy,
William Sellers: customer of yours that we’re
Max Clark: energy customer of yours that we’re gonna focus and spend time on
William Sellers: had in the
Max Clark: had in the same thing.
William Sellers: true,
Max Clark: It seems to be true and true. Right.
And reading this, of course, I project a little bit, ’cause I see this a lot and I talk about this a lot, but
William Sellers: promise the cloud
Max Clark: the promise of the cloud, like, oh, the cloud is here, the cloud is great, move everything to the cloud.
William Sellers: of enterprises
Max Clark: a lot of enterprises following that and pushing [10:00] into these digital transformation projects and cloud transformation projects, and then getting to a point where it’s maybe not the right path for them.
And then, and then, and I’ve seen this a lot of different ways and I, I really wanna talk, let’s, I wanna start with asking you and talking about just that set of background, what drove them to the cloud and which cloud, what was the promise and, and what was that inflection point for them really of like maybe this.
William Sellers: Yeah.
Max Clark: Isn’t what we thought it was.
William Sellers: this was a customer, um, in the energy sector, right? And they were a current customer of ours, um, consuming, you know, co-location at the data center. They had, uh, they had some workloads in our multi-tenant club and, um, managed firewall, right? And things, life is good, right? They’re, they’re consuming our resources and they’re, they’ve been a customer for several years, I mean like five or six years. And, know, they had an initiative to build an application. And to get it done [11:00] quickly and like, I’m not saying shadow it, but they go to, they’re like, Hey, take our credit card. Let’s swipe some measure and let’s start building the application let’s build
Max Clark: Mm-hmm.
William Sellers: you know, see if, see if it, it’s gonna work for us.
And they built an application and deploy it into Azure all good. They actually had some, uh, express route to go between our data center and Azure. And what they found is they were able to get the application up and tested and do many iterations until it became the perfect app for them. Um, but they started becoming successful with it and actually uploading more and more and more and more and more data. And The, price kept going up and up and up and up and, um, you know, they came to us. You know, they, they came to us and they were like, Hey, look, you know, this is getting expensive more so than what we’re paying, you know, for, for to host our workloads in your cloud. [12:00] And, you know, they came to us and they’re like, what we need is we need to migrate Azure to, to our Baton Rouge multi-tenant cloud here at Dark Points. And um, you know, some of the things that they really were looking as like, Hey, we want to, you know, reduce the cost of the solution, right? We need to, we need to save money. their spend with Azure was up into like the 30 KA month cost like, we need to reduce this. And a lot of it’s dis storage and a lot of it’s, you know, bandwidth and throughput and things like this.
And, and also when the application spins up, resources become more, more taxed and that equates into bigger bills, right? So that was really the first thing is lower the cost. Um. They also wanted to consolidate their cloud spend. They’re like, look, you know, we’re paying Azure. paying money for express route. We’re also paying you guys for co-location. And all of the services that we get, they’re in the Louisiana area, [13:00] so they’re relatively close to our data center. So they’re like, we wanna really wanna bring this application back home. Right? And then, you know, they wanted to say, okay, this is where we want to go.
It’s gonna reduce complexity, it’s gonna reduce cost, right? So we can migrate, but we need to migrate this, this solution out of Azure with very small amounts of downtime, We, we, we can’t be down ’cause it’s now a critical application, right? You develop it, you know, let’s just see if it works. Okay, it works, it becomes relatively successful and now it’s a dependent application. And they were like, you know, we gotta migrate this data. of, out of Azure to save money and to be, be less complex. Right? We can maybe shave some things off. So, you know, they came to us and said, listen, um, what do we do? How do we get there? And, um, you know, of course the thi the thi first things we do when we are working with [14:00] customers like this we do the, the obligatory workshop, right?
Let’s see what the application does. Let’s try to understand what, what are your challenges? And, um, you know, we, we do the, let’s see how much it’s gonna cost coming to us. Let’s see if it’s feasible. Uh, let’s make sure you’re ready to move from Azure to, you know, an on-prem or, or a data center like dark points.
Let’s make sure it’s the right, it’s the right fit. we go in and we like, you know, look and see, are you consuming any Azure resources or, um. Or solutions we can’t provide. You know, if it’s something like in, uh, platform as a service or something that’s not a great fit, we wanna identify those before talking about billing and migrating and, you know, actually the steps it takes. So, um, you know, our game plan was to do a, we sat down with ’em, we did the workshop, we did the whiteboards, we understood the application, [15:00] and the customer had kind of was already, you know, and our, our goal is to convince them or, you know, convince them it’s the right solution or convince them it’s the wrong solution.
Stay in Azure, maybe that’s been in Azure is justified. And your best, you know, your best, um, you know, to success is staying where you’re at, you know, so we take a very consultative approach when we work with customers, right.
Max Clark: The, I’m, I’m scribbling notes here fast. Um, what’s interesting about this is.
William Sellers: Depending on the company
Max Clark: Depending on the company organization, right? You start talking about org org hierarchy and what does the actual org look like? And, and then you say, okay,
William Sellers: Application. And
Max Clark: the application? And how does that application then get budgeted and, and reflected internally?
Is this an externally facing revenue generating application? Is this an internal
William Sellers: Mm-hmm.
Max Clark: productivity tool that’s used by the [16:00] organization?
William Sellers: that
Max Clark: And then,
William Sellers: a lot
Max Clark: and then that dictates a lot of, you know,
William Sellers: Like who actually
Max Clark: as well, like who actually owns these things. And you have organizations that still have,
William Sellers: IT infrastructure
Max Clark: traditional IT infrastructure roles and maybe that reports up into the CFO or the finance administration side of the house.
