Follow me:
Listen on:

Day Two Cloud 104: Cloud Is No Escape From Your Data Center Dumpster Fire

Episode 104

Play episode

IT and business leaders need to be more nuanced in their approach to the cloud. On today’s Day Two Cloud podcast, guest Melissa Palmer discusses why cloud hype is a problem for organizations, the need to understand why you’d use cloud, what problems you’re trying to solve, and how your premises data center can often be a better option.

Melissa Palmer is Senior Technologist & Team Lead at Veeam. She also blogs at

We discuss:

  • How cloud hype can obscure better solutions
  • Poor assumptions that the cloud is “better”
  • The value of the cloud’s operational model, and how it can be incorporated on-prem
  • Good and less-good cloud use cases
  • Correcting cloud mistakes
  • More


  1. Always ask “Why?”
  2. The cloud may not be the answer (and that’s OK)
  3. Take the time to do cloud right

Show Links: – Melissa Palmer’s blog

@vmiss33 – Melissa Palmer on Twitter

Why I Hate The Cloud –

vMiss on YouTube

Melissa Palmer on LinkedIn



[00:00:05.210] – Ned
Welcome to Day Two Cloud. And today we’re talking to someone who maybe doesn’t like the cloud all that much, you could even say that she hates the cloud. We’ve got Melissa Palmer. She’s the senior technologist and team lead over at Veeam and she might have a bone to pick with cloud. What stuck out to you there, Ethan?

[00:00:23.990] – Ethan
It’s not so much that she has a bone to pick with cloud. It’s how people use the cloud, the hype of the cloud, the dumb decisions people make around cloud and the rage monster comes out a little bit Ned the rage monster comes out with Melissa.

[00:00:37.550] – Ned
And there’s lots of wheels being spun. And if you’re not sure what that’s all about, while you’re about to find out. So enjoy this show with Melissa Palmer from Veeam.

[00:00:46.320] – Ned
Melissa, welcome to Day Two Cloud. Before we get into the topic of why you think cloud is awful and terrible and no one should ever use it, why don’t you tell the folks a little bit about yourself and what you do?

[00:00:57.750] – Melissa
Well, first of all, thank you for inviting me on this Cloud podcast. I was very excited to get your invitation. My name is Melissa Palmer. You might actually know me as vmiss. I blog over at on a lot of different IT infrastructure type of stuff. And I am currently a senior technologist and team leader at Veeam Software.

[00:01:16.710] – Ned
Awesome. So as we all know, Cloud is like the best thing ever. We know that that’s that’s just an undisputed fact. It represents everything that is great and incredible in the world. So tell me Melissa, why do you hate things that are awesome?

[00:01:31.440] – Melissa
Well, how about first you tell me, why is the cloud really that amazing?

[00:01:38.500] – Ned
If you want a serious answer and not the flippant marketing answer, in my estimation, the thing I like about Cloud the most is the availability of APIs to construct services and the fact that I can do it in an On-Demand fashion and experiment with new features and services.

[00:01:56.660] – Melissa
Those are very valid reasons to like the cloud. However, not everybody has the same reasons to like the cloud, but you kind of stated those might not actually be requirements for everybody and they might go in thinking, OK, I can do all this good stuff, but do I actually need to do it or do they really need to do that to accelerate their business? In some cases yes, in some cases, no. It’s kind of like the classic IT answer of it depends.

[00:02:22.660] – Ned
It always depends.

[00:02:23.670] – Melissa
It always depends.

[00:02:24.940] – Ned
Yeah, well, maybe we can back up and we’ll set the stage a little.

[00:02:28.400] – Melissa
Yeah, absolutely.

[00:02:29.420] – Ned
You know, we all encountered the cloud in different ways and maybe had some different impressions in our first experience. So tell me, how did you first encounter the cloud and what was your impression?

[00:02:40.370] – Melissa
You know, I was thinking about this this morning. I’m trying to remember the year that the cloud took off from a kind of buzzword perspective. And I want to say maybe like 2010, 2011, somewhere in that range. I know it was before twenty twelve. I know that for sure. I can remember exactly what it was like. You know, CIO magazine had this great article in The Cloud and everybody read it and we have to go do the cloud.

[00:03:05.600] – Melissa
I remember sitting with some of my coworkers and we were actually in the data center and we’re all like looking at each other. We’re like. But we are the cloud, because at the time, the company I worked for was one of the top five biggest VMware implementations in the world. So, you know, they’re telling us all the stuff this cloud is going to do for years. But we we do all of that already here. So why are we going crazy over the cloud?

[00:03:29.010] – Melissa
And that’s kind of what put the first bad taste into my mouth about the cloud. Right. And not even a bad taste, but just the hype factor. Right o cloud is cloud is cloud this cloud this. And I kind of saw this. Hype factor going and going, going, and I think we still have it today, and I think a lot of people don’t even ask themselves why they are using the cloud. They’re just going to do the cloud because it’s cool.