Or you have a software development team that goes to A CTO and like, right? So you get in all these different like org hierarchies,
William Sellers: around
Max Clark: conversation around
William Sellers: costing us too
Max Clark: this is costing us too much money.
William Sellers: this is gonna
Max Clark: We’ve realized this is gonna continue to cost us too much money and we’re gonna continue to scale and invest in this.
And maybe this is something we don’t want to, that sounds like very much like a finance conversation. Like somebody,
William Sellers: pressure
Max Clark: know, that’s, that’s pressure from
William Sellers: at
Max Clark: a CFO looking at numbers and the numbers are only going one direction. And then, and then, and then really, you know, ideally, I mean, look, you know, in my world, like cloud spend going through the roof is okay, as long as the revenue spend is going up, you know, with it, right?
Or revenue, the revenue coming in goes up with it, right? But when you, when you start looking at that and you say, okay, cloud’s going up, but revenue’s not, [17:00] or that they’re not directly correlated with each other, um, that should get people’s attention.
William Sellers: also
Max Clark: But then also that drives, and, and you touched on this a little bit,
William Sellers: process.
Max Clark: migration process is really complex.
You know, like this isn’t from, for a lot of people, this isn’t just. You know, even if, let’s just say in the easiest way, say, uh, you, you’re, you’ve completely containerized, you’re not serverless, you’re not completely integrated with, with, um, any of the cloud cloudy tools. And you have an application you can forklift out really easy.
It still requires a lot of human capital to move your application back. And, and in this case, um,
William Sellers: like
Max Clark: you know, it’s like, what’s, what’s the expression? The juice has to be worth the squeeze. Right? Like, and I, and I find a lot of times these projects stop because the perceived risk or the perceived level of effort to do that migration just becomes this point of paralysis.
Um, but in this case, they migrated,
William Sellers: That,
Max Clark: they pulled out so that, you know, somebody inside the organization, there’s like strong executive, like, we’re gonna make a [18:00] change here, kind of push. Mm-hmm.
William Sellers: for us. ’cause I mean, when we are, when we’re working with customers, we have to be, you know, IT administrator, solutions architect, finance person. We have to talk about all the challenges that a customer has to face when migrating to a data center or away from a, a hyperscaler, right? So it’s not just, you know, can we do it? How do we do it? It’s like, why do we do it? Is it really gonna save you money? Is it gonna save you headache? And when we were, we actually had the, the IT director, the system administrator, and I think we had the finance person in our data center, right? And we were whiteboarding and we were looking at like, know, let’s make sure it’s the right decision.
And, know, at the end of this conversation, the, the finance person said, listen, you had me at migrate out of Azure. Thank you for
Max Clark: Oh no.
William Sellers: shopping with us. It’s really great. It, but basically we had done the homework already because they were previous cu They,
Max Clark: Yeah.
William Sellers: were already a customer.
So [19:00] they’re like, you know, William, I appreciate all of the diligence and the, the, the, know, the, the architecting and the consulting, but we’re ready to get outta Azure, um,
Max Clark: This is a pretty big divide for a customer profile now within cloud environments, right? Because you have, you have the, the original promise of, um, elastic resources or ephemerality and
William Sellers: down
Max Clark: to scale up and scale down and unpredictable workloads. And then you’ve got
William Sellers: that
Max Clark: most enterprises that are relatively static or steady workflow.
I mean, that’s, there’s a high degree of predictability of an application once the application has production, right?
William Sellers: that’s, you know, the thing is with that is, that’s true in some cases and in other cases it, it becomes, you know, the old saying, pay as you go broke, um, you know, pay as you go is great. And, and, and having that al elasticity is fantastic. The problem is it’s tied to your credit card. And, and, and that, that [20:00] becomes a source of, of pain, right? Um, especially if you become successful or the application takes off or you start using it more. this, in this particular solution, it was an internal application that supported some of their energy. Um. Uh, research and, you know, it it, like I said, they had already done, they had already made the decision for the most part before they came to us.
They’re like, look, we, this is what we’re doing, William, but let’s, let’s walk through the process. And for them, the, the way how we leverage cloud, really, um, we give you the customer the flexibility of a pay as you go architecture, right? Where they can, uh, allocate additional resources when they need. the, the one thing that’s really good about the solutions that we have, especially when it comes to cloud, whether it’s multi-tenant cloud or single tenant, uh, cloud is the customer is very simple.
Billing, right? You pay for disc, you pay for CPU, you pay for [21:00] RAM network and bandwidth and, and it, and, and that’s what you get. And, uh, no matter how much you use, the, the, you know, the, the, the, the actual compute, it’s not gonna spike to where. You, you, you get outta whack. Plus again, we’re very flexible and nimble on what we can do for customers in terms of creative billing, creative invoice invoicing, and some of the things we do to kind of alleviate the, sting of getting a bill at the end of the month, that’s $30,000.
It’s, or it’s, it’s, it’s more than what you were expecting. Um, so really, you know, after we went through the workshop and made sure that the customer was really cool with getting out of Azure, we started looking at what’s the easiest way? What’s the way to go? And, um, it, it was, it was funny. So we’re a VMware based right now.
We’re a VMware based cloud, whether if it’s, um, [22:00] single tenant running vCloud Director in a multi-tenant, uh, environment. Um, and at the time, my, my role as solutions architect, I also did a lot of research and development, or I did a lot of like product testing and, um, you know, going with the customer’s desire to have minimal downtime, you know, um, this was important to them.