[00:03:52.530] – Ethan
Well, contextualizing going back to the time frame you’re talking about, Melissa, back in 2010, there was a lot of hype around it. These offerings were not mature. What Cloud even meant was up for discussion and if you fast forward to now, I think a lot of the hype, a lot of the promise has been delivered, even though it’s still a rapidly evolving service and maybe there’s somewhat less ridiculous hype than there was. Do you?

[00:04:14.460] – Melissa
Yeah, I wouldn’t say it’s completely ridiculous anymore. But again, it comes back to why why are you going to move the cloud? What benefit is going to the cloud going to give you? I’m asking why a lot. But how how do you get there? Right. Just because the CIO says our new directive is to go to the cloud and everybody scrambling to make it happen, are people really having the right discussions to get there in the right way?

[00:04:36.300] – Ethan
So are you asking why in part going back to your very early days impression? Hey, we’re already delivering a lot of the same services and functionality that this so-called upstart cloud is delivering. So why bother? Do you still feel that way today that we can deliver on Prem?

[00:04:52.320] – Melissa
You absolutely could. But, you know, there’s different things to think about now, especially in the last year or so, right? Let’s say I was 100 percent on Prem. What if I can’t get into the data center to rack and stack new equipment? What if there are supply chain issues and I literally cannot get the hosts I need to expand my vSphere cluster. So there absolutely are use cases for the cloud. There absolutely are drawbacks for staying one hundred percent on premises these days.

[00:05:14.350] – Melissa
I don’t feel like it’s all kind of hype anymore, but I feel like a lot of people still do get lost in the hype and maybe they’re doing themselves a disservice by doing that.

[00:05:24.600] – Ned
Right. So in your case, you were delivering all this functionality cloud like functionality at your organization. But there’s lots of organizations that have very dysfunctional data centers that might not be delivering all the things. So for me, it seems like cloud is a slam dunk because, you know, you’ve got a dumpster fire on Prem. You might as well move to the angelic clouds that are above.

[00:05:47.850] – Melissa
Let’s ask again why? Why is your why do you have a dumpster fire in your data center? Some of those reasons you might try to put them in the cloud, you’re looking at a dumpster, fire and cloud, right? So why you have a dumpster fire in your data center?

[00:06:01.300] – Ned
Valid question. I never did. My datacenters were always pristine.

[00:06:05.880] – Melissa
Pristine, majestic.

[00:06:06.930] – Ned
But I’ve heard from other people. So if those are the things that cloud can’t do for you, what is the cloud good for or is it good for anything?

[00:06:19.080] – Melissa
The cloud is good for a lot of different things, depending on what your requirements are.

[00:06:24.780] – Ethan
So give me some requirements that would indicate to you.

[00:06:29.090] – Melissa
Well let me give you a really good example.

[00:06:29.090] – Ethan
Yeah, yeah, go ahead.

[00:06:30.510] – Melissa
Obviously I work for Veeam software and we do a lot of backing up and recovering of stuff. Right. It’s kind of like what Veeam does. It’s expensive to keep a big old disk array in your data center to keep your backups for seven years or the really old stuff or blah blah, blah, blah, blah. The cloud could be a really good use case there. Send it up to the cloud. I don’t have to pay for power, cooling, maintenance, et cetera, for this big honking disk array that just sits there and collects dust. And when it gets too expensive to maintain, I don’t have to keep maintaining it or migrate off of it. I just send it to cloud. It sits there for however long. I need it to sit there due to my requirements and we’re good to go. That is an excellent use case right there.

[00:07:10.510] – Ethan
That’s the most boring use case too. I got to say.

[00:07:14.280] – Melissa
I’m a very boring person Ethan, I will have you know, I’m one of the most boring individuals you will ever meet, but so.

[00:07:20.820] – Ethan
OK, I understand your point, though, when you point to that use case.

[00:07:23.640] – Melissa
That’s a good starting point too, that’s a good starting point.

[00:07:27.630] – Ethan
There’s a there’s a cost and cost benefit analysis is being done there and there’s a performance requirement being done there. If you’re doing deep archive backups, you’re not touching most of that data. Most of the time it’s just going to sit there in case and hopefully in case it doesn’t come up very often. All right. Well, let’s flip this on its head then and go from what would be kind of a business necessary but boring use case for the cloud.

[00:07:50.610] – Ethan
To something more on the compute side. When am I putting a workload in the cloud in your mind? And that would make sense.