So I was looking at, one of the products we use for our DRA solution is a product called Zerto, right? And it allows customers to do realtime replication of their virtual machines from. And Mo and, and how we use it. We typically see customers doing VMware to VMware replication, whether that’s VMware to vSphere, you know, VMware vSphere to vSphere or VMware vSphere to, uh, vCloud Director. We also have some customers that do Hyper V to, you know, across hypervisors, into [23:00] VMware. Um, but one of the things I was testing because, you know, I’m, I’m very interested in what people and how people leverage public cloud, meaning the hyperscalers, looking at ease of, you know, ease of moving workloads around. You know, when I first started running virtualization back in the day, it was crazy awesome technology. the issue was how do you move the VMs around? How do you make them mobile? And one of the tools that we really invested in and really, um, use a lot even today is, is the Zerto product. Right. And um, so I was in the lab working probably a few months before this customer came to us.
And I was testing Azure to vCloud Director replication where, you know, you have some workloads running in Azure. You go to the Azure marketplace to say, oh, there’s my virtual, there’s my Zerto cloud appliance, let me deploy that. and I had done several [24:00] testing, uh, you know, did some test, several testing, um, you know, sessions, right?
Even did lunch and learns and like show and tell to the rest of my engineering team. I’m like, look guys, if we ever have a customer needs to get in or out of Azure, this is the tool that we can use. And, uh, you know, we did, I, I did my testing with our team, did, uh, documentation how tos so that our, our offering to the customer was kind of already built a couple months before we actually was talking to them. Um. So we had really good success in lab and uh, you know, when we had a follow up session with the customer, we said, Hey, we want to use this product. This is gonna be able to replicate the VMs from Azure’s prim into, into ours. Piece of cake. Right? Uh, it’s never, always a piece of cake, but in this particular case, um, it was, and it wasn’t, a lot [25:00] of, it wasn’t a lot of VMs.
We had maybe, think we had maybe five virtual machines that we were, we were migrating, right? four of the machines, uh, were relatively small. Couple of hundred gigabytes of storage, some sql, they were all windows. And, um, those tested really well. One of the things we like about the Zerto product is we can. Before we actually fail over, we can do a failover test, right? Non-disruptive, let’s see how these machines fail over in a isolated environment. And all of our testing for four of the five machines, beautiful. No problem. We were able to fail over from Azure. Uh, we could fail back if we needed to. So we, if, if we had an issue where something just wasn’t right, we could fail back or, or, or back out of the, of the, um, the failover. Uh, but one of the machines was quite large and when I say [26:00] large, it was about a hundred terabytes, 106 terabytes worth of data. And, um, you know, so. At the time we didn’t, wasn’t really a problem. We, we set up the replication target, we started replicating and we were like, we told the customer ahead of time, it’s like, listen, this particular VM probably is going to take a little while to replicate even replicating over your express route connection.
Right.
Max Clark: Mm-hmm.
William Sellers: um, know, so three of you know, the four of the machines, no problem. We got to the fifth machine, the really super huge, super sized me virtual machine. And we had issues. Uh, it would get to, you know, during the replication it would get to like 95%, 99%, and it would just fail. And, um, of course, you know, that’s, that’s not good because this machine was like, really it was the crown jewels.
It was actually, it wasn’t a very complex machine. It was, it was a file server. Uh. [27:00] um, you know, it was failing on us. So of course, you know, we, we stopped, we halt processes. We, we, you know, halt what we do. We call our vendors, we call our support. We call, you know, VMware and Zerto and collaborate on how to fix the problem. And it comes, it came out, you know, when we were working with Zerto, um, we discovered that there was a bug in the Zerto application, right in the actual build that, uh, was the latest build at the time. And it had a problem with discs that were over terabytes. Okay. Well, the, the machine that we were working on basically had, you know, four or five, to 30 terabyte discs.
So really, really
Max Clark: to it and.
William Sellers: And, um, you know, we were like, okay. We, we have this problem, how do we fix it? You know, [28:00] Zerto comes back to us a couple days later and they’re like, listen, um, you know, we’re, we’re gonna have a fix for this. And, uh, it may take a little bit of time, it may come out on our next big build, but, yeah, hang tight, right? So we go back to our customer and we’re like, Mr. Customer, Mrs. Customer, this is the problem we’re having. It’s not a problem that we can easily solve. And they, you know, the customer was very understanding, but they said, listen, is there anything else we can do? Um, you know, we need to get this machine over to, over to dart points. um, you know, I was like, well, it could be a month before. Um, the, the Zerto patches out. And then of course, once it comes out, we have to test it. We have to go through but this is the situation when we’re in, and then we’re, like, look, we don’t wanna spend another month of, of Azure services. We really are ready to get out. So, you know, we kind of took a [29:00] step back and said, okay, what’s the, maybe not the slickest way to migrate this machine, but what is a way that will get it done? Right? Uh, the good thing is it was a file server, windows file server. Not a lot of applications, not a lot of complex, you know, we’re not running Oracle or SQL Server on this machine.
It was basically just a file server. So we, what we ended up doing was excluding the discs that were the super huge machine, the super huge volumes, right? So we only replicated the, the boot, the boot drive, like the C drive and we. You know, set up the replication again. We were able to fail over the virtual machine with just the boot drive so it would come online and would function, right? it just didn’t have all of its data. what we ended up doing was failing it over our multi-tenant cloud. So from Azure to to [30:00] Baton Rouge. We brought the machine up, changed the machine name so we could have both machines up and. running at the same time. They would see each other, but you know, nothing like net bios, conflicts, IP addressing nothing to
Max Clark: Yeah.
William Sellers: us.
And then what we did is we said, let’s just build your virtual disks back right in our cloud. So we, we recreated blank virtual disks, then we resorted to good old fashioned s copy. And we started copying the data, not, not, not a sexy solution. But it actually worked fine. Um, you know, it still took us few days to replicate all the data, to get the data from point A to point B, but, um, you know, we fell back to a tried and true method. Um, it worked fine for them. We were able to then on the date, it was our cutover date, which we did [31:00] make our, you know, within the month, right? So the customer was very happy of that, we were able to fail over the four machines that were like, easy peasy, came up in our cloud, no problems. And then on that final machine, we did a final synchronization of the data over an overnight session and failed that machine over and turned them off in Azure.
And at the end of the day, the customer was like, this was great. You know, we, we were able to get out of Azure without too much and stress. And, uh, it worked out well for them.
Max Clark: There.
William Sellers: Customer,
Max Clark: Every customer, every enterprise, every everybody. When you ask them and you start talking about migrations, right? This idea of minimal downtime comes up and you know, if you’re, if you’re trading a market, right? Commodities, equities, whatever, right? You’re downtime matters to you. If you’re e-commerce and you’re doing transactions, you’ve got some sort of OLTP that’s running [32:00] internal applications that have a finite number of users that maybe you’re not, you know, actually
William Sellers: second by second.
Max Clark: second by second, critical, how do you walk through that with the customer and figure out what an appropriate migration time really is?
And setting expectations, because we’re talking about complexity and cost, right? Like sim, simple migrations, turn everything off, move it, turn it back on, right?
William Sellers: less
Max Clark: Less complexity, less cost, more time, maybe, versus trying to run two systems online. So how do you walk through that process?
William Sellers: it comes down to, you know, number one, does the customer really understand their application from, from soup to nuts?
Max Clark: Mm-hmm.
William Sellers: also do they understand the impact? And this particular customer, it, it was an internal application, but it was servicing their research group. Research means money. If they’re doing, um, energy exploration or anything like this, um, you know, they, they don’t take [33:00] time off, right? Um, the weekends maybe not. So, you know, that’s where they have a little bit of time, When we have con discussions with customers like this, you know, honestly when we talk to a lot of customers, it’s, we can’t have minimal, we have to have zero downtime. We have to, it, it, we can’t miss a ping, or something like this.
Max Clark: Yeah.
William Sellers: really,
Max Clark: Yep.
William Sellers: isn’t until you do the, the workshops you flush out what really is like, what is the high order bit, right? What is really most important sometimes when you look at, when you actually start breaking down and actually talking to the subject matter experts, the, you know, the administrators of course who, who usually say we can’t have any downtime anytime ever. But then you start talking to like the finance people or maybe some of the other decision makers and it’s like, well it costs us this much, we would, we are okay with a few hours of downtime or we’re okay with [34:00] a day of downtime. Right.
Max Clark: All of a sudden, Yeah.
William Sellers: know, that’s why. When we have consulting time with customers, these are things that we, we flush out through the workshops, right? We will either do a workshop in person. In this particular case, we, we went into our, you know, into our conference room with a big, big wall size whiteboarding, and we started through the technical challenges, but also, you know, the impact to the company. know, um, in this particular case, the impact was, you know, we can’t be down during the week because these engineers that are working in our company, get paid a lot of money. They’re the, they’re the really, they’re the big brains, right? They’re the scientists who are looking at. You know, um, energy data, geographic data, looking at, um, you know, things actually on the planet earth, right? And we, we don’t wanna [35:00] pay them a salary when they’re just sitting downtime. So they, their time is very important to us. And, you know, of course we go, well, how, how important, you know, what can we, can we quantify, can we qualify that with dollar symbols? sometimes customers can come up with that information and, and, and make the decision themselves. And other times it’s like, you know, it’s just a pain point that they wanna try and alleviate. Um, but in this case, you know, we definitely, we definitely a home run using the realtime replication product with Zerto. Uh, and it’s something that we were very comfortable with. Through our DR as a service practice, right? Because that’s something we do, a lot of right? Where we, we will replicate people from their premises into our cloud platform, you know, multi-tenant or single tenant. Um, and when disaster strikes, it’s very easy for them to hit the red button and fail over to our data centers [36:00] anyway. But for this customer, um, you know, it, it, it can be, it was the right product. have, know, some challenges, but we were able to overcome them and get the customer what they needed.
Max Clark: Today we have
William Sellers: Almost
Max Clark: cases only two, almost too many options for people in terms of how they consume or deploy compute. I’m just gonna use compute as like the, the, the overarching, you know, tool, right? So
William Sellers: data
Max Clark: you’ve got data centers and co-location bring your own resources into a facility and get space power and usually ping, right?
Um,
William Sellers: metal.
Max Clark: and then you’ve got bare metal, right? The rebranding of dedicated servers from the good old days. Now we call it bare metal. And then you get into, you know, you’re talking about private cloud options, where it’s, it’s um, you know, it’s, it’s, it’s it’s actual private cloud where resources are dedicated to one customer or you’ve got multi-tenant, you know.
William Sellers: cloud, I mean [37:00]
Max Clark: Private cloud. I mean, in this case it’s
William Sellers: environment
Max Clark: like hybrid cloud environment where the,
William Sellers: Points has
Max Clark: where Dart points has a vCloud director deployed, and then you, you kind of just keep working all the way up the stack of all these different, and then the application abstraction. Are you doing serverless? Are you doing containers?