[00:07:58.030] – Melissa
So, again, it depends I think one really good use case out, I don’t really want to talk about networking today, but I hear a lot of people know a lot about networking. And I would say I’ve actually been immersed in the cloud a little more than usual the last couple of weeks. Networking, I think, is one of the most annoying parts of it. Right. So let’s say you have all your networking ironed out and everything works great before I give this use case first workloads.

[00:08:21.910] – Melissa
Right. I have maybe I’m a business that I see a big seasonal increase for whatever holiday is coming up. Right. And for me to be able to accommodate that workload, I can’t do it on Prem so I can go spin up whatever extra capacity I need to in the cloud, burst up there for whatever holiday seasonal use case I have. Maybe it’s something else like end of quarter processing. I don’t know, something that isn’t going to be around for a long time.

[00:08:47.980] – Melissa
Do everything up in the cloud and then throw it away when I’m done. Right. That’s another boring use case they’re going to say, isn’t it?

[00:08:54.250] – Ethan
Well, it’s not boring so much. But you know what both of those use cases you’ve cited have in common. Neither of them are primary like this is a major way I would use the cloud as as my primary place to put workloads and stuff. But I mean, I’m not digging for another use case, Melissa. But I, what is helpful to understand, since the premise of this is you kind of don’t like the cloud, you kind of hate the cloud.

[00:09:17.960] – Melissa
All right. I’ll give you a how about I give you a legit one.

[00:09:20.800] – Ethan
OK, go ahead.

[00:09:21.580] – Melissa
OK, let’s say that again, I’m at capacity in my physical data center. I’ve reached power cooling max, electricity max data centers full. I either need to build a new data center or find someplace else to host my workloads. Right. So once I do the cost analysis on building a data center, getting the equipment, building out, having the people to run it, or I can look at a major hyperscale or cloud provider, even a service provider.

[00:09:48.940] – Melissa
Right. That’s another form of cloud. Some other place to put my data, I might say, you know what? Maybe it’s time that we think get out of this data center business because us running operating a data center doesn’t really drive value for our business. We’re a, trying to think of, we’re a gluten free cookie company. We don’t care about a data center. It doesn’t get us anywhere. We don’t innovate through the use of technology or data center. So maybe it makes sense for us to start getting out of this game and not having to maintain these facilities and equipment, all that stuff.

[00:10:19.270] – Melissa
And maybe the cloud is a really good use case for us because we can focus more on making great gluten free cookies instead of making sure the data center is running properly. Is that less boring? Is that less boring?

[00:10:31.000] – Ned
Talk about trends and hype gluten free? I just bought some cured meats like a couple of days ago and it had zero trans fat because remember, that was the thing for a while that was and it said gluten free on it. I’m like, yeah,

[00:10:43.870] – Ned
Shouldn’t meat be gluten free?

[00:10:45.220] – Ned
It’s meat.

[00:10:45.850] – Melissa
I eat meat, because it’s gluten free.

[00:10:48.180] – Ned
Like, I swear there was something else on the label too, and it was always just

[00:10:51.490] – Melissa
I hope it wasn’t vegan friendly.

[00:10:54.550] – Ned
Yeah. Why not. So I mean cloud a lot of hype. And you cited that as one of the things that really bothered you is there was so much hype around cloud, but not necessarily all the good use cases to back it up, especially in the early days.

[00:11:08.830] – Melissa
Right. Absolutely.

[00:11:10.180] – Ned
What other aspects of the cloud do you think are are negatives or things that would cause you to stay on premises as opposed to adopting a public cloud?

[00:11:20.980] – Melissa
I think a lot of it will come back to your people, right. If you have an organization full of people at me that perceiving we don’t like the cloud or don’t want to learn a new technology, that could be a big problem for adoption, too. Right. Or maybe people just say, oh, well, you know, I’m used to X, Y, Z, hypervisor. That’s what I use that I’m comfortable with. I’m not going to go learn that new platform.

[00:11:42.100] – Melissa
So there’s a human element, too. But again, you could have the CIO send an email and say our new directive is amazing, majestic cloud. So we all need to start using it and maybe that’ll help. But again, I’ve seen a lot of those emails go out and I’ve heard a lot of customers say, well, the CIO says so, but not they haven’t always explained their logic. And I feel like when you get to that C level leadership level, it’s their responsibility to to educate the organization of why we’re going to the cloud, not just because I said so.

[00:12:10.840] – Ethan
Is their vision coming from the C suite, Melissa, these days, at least from what you’re hearing, where it’s migrated, where it’s migrated from, the old argument, we’re going to save money and gone to a maybe a new argument about agility.

[00:12:24.970] – Melissa
The cloud isn’t a charity. You aren’t going to save money necessarily.

[00:12:28.050] – Ethan
Well, exactly. And that’s the root of the question, because the C suite does seem to be finally getting it. This isn’t about saving money, moving to cloud. It’s really about transforming how we do ITOps.

[00:12:36.790] – Melissa
It’s about aglity and, yes, transforming ITOps and getting out of that physical real estate data center business too, right. Because, again, unless you’re unless you’re innovating in that space, what does it get you at this point other than control right? Like, if for some reason, you need control. All right, then I guess you’re going to do everything in your data center.

[00:12:57.420] – Ned
Yeah, so the cost thing is definitely I think we’ve seen that ship sail and some of the newer arguments that I’ve seen have been more around that agility. Do you think there’s the possibility of moving the cloud operation model from cloud, actual public cloud to just adopting it on Prem instead and not making that migration?

[00:13:21.480] – Melissa
Well, it depends on what how you’re going to consume the cloud, I think that’s one of the reason that some of the VMware branded clouds have done so well because you can basically keep your operations model. Ninety five percent of the same, I’m going to say, because somebody still needs to go into the cloud control panel or portal or whatever they call it and actually deploy vSphere, I still have to log into the cloud at some point, so I have to do a little bit of cloud stuff.

[00:13:44.050] – Melissa
Right. But I think that’s part of the reason that model’s been successful because hey, I get out of the data center game, but I can leverage those same operational skills. Look, any kind of skill change in an organization is always a struggle. People don’t like to change. And so it’s like a human thing, like people don’t like change. If I’ve been making the same widget the same way for 20 years and you’re going to tell me, well, we’re going to make the same widget, but we’re going to do it a new way.

[00:14:08.290] – Melissa
I’m used to my way. My way is great. Why do I want to use this new way? Unless you can prove to me that it’s better.

[00:14:14.960] – Ned
Right, so let’s say you are a VMware architect working at a company. Imagine that and.

[00:14:22.860] – Melissa
So funny.

[00:14:23.690] – Ned
And someone rolls along and there they said, you know, we’re going to start migrating our applications to the cloud, what are the key questions that you would want to ask them to dig deeper behind that?

[00:14:34.700] – Melissa
Obviously, I’m going to ask them why they’re doing it again. Right. And agility, perfect answer. Cost? Hmmm. Maybe we should go back, but it goes back to why are we doing this? And then proper planning, taking the time to say, OK, how are we going to do this or are we going to do it in stages? Are we doing it by application or are we doing it by data center? Are we doing it by vSphere cluster.

[00:14:53.090] – Melissa
Let’s start down and have a good session where we understand everything we have today in our data center and we put together a plan to get it all into the cloud. And how long does that plan take? Do I have time constraints around it? Can I take my time and do things the right way, or is it just. No, you need to go to the cloud right now because we have no more space on prem. Like, go, go, go.

[00:15:13.880] – Ethan
You just said the right way. And it sounded like earlier you were describing kind of a lift and shift. Let’s figure out what we’ve got and get it up there.

[00:15:21.440] – Melissa
Yeah. So it kind of depends. Right. If you just like, oh, man, I have this really troublesome application. It runs like garbage. That’s my first candidate to move to the cloud and I’m just going to shift it right over. And it might still run like garbage because there’s probably something else wrong with it that the cloud is not going to fix. And I think a lot of people don’t realize maybe problems they have on Prem and then they might notice them in the cloud because, you know, I’ve never really looked at this application before, but they complain sometimes but I kind of ignore them.

[00:15:52.370] – Melissa
But again, moving things is a great opportunity for app owner to just be like, well, it was kind of OK yesterday and it’s no longer OK now. And I know you moved it to the cloud last night, so you need to get in there and you need to fix it.

[00:16:03.440] – Ethan
It was garbage. Now it’s hot garbage.

[00:16:05.360] – Melissa
Yeah, it was garbage, but now it’s like a flaming dumpster. So you need to fix that.

[00:16:11.570] – Ned
Yeah. And so often that comes down to things like poor planning, which you mentioned, and not understanding the application dependencies in your environment. So that application that was just kind of functioning OK, well it was relying on this other application and

[00:16:25.370] – Melissa
That we didn’t move yet. Oh no!

[00:16:27.140] – Ned
So those two applications are no longer sitting next to each other. They are now sitting fifty milliseconds or one hundred milliseconds away through a pipe that might not be that big yet. And Oopsy Daisy.

[00:16:38.930] – Melissa
It happens.

[00:16:40.610] – Ned
It does. Yeah. So, OK, so if I am thinking about the cloud or I’ve gotten that directive.

[00:16:47.630] – Melissa
Please take your time, please take your time. Please figure out why you’re doing this and take your time and plan it properly. Right. Would you just. OK, I shouldn’t answer this because I literally have done this. Would you just like go migrate your whole data center? I’ve done that actually. Like we had a project where we’re closing the data center in six months.