You know what, you know, it’s like, it’s like there’s so many different flowcharts and you go all the way up and you say at the top, now you’ve got this like public cloud,
William Sellers: you know,
Max Clark: um, you know, the Azures and AWSs of the world.
William Sellers: of things that I noticed
Max Clark: One of the things that I notice a lot with people when you start into this conversation is really trying to understand the differences between, you know, if we’re here, you know, and there’s all this other infrastructure available to us, how do you decide like a,
William Sellers: the right
Max Clark: which infrastructure’s the right place to go to?
William Sellers: we
Max Clark: You know, do we, do we come from,
William Sellers: Azure
Max Clark: uh, Azure into a, a, um, a multi-tenant v cloud? Or do we go down to a private vSphere or do we actually wanna deploy our own resources or, you know, go to HyperV, all these different things. And the second part of this, I, I, I want to explore a little bit, I’ll just [38:00] prompt you with this, becomes understanding,
William Sellers: differentials
Max Clark: the cost differentials between them.
William Sellers: If you’re in, a
Max Clark: if you’re in if you’re in a public cloud,
William Sellers: you
Max Clark: there’s a lot of things you don’t know, like what is an actual VCPU mean on these platforms? How do you size your environment? How do you figure out what the overs, you know, oversubscribe rates are on a, on a processor. And,
William Sellers: You know, and, I see this a
Max Clark: you know, and I, I see this a lot where people say, oh, this is what our resources are in the public cloud, so we have to replicate this in our private cloud.
And you’re like, no, you actually don’t. You know, and, and, and so as you’re, as you’re going through and you’re, and you’re, you know, in this case, you know, more one-to-one VM to VM kind of migration project,
William Sellers: into a
Max Clark: but it still turns into a big part of the, of,
William Sellers: go
Max Clark: of like the where does it go and then how do much does it cost and how do we get it there?
Part of the conversation.
William Sellers: something that, you know, when, when we were working with a particular customer when I first, you know, was working with this super-sized vm, I’m like, do you need this much storage? know, do you need
Max Clark: Huh.
William Sellers: data and do you need all this free space? Right. Because [39:00] they did have, so they had, you know, some, some data that was landed on disks that was actually bits, bits on silicon, right. Um, for their projects. And then there was a lot of storage that was, that was empty. Right. And they, and, and it, it wasn’t something they could just resize or get rid of because the way how their, their workflow was, was to copy data between lungs, between discs in terms of compacting projects.
Right. So we’re talking about really, really big vials um, that contain, you know, seismic data or whatever. And, um, you know. It, it really comes into that, that talking to the customer and trying to understand what their workflows are, right, and, and understanding how their application works. Uh, in this particular, in this particular migration, it was relatively simple because we weren’t dealing with containers, we weren’t dealing with, um, you know, a, a, an overly [40:00] complex solution.
It was basically just I as, as a service. Um, so for us, you know, for, for doing this migration, it was relatively straightforward and, and we didn’t have to do a, you know, too much digging into how the application works. And we’re, we’re not, you know, dark Points is not an application centric company, right?
We are a, a data center focusing on co-location and cloud and I, and infrastructure as a service. So that’s kinda what, what, you know, when we worked with this customer, um. You what, we were able to, you know, the challenges we were able to overcome relatively easy. What’s funny is right after, uh, the migration, uh, Zerto came back and said, Hey, we fixed it. Okay,
Max Clark: like, great. Fantastic. We’ll keep that in mind for the next time, but I mean, how do you go through that process of deciding, you know, this workload, you know, like, like how much of this performance, how much of this finance, how much of it administration, you know, saying like, we’re gonna put this on vCloud, we’re gonna put this on private vSphere, we’re gonna put this [41:00] here,
William Sellers: what’s
Max Clark: you know, what’s, what’s, what’s driving that now for you?
William Sellers: so I would say, you know, when, when we do analysis of the machines, right? And we, we, we, we look and see, okay, do you? really need the resources you need? And in most cases, when people are sizing things in, Azure or AWS, they, they kinda like, it’s there.
Let’s use it. Let’s, let’s, let’s just
Max Clark: Mm-hmm.
William Sellers: let’s take that compute you know, we have some tools that we, we can deploy. On behalf of the customer that kind of looks and sees much resources are they really using and is it the Right. choice um, you know, to, to have a one for one, maybe we could right?
Storage is one of those things for this particular customer that was kind of like something they didn’t want to change on, but when we were looking at like CPU and ram, um, you know, we were like, you’re really not using that much, that much. CPU, you’re not using that much ram. Let’s [42:00] rightsize
Max Clark: Mm-hmm.
William Sellers: deploy in, into, um, you know, in hyperscalers or in, in our own environments that we can use to kind of, let’s see what the, what, what the application really is taking and can we sh shorten it up? then when we look at the decision. To as to where the target will be. Um, we can ask the customer, listen, you know, is a multi-tenant cloud gonna be good for you? Right? Um, just because it’s in cloud doesn’t give you, you know, the, the, the comfort that maybe you need to have hypervisor access. Maybe you’re running application that needs to talk to vCenter or vSphere, in which case private cloud, um, public cloud on something like vCloud Director maybe isn’t the right choice. you know, during that workshop, and It can be a day or it can be, you know, a few hours and spread it over a couple of days, these are the questions we ask, right? To really make [43:00] sure we’re going in the right direction for our customer. So when we, when we, when we look at the solution as a whole. Um, you know, we look at our metrics that we get from the re from the, the hyperscaler, or even if it’s in a customer prem location. get those, those metrics and we look and see is it the right choice? And then we look and see where are we going to, and is that target the right choice for the customer? Really messy.