[00:17:03.440] – Melissa
You’re going to get it all out because we’re going to close it. Just go, go, go. And it was just go, go, go. And it was challenging, but that’s how it was done because there was this artificial time constraint. But whatever.

[00:17:15.770] – Ned
That’s a really interesting situation and I feel like you might have some scars from it. And I don’t want you to revisit trauma.

[00:17:22.000] – Melissa
Still bitter.

[00:17:23.180] – Ned
I would like you to dig into that a little bit more, because I think especially for the audience, that sounds like a really interesting project where you probably learned a ton that maybe you didn’t want to.

[00:17:32.800] – Melissa
Yeah, so my company was acquired and they just like looked at the calendar. So we want everything out of the data center by then. Go do it, go do it, and you can’t do the things that, like, we will come up with all these ideas. No, no, no, no, no. And it got to the point where we just had to pick a plan and go. It’s always interesting. And it’s funny because, again, this was this was probably 2011 there.

[00:17:58.080] – Melissa
Literally the most ironic thing to me is one of the products I work with at Veeam. It’s called Veeam Disaster Recovery Orchestrator. It literally does exactly what I did manually for six months of my life, like I could, like, click a button and be done besides a P2V part. OK, I could back it up, restore it to VM, you know, whatever. But besides the P2V of the actual OK, so we P2V’d it. And then we migrated to the storage and we replicate it for a week and then we turned it off and we mounted the storage and we registered it to vSphere and we ran a bunch of stupid scripts to convert it to the new systems like that’s all I did for six months was horrible.

[00:18:36.960] – Melissa
And today there’s stuff out there like, oh yeah, OK, I’m done. Have a nice day.

[00:18:41.400] – Ned
Yeah, I’ve worked with I haven’t worked with the Veeam software specifically, but Azure has a migration tool. It’s a disaster recovery tool. But the thing about disaster recovery is.

[00:18:50.850] – Melissa
It’s just uh, migration is just a disaster recovery you actually plan for. It’s same thing.

[00:18:55.950] – Ned
And you have to be able to deal with the whatever the loss that you would have. So you have to gracefully shut that application down, do the last replication and bring it up in the new environment. But the nice thing is, is you can you could test bringing it up in the new environment.

[00:19:08.070] – Melissa
You can work until you get it there and you’re like.

[00:19:11.850] – Ned
I don’t know until users start hitting it and they’re like, oh, we totally mis-sized this thing, we should have put this on fast disk.

[00:19:18.300] – Melissa
Well, I was doing P2Vs while I was migrating on them. So it came to a point where we had to like P2V and then like I think we ended up P2V’ing and then waiting two weeks and then cutting over because when we were doing it so close together, like people were just going nuts and like they were like making up issues like because it was because you P2V’d me no, it wasn’t like crazy, but that was like a thing because people knew things were changing.

[00:19:43.710] – Melissa
Well, you converted me to from physical to virtual, OK, that old garbage server that, like, doesn’t even have support on it anymore and is completely trash, like one CPU and two gigs of RAM. Trust me, the VM is better. Like you’re not worse off than you were before.

[00:19:58.780] – Ned
Oh I’ve fought that battle.

[00:20:01.350] Oh P2V is clear. At least the cloud, you don’t usually have to do P2V. You’re usually sending virtual stuff or building new, you lose that P2V element. Can you imagine doing a P2V to the cloud? Maybe I’m going to feel like I’ll do that later, actually, like just for fun. I think I would go for it.

[00:20:18.090] – Ned
I’ve actually done it, so.

[00:20:19.740] – Melissa
I’ve never done that before.

[00:20:20.860] – Ned
Now, I moved a RHEL6 that was running SAP up to Azure using their migration tool and it worked seriously. It took a long time, but it did work eventually. Yeah. So that sounds like a nightmare from a planning perspective too, because you have to coordinate not just between the technical teams, but also.

[00:20:43.070] – Melissa
Yeah the application owners too. Right. And again, those application dependencies. I can’t tell you how many of those we missed. Then it was like a week or two of because we used to call it the bus right.

[00:20:51.570] – Melissa
We sent the bus down to the new data center, which was just the replicated volume. So we let it replicate for a week and then cut over. You’re watching the bus all week long to make sure it would make it on time like it was crazy. But like, sometimes we will have to change a priority because, you know, the bus parked in the new data center and we realized we forgot a bunch of stops. All right. That goes right on the next bus and bounce that stuff off.

[00:21:10.710] – Melissa
And it was just poorly executed.

[00:21:13.380] – Ned
Oh, that’s OK. Well, I’m glad you survived.

[00:21:16.830] – Melissa
I did, still here.

[00:21:18.090] – Ned
That’s what I can say at the end, but I can see how that might make you unhappy with the cloud a little bit. Or at least that project in particular. Sounds like you could use a therapist.

[00:21:28.860] – Melissa
Really. Do they make like therapists for the cloud.

[00:21:32.370] – Ned
Well, they do. My buddy Bobby Allen, who’s now over at Turbonomic

[00:21:36.990] – Melissa
Joining IBM?

[00:21:38.270] Well, yeah, I guess now. Wow, that did that last week. That happens. It’s going to take me a while to process that one. So he has called himself a cloud therapist. In fact, that is his title on LinkedIn is Cloud Therapist.

[00:21:50.950] – Melissa
Well can I read what he calls a cloud therapist? Allow me to read this.

[00:21:54.360] – Ned
Yeah, absolutely.

[00:21:55.410] – Melissa
“Cloud therapy to me is really two things. One, I’m focused more on how I can help you than what I can sell you. And number two, I’m there to tell you what you need to know, not what you want to hear.” I actually agree with that. A thousand percent. I think I might be some kind of cloud therapist, cloud coach, too, because I’m always trying to dig down to why are we doing this? And here’s the stuff you really need to know.

[00:22:17.010] – Melissa
Like you’re trying to, like, ignore. I’m just going to send that flaming dumpster up there and maybe it will be OK. Like it’s not going to be OK. I like it. It’s just not. I actually really like that. And I think we need more of that instead of just like, hey, let’s go to the cloud and everything will be fine. Like how many how many vendors will be like we need to like stop and like plan this like instead of yes, please give me your money and let’s go to the cloud today.

[00:22:41.730] – Ned
Yeah. I mean, as someone who worked for a VAR and did cloud style consulting, we were not beholden to a number with Azure or AWS. So we actually were able to ask those questions and say, do you really need to move? Because if you move to the cloud, that’s fine. But if you wanna stay on Prem, I can sell you hardware.

[00:23:02.910] – Melissa
I don’t carry it around. Kind of like that with Veeam like Veeam kind of has a solution for on from cloud. Like I don’t really care what you want to do. I’m here to help you and I’m here to help you in the best way possible.

[00:23:11.520] – Ethan
So there’s a catch here with this moving to the cloud thing, though Ned, this has come up a bunch of times and it goes back to that idea of doing it right. That is, if you just going to pick up a workload and dump it in IaaS and call it done, you didn’t change. Nothing changed except for where that compute is happening. That move is all about operational change. And that’s that that’s the sticking point.

[00:23:31.980] – Melissa
The human element right.

[00:23:33.360] – Ethan
Well, it’s the human element, but there’s also an architectural element. If you need to rethink how your app is delivered so that it is cloud native, cloud friendly.

[00:23:41.940] – Melissa
Like this is an opportunity, right. Is this not almost an opportunity.

[00:23:46.110] – Ethan
But who’s got the time for that.

[00:23:46.110] – Melissa
To start the kind of application that’s a problem? Right. Because I got this garbage application that’s like so important to my company, but it’s trash. Maybe we take the time to rearchitect it up there, take advantage of the things you can actually get out of the cloud that you don’t have on Prem and do it the right way and make it better going forward.

[00:24:07.401] – Ethan
Or just modernize it.

[00:24:08.130] – Melissa
Instead of we’re just going to send it to VMC and it’ll be fine.

[00:24:12.670] – Ethan
Sometimes that modernization, the opportunity, taking the time, putting the project together, getting the people on board to make it happen is sometimes so difficult that you can’t pull it off. Going back 20 years This has been a problem in IT. But pre-cloud I was supporting a shop that had a big app running on a mainframe. That mainframe cost some ridiculous six figure amount of money per year to keep that thing going. There was a project instantiated to take the app that was sitting on that mainframe, move it off to I think it was some kind of an IBM platform that was more client server friendly, smaller, x86 based, and it would save an enormous amount of money.

[00:24:47.950] – Ethan
Didn’t happen because couldn’t put it together, couldn’t get the people behind it, couldn’t drive that that project ahead. And it feels like it’s 20 years later. Now it’s the same thing. Instead of a mainframe, we’re talking about a data center and we’re talking about trying to get things really architected for cloud. And we just can’t just can’t do it somehow.

[00:25:06.040] – Melissa
I mean, stuff is kind of hard too again, it comes back to people and what’s the function of an organization. Right. I can’t tell you how many times I’ve even seen that. Just going to a new VMware environment. Right. Well, this is garbage. We should fix it going forward. But we didn’t because no one could agree on anything. So I think it’s really hard. And part of that kind of almost cloud analysis phase, when you identify the application level stuff, maybe we should, like, try to put some dollars around it and say I could, not that I could save money with the cloud, but like, if this application ran better or I could make it faster or I could do it with less or whatever, what does that mean to my business?