Max Clark: It gets really messy though, right? I mean, ’cause now it’s, it’s like, you know,
William Sellers: does it
Max Clark: this go to Azure? Does it go to vCloud? Does it go to s vSphere?
William Sellers: And I, and I
Max Clark: You know, and I, and I feel
William Sellers: um,
Max Clark: I.
William Sellers: I’m
Max Clark: Um, I’m very sympathetic to it. You know, you see people trying to struggle, you know, struggling and trying to go through these decisions and, and figure out what goes where.
And, and especially with the,
William Sellers: of it,
Max Clark: you know, other side of it, which is we just, you know, we deployed this whole thing and we’ll just pick on Azure, you know, we deployed this whole thing in Azure and, and [44:00] now like, whoops. You know, we kind of, kind of, you know, and, and, you know, there’s lots of things that drive that conversation, but, um, a lot of times, you know, uh, finances become a, a, a fixed resource.
Like it’s a, it’s, it’s a limiting factor, you know, for, for every, I mean, I don’t care who you are, right. At some point. Like you just can’t spend your way out of every problem to a certain point. Um,
William Sellers: how, how much has this
Max Clark: how, how much has this changed,
William Sellers: year
Max Clark: you know, in the last year in terms of conversations you’re having with companies in terms of
William Sellers: and how
Max Clark: where they’re selecting and how they’re.
How they’re choosing because,
William Sellers: business?
Max Clark: know, the, the data center businesses,
William Sellers: center operator
Max Clark: all, all data center operators had to, I think they, they fought the cloud and now have all embraced the cloud and they’ve, you know, and, and it’s just this, I understand it’s like some stuff is gonna be in Azure and we’re gonna help you with your stuff in Azure and some stuff’s gonna be with us and we’re gonna, you know, so how, how, you know, has this shifted, um, with Dark Points and your, your approach over the [45:00] last year?
William Sellers: Yeah, that’s, that’s a great question. I, I really feel it comes down to the consulting that we do and, and, and really going from, you know, going from just having a technical discussion to also going into, you know, kind going into financials, right? And, and really getting the
Max Clark: Mm-hmm.
William Sellers: about, um, you know, what is it, what is, what is the point where it makes sense, right?
To move to, uh, away from a hyperscaler, like AWS and Azure Google Cloud platform. Um, has it changed? Um, I mean, somewhat, I mean, because, you know, we still have customers that are traditional colocation, or we have customers that are, um, dealing with things like the VMware attacks, right? The, the whole disruption with VMware and Broadcom. that is something that, um. You know, I, I think [46:00] it’s, approach has always, it is kind of been the same and has it changed over the years about, you know, how we leverage it in, in, I’m, I’m kind of, I’m kind of like not answering this question well. Um,
Max Clark: Well, you bring up the VMware, attach, you know, tax, and I don’t feel like,
William Sellers: really
Max Clark: I mean, look, Broadcom is really good about acquiring things and then maximizing their investment, right? You know, you buy, you buy something because you think that there’s an opportunity for you to make money with it, right? So.
William Sellers: Stuff.
Max Clark: Just the way, that’s just the way stuff works.
It’s, it’s more like, I don’t even think, it doesn’t even feel like Broadcom knows what they’re doing and, and their strategy’s like, it’s like, it’s like, you know, not to make a pun here, but like, you know, throwing darts at the board and then like, oh, that one hit and that one hit and this one didn’t hit, and so let, let’s adjust.
And it, and that’s creating a lot of confusion and uncertainty, but for the, for the service providers as well as for the customers. And
William Sellers: Disruptive.
Max Clark: know, that’s been very [47:00] disruptive recently.
William Sellers: definitely, so we are, we, we, we are and will always be a VMware partner and, and leverage that technology. But, um, some of the things we’re looking at, you know, we, we recently. had, um, a lot of growth in our company, right? We had a, uh, a capital investment from a company called Nova Infrastructure, right?
Which was, was on our LinkedIn not too long ago. we’re expanding, um, you know, uh, power in some of our, our critical data centers, right? So we’re able to offer really heavy, heavy, dense power. But we’re also looking at, um, kind of new cloud platforms, next generation cloud where we’re looking at things like, you know, I won’t name the providers, right?
Because we’re still evaluating and, and, and getting our chops
Max Clark: Mm-hmm.
William Sellers: the, the, the new, uh, offerings. But, you know, really giving the customer the flexibility. To choose and have an o more of a flexible portfolio than just VMware is really [48:00] critical. And we see customers, very frequently come to us and they’re like, look man, it’s getting expensive.
What do we do? You know?
Max Clark: Yeah. Yeah,
William Sellers: some customers are like, well, we’re just gonna pay that. We’re just, we’re, we’re, we’re gonna run VMware.
Max Clark: yeah.
William Sellers: And that’s fine. And we can still provide, provide that solution base to the customers. we have some customers that come to us and say, listen, we really don’t care what, we just need our virtual machines to run. We, we don’t
Max Clark: Mm-hmm.
William Sellers: hypervisor we’re on. And that, that opens up a lot of discussion and a lot of flexibility that we’re gonna start offering to customers, right. To give them that choice.