[00:25:45.590] – Melissa
Can I make more money if I make this better? Can I lose less money if I make this better? Like, what can I do and how can I tie that to some kind of business driver or dollar amount to make the people in the organization see that this is something that we should spend time on.

[00:26:03.010] – Ethan
There you go. If there’s a business opportunity that’s being created by moving to this new model of delivering IT services, can you then get the force of will required within an organization to drive the project ahead? And I think the answer is yes, but it’s an incredibly hard conversation to have.

[00:26:19.900] – Melissa
It is.

[00:26:21.100] – Ned
Melissa, you’ve mentioned this several times. People don’t like change. They don’t want you to move their cheese. I don’t know if you’ve ever heard that wonderful expression I did.

[00:26:28.660] – Melissa
What like the mouse and the cheese?

[00:26:30.550] – Ned
Kind of like you moved where my cheese is. There’s a book called Who Moved My Cheese, which I refuse to read just based off of the title alone. But I did this one consulting engagement where one of the people there had read the book and was clearly obsessed with it and mentioned people don’t like their cheese being moved in every conversation we had, which really, you know, drilled into my head.

[00:26:50.170] – Ned
But the point is people don’t like change and people don’t like changing the way they do things. They also don’t like things being in a different place. They have an expectation and if you don’t. It’s sort of gently prod them towards new expectation, then they’re going to react badly.

[00:27:08.930] – Melissa
I mean honestly, you have the same problem on premises when you go to the new version of the vSphere client. I know there hasn’t been like a new one in a while, but from the C# client to the Web client to the now HTML5 vSphere client, right?

[00:27:23.470] – Melissa
That was a big operational change, like each of those three transitions. That was gigantic. And if you survived it and I keep going back to VMware example, if you survived that, well, you can probably survive logging into whatever cloud is your favorite at this point. Right. But again, it comes back to education, training, getting people used to things. And like, that’s a good example. Like, you don’t almost don’t have a choice in some of those cases.

[00:27:48.180] – Melissa
Like literally there’s no more C# client after a certain version. You can’t anymore. Like, they forced it eventually. Right. So, you know, people are just resistant today. But we maybe part of it is tying it back to examples of times that, you know, OK, whatever software we use, well when we updated the latest version, it completely changed the UI. You survived that. So maybe we can survive this to.

[00:28:09.450] – Ned
Right, people like doing things the way that they do them, and you have to show them that tangible benefit and businesses are this are just like a bunch of people. Businesses are made of people and businesses have momentum.

[00:28:20.580] – Melissa
Can I click on that? Businesses are made of people.

[00:28:24.180] – Ned
I think I think people forget that sometimes that businesses are just made of a bunch of people that all have some sort of cultural momentum behind them. And if you’re going to shift that momentum, there has to be really good reasons behind it or you have to do it by fiat. And when you do it by fiat, that CIO directive, it’s not going to go well.

[00:28:42.660] – Melissa
Unless, again, your CIO is explaining here is why we’re doing this. And it makes sense for our business and our organization versus no, I spun the wheel of cloud and I landed on.

[00:28:55.530] – Melissa
Azure, we’re going to Azure.

[00:28:58.350] – Ethan
But so Melissa, a serious question here, the premise of this has been that you you hate the cloud when the right reason comes out of the C suite for why we’re moving to cloud. Are you OK then?

[00:29:10.950] – Melissa
Absolutely. If the cloud meets your requirements, go for it. I love the cloud. If it does what I want it to.

[00:29:16.830] – Ethan
You’re not arguing that the cloud doesn’t really live up to the hype at this point or that we can just deliver it all on Prem and do pretty much the same thing?

[00:29:25.410] – Melissa
I mean, if you had enough money, you could do anything anywhere, right? No, I kind of say I hate the cloud because to try to counteract some of that cloud hype. Right. Because I feel like it is a little too hype still. And because of the hype, people are just going in and not planning properly. That’s probably my biggest pet peeve with the cloud that people like. Oh, I got my credit card. Let’s go.

[00:29:48.790] – Melissa
Woohoo. And you don’t do the proper planning up front. You’re not going to have a good cloud experience and you’re going to end up hating the cloud. Hello.

[00:30:00.300] – Ned
Yeah. Yeah. Oh, jeez. So if you do make that move to the cloud.

[00:30:04.470] – Melissa

[00:30:04.980] – Ned
And things are not working well, do you just tough it out or is there a use case or is there like a case to be made for moving stuff back?

[00:30:13.080] – Melissa
You know, I think it depends on when you say it’s a dumpster fire, is it a dumpster fire operationally is a dumpster fire because it’s not working right, and we’re losing money. Like, what is the reason for the dumpster fire? I don’t think there’s anything wrong with bringing it back until you can plan it the right way. There’s nothing wrong with saying, you know what, this isn’t serving our business. Let’s revert until we can figure it out.