Max Clark: I mean, this comes back to like how integrated you are with the cloud platform, right? You know, you talk about the beginning, you know, part of the big process here is understanding how integrated are you with Azure and are you using Azure stuff that can’t be replicated and has to be changed, right? And so,
William Sellers: value of
Max Clark: you know, the value within VMware and vSphere and vCloud is, there’s a lot of [49:00] tooling that you get with it.
And not everybody uses all that tooling, right? So,
William Sellers: and, and we’re
Max Clark: um,
William Sellers: and more customers, you know, as we have this discussion with our, our current customer base, who or who, our customers or customers that come to us and go, what the, what can we do? This is crazy. We’re, we’re, we’re like, you know, and this is mainly people who are, off-prem, that are either replicating to us or have something in colocation. Right. And they’re like, look, we’re doing it ourselves. We’re, we’re, we’re paying for the infrastructure, we’re paying for the licensing. just getting out of hand.
What do we do? And, um. You know, we as a service provider can do certain things and, and alleviate some of the cost and complexity of, know, retooling hardware and, and, and, and whatnot. But most everyone that comes to us says, listen, you know, other than doing things like backups and replication, don’t really need to talk to the hypervisor.
We just need our virtual machines to run, you know, and, and [50:00] understanding what plugins or what integration that a customer will need to have going to a next Gen cloud, we can solve that problem, it makes it a lot, uh, a much more easier discussion of migrating into something that’s completely different than VMware.
Max Clark: DAR points has a, uh, data center footprint, and then as a result of that, a customer focus that is different from a lot of the market. Right. You know, there’s not very many people that are gonna say, we’re operating data centers at scale in Louisiana. Right. It’s just not you, you know,
William Sellers: How would you
Max Clark: it’s a, it becomes, how would you describe, like, your approach to the market and the customer that you’re looking for as a result, and how, you know, where your, where your facilities are based, where people are,
William Sellers: is that?
Max Clark: you know, what, what is that?
Mm-hmm.
William Sellers: honestly, we, we have, so in our, in our locations, we have, we have several data centers in the South Carolina area.
So we have data centers in Greenville, Spartan, Ville, Spartanburg, Columbia, and, uh, we actually have one in, in North Carolina [51:00] and Asheville. our, our data centers are sort of spread between the south central, all the way up to the southeast and Midwest. So we have data centers in South Carolina, North Carolina, Ohio, and then Indiana, and then down here in Louisiana, which is where I’m at.
I’m here in the Baton Rouge data center. We actually have two data centers are in the Baton Rouge campus. And then we have one in Shreveport, Louisiana, which is up north. Um, so really our, our customer base is customers who. Maybe need to bring in, um, you know, they, they, they, they already have their own gear, whether they’re running infrastructure, for conventional applications or running something that’s AI centric and they need to pull, you know, tons of power into a cabinet.
Right? you know, typical cabinets anywhere between five kilowatts up to 15 kilowatts all the way up to like 25 kilowatts. And even north of that, right? When we’re looking at really, really dense workloads, [52:00] um. So those customers were able to, to cater to, right? And they’re just like, I need power, I need connectivity, I don’t, I need to stay up all of the time.
I don’t need to go down. So those are the types of customers that, that we, we cater to. We also cater to customers who come to us and say, listen, I don’t wanna put anything in co-location. I don’t wanna own the hardware. I don’t wanna own the hypervisor. want to come to you and consume the resources you guys have. And
Max Clark: Mm-hmm.
William Sellers: wanna pay for what I use I don’t wanna own anything. don’t wanna own the infrastructure. And, and those are really the, the, the two major comp type of companies we host at Dark Points, how
Max Clark: Nowadays everybody refers to it as ai, but it’s really like. GPU compute, right? Like the, you know, you look at, look at data center, you know, like footprints. I mean, even today from a, um, you know, an A MD or intel socket [53:00] basis, the, the watt envelope on a socket hasn’t changed too much over the last decade plus, right?
Like, you know, you’re, you’re, you’re factoring 200 watt per per socket on A CPU. You know, disc consumes what disc consumes. So, you know, 12 and a half kilowatt is a relatively dense cabinet in that kind of environment. And then you introduce like, oh, we’re gonna put a lot of GPUs into a cabinet ’cause we’ve, ’cause
William Sellers: You ’
Max Clark: cause we’re doing, you know, something that’s using them.
And then you’re like, okay, you know, we need 25 kilowatt, 50 kilowatt, a hundred kilowatt. You know, and that density changes
William Sellers: know,
Max Clark: that, you know, changes, um,
William Sellers: pressure on the
Max Clark: that puts a lot of pressure on the data center in, in terms of.
William Sellers: hint. Hence, we, we building in more power recently because we, we, we, we see the need in the market where people, you know, legacy compute WA was like you said, five kilowatts, right? Uh, you know, I have several servers I wanna put in a 4 2 42 U rec. Machine. And if you’re running
Max Clark: Yeah.
William Sellers: and compute, [54:00] such a big deal. You know, you can, you can definitely easily get away with, um, you know, 10 kilowatt cabinet or five kilowatt cabinet. And, but when you start putting in, you start talking to customers who are like, listen, I need to bring in AI workloads or GPU, you know, Nvidia type, Nvidia
Max Clark: Mm-hmm.
William Sellers: um, racks, Nvidia, you know, uh, deployments.
Max Clark: Mm-hmm.
William Sellers: really drives up that, that, consumption of power. Um, and we’re seeing more and more of this as, as we talk to customers, I mean, ’cause people are embracing ai, uh, it could be that they’re just putting their toe in the water and just bringing in, you know, something like a Lambda server, which is maybe something relatively small that we can kinda like deal with. Or they’re bringing in full NVIDIA stacks where they just need to just chuck power. Like it’s, you know, like it’s going outta style.
Max Clark: The early two thousands,
William Sellers: Arms
Max Clark: the [55:00] CPU arms races drove a lot of power. We saw huge spikes in, in power consumption. You know, late nineties it was two kilowatt cabinets and all of a sudden it became foreign. And five, and then 10. Right. And this, and, but then it plateaued, right? And the, and the CPU manufacturers, I think all, all kind of came to the grips with like, there is a certain power envelope that we need to stay in, and that becomes the performance barrier that we have to achieve.
When does that happen with the GPUs?
William Sellers: what
Max Clark: I mean,
William Sellers: much?
Max Clark: what point is too much? Too much. Right. You know, a hundred, a hundred kilowatt cabinet. I mean, that’s, that’s a lot, you know, like
William Sellers: Um, I, I, I honestly think that, you know, as, as AI becomes more and more prevalent, That wills i i a hundred kilowatt cabinet, you know that, that’s, that’s a pretty big, that’s a lot of compute. That’s a lot of silicon. Right. And that’s a lot of, lot of
Max Clark: Yeah. Mm-hmm.
William Sellers: you have to think about heat dissipation. Are we doing something like water cooled, um, deployments? Um, I don’t know. It, [56:00] I, I’m, I think there’s a lot of change and disruption in AI right now where if you look at, you know, are you doing true GPU compute? Are you doing inference? Um, I think a lot of this is unfolding. The one thing I’m sure of is the dense, The need for dense cabinets and dense power at our data centers is continuing to increase. And that will not be going away in, in the cons, you know, in the near future. So I think we’ll probably hit a plateau, like you said earlier, I think there will be a point where we’re like, okay, you know, XX, X amount of kilowatts per rack is what you’ll use until the next big thing comes out.
Max Clark: The first data center I stepped foot in, the floor plate was sized at 50 watt per square foot. So now, you know, you had cages and cabinets, so you had a little more density when you actually got into the, um, you know, machine space, quote, unquote. Right. But I think I think about that a lot of like 50 watt per square foot.
William Sellers: [57:00] versus
Max Clark: And then versus,
William Sellers: Okay.
Max Clark: know, now you say, okay, you know, a hundred kilowatt a cabinet, you know, you’re, you’re, you’re not talking about,
William Sellers: 12.
Max Clark: I mean, 12 square feet worth of floor space, you know, if you account, you know, allocate aisle space into that as well. Like, it’s,
William Sellers: Pretty
Max Clark: it’s, it’s pretty remarkable. Um,
William Sellers: what is something
Max Clark: what is something that I’m, I’m gonna make this, um,
William Sellers: this
Max Clark: I wanna make this generic, but, but somewhat specific companies, people sitting in around a conference room table and having a conversation of, do we move this application to this cloud, to a cloud?
William Sellers: out of the cloud?
Max Clark: the application outta the cloud? Right? And that conversation that like battle is, is, you know, politely waging, right? What’s something you wish more people understood or knew
William Sellers: These
Max Clark: that were having these conversations internally
William Sellers: an
Max Clark: that would benefit from an outside expert? Like just being present and saying, have you thought about this?
William Sellers: Um, I [58:00] think in this particular customer that we were just talking about, I think having all of the decision makers, or at least the people, the kind of the, people who are the decision makers from the standpoint of finance, from the standpoint of the application developer, IT guy, you know, the guy who’s running the infrastructure, having them in the same room really levels the field because when you talk to, you know, when, and I’ve been in situations where we only talk to the IT administrator and we definitely have a. A narrative that is right? We’re we’re talking speeds and feeds and, you know, oh, we can’t have, we ha we, we can’t have zero, it’s gotta be zero downtime. But when we,
Max Clark: Mm-hmm.
William Sellers: include the, you know, the application owner or maybe the application development person as well as the financial person you know, the, the CIO or, or CTO, if we [59:00] have those folks all in the same room, it kind of opens up the discussion and it, it makes it easier for us to really come to a good decision where we’re taking everyone into account.
So, if, I would say a takeaway of what I wish when we do consulting or workshops with, with customers is we do get, you know, the entire IT or technology. Kind of technology, uh, board right in, in the discussion.
Max Clark: Mm-hmm.
William Sellers: ’cause when you’re just talking to the IT guy or IT manager or or engineer, it’s, it can sometimes be very limiting onto the entire scope and the 30,000 foot view that you get when you include, you know, really all of the application and financial people
Max Clark: I still end up in arguments online. I know you’re not supposed to argue with people on the internet, but I just sometimes can’t help myself. Um, I still end up in arguments [01:00:00] online with this belief system that going to the cloud, going to the public cloud is cheaper than running your own infrastructure or being in some sort of multi-tenant private infrastructure with provider.
What do you say to that?
William Sellers: it, two words, it depends, You, you, you, you really have to, to understand what you’re doing. And like you said earlier, you know, is your spend on public cloud or hyperscalers matching what your revenue is, right? And if it is, and you’re okay with that, spend, life is good.
Continue on. Um, but um, yeah, being, being able to, to really understand what, what, what the customer need is, is really critical.
Max Clark: So get everybody at the table. Get everybody at the table and have a real conversation about what you’re doing
William Sellers: doing right.
Max Clark: and is what you’re doing right or should you make changes.
William Sellers: Absolutely.
Max Clark: And then, and then get some outside help. William, thank you very much. I appreciate your [01:01:00] time. This is fascinating. I.