[00:30:33.060] – Melissa
I also think that’s I don’t know if I should say this or not. Right. The VMware based Cloud Solutions can also be a good intermediary type of thing, like, OK, I have no choice. I have to get off now because I’m out of space in my data center. Let me go here. That gets me out and take my time and evaluate would these applications be better served deployed in the cloud native manner versus as a VM? At that point maybe some of your workloads become cloud native or maybe some of them are OK as VMs and they stay that way. That’s fine too.

[00:31:01.140] – Ethan
Yeah. VMware has made that argument that if you go to VMC, then it could be that for the life cycle of that app, whatever it is that is, it’s final resting place.

[00:31:09.030] – Melissa
Maybe it stays there and next time you go cloud native. I’m actually doing a product project on a unnamed VMware cloud. I will not say which one I’m playing with right now, but I like literally have cloud native stuff, talking to my vSphere environment, because I have a reason to do that and it’s actually pretty cool, I’m going to say that.

[00:31:28.890] – Ned
One of my favorite parts about and it’s not really cloud specifically, but it’s something that was enabled by cloud is the automation and sort of CI/CD pipeline, infrastructure as code, all of those things.

[00:31:39.510] – Melissa
Anybody remember Lab Manager back in the day. Lab Manager?

[00:31:43.380] – Ned
Why would you bring that up?

[00:31:44.520] – Melissa
Long live lab manager.

[00:31:46.500] – Ned
I mean, for listeners who aren’t familiar, do you want to describe it like a little bit?

[00:31:51.370] – Melissa
Lab manager eventually became what does it even call now? VMware Cloud Director.

[00:31:56.010] – Ned

[00:31:56.950] – Melissa
Used to be called vCloud director and I had another name before that, between lab manager, and vCD I can’t remember what it was, but it was kind of like the beginnings of some kind of self-service VMware stuff. It was pretty cool. And we were using it in our little cloud on Prem. Why not? It was the cloud. So I was like an early adopter of a lot of really weird VMware stuff that didn’t make it necessarily.

[00:32:18.720] – Melissa
But every time I see the vCloud director like Anthony Spiteri on my team is like the vCD guy. Every time he does something on vCD, I just like tweet back, long live lab manager. Why not?

[00:32:30.510] – Ned
I used it a little bit and I was like, oh this is amazing. And then, well, it got more complicated and I just I don’t know. It always made me a little bit sad, I think, because I feel like VMware had a moment where they could have copied what AWS and Azure were doing with their portal and made it very friendly to the end user, not the VM admin, but the actual end user. And they didn’t end up doing that for for whatever reason.

[00:32:57.660] – Ned
And I was like, that was a miss. But that’s the thing I like so much about the cloud, is it does have that friendly console for the end user. It does have the API.

[00:33:06.150] – Melissa
Have you logged into the cloud lately? I actually I will say this and this is how far I’ve come in the last year and a half talking to someone on my team about something. Like does it have an API? Can I just use the API? I don’t want to do this. Does it have an API?

[00:33:19.890] – Ned
Oh my goodness.

[00:33:21.240] – Melissa
Like that is how far I’ve come in the last, I actually figured out, I’ve been like dragged like seriously into the cloud for about a year and a half now. Rick Vanover at Veeam is my boss and like one day basically told me, like, you’re doing Cloud now. And I’m like, oh kay. But I actually really do enjoy, like, the APIs and stuff like that. And I even have started using APIs not in the cloud with on prem stuff, believe it or not, like I’m actually digging some of this stuff. So there definitely like lessons you can take from the cloud and apply in your data center to make things better too as well. Right.

[00:33:55.740] – Ned
Yeah, so.

[00:33:56.730] – Melissa
By the way, it didn’t have an API, I have to do it manually.

[00:34:03.030] – Ned
Shed a tear. Pour one out. That is sad. Oh no, no.

[00:34:07.660] – Melissa
Oh god can I just. No?

[00:34:09.150] – Ned
Can I please? It always amazes me when any cloud introduces something that’s only available in the portal. I’m like, what are you doing? So you’ve been immersed in the cloud for, you know, a little while now?

[00:34:21.840] – Melissa
I’m still pretty immersed in the data center, too, like I’m pretty still.

[00:34:24.900] – Ned
So are there some positive examples that you’ve been or projects you’ve been on that have been adopting cloud properly? And if so, what were the lessons you learned from those?

[00:34:35.280] – Melissa
So I’ve actually seen a lot of customers be really successful in the cloud and it’s usually when they’ve taken the time to decide what is our cloud strategy as a company. I see a lot of people using multiple clouds as well, but I haven’t decided. I haven’t spun the wheel of cloud and decided on Google and just stuck everything Google. Right, because you guys are going to have another rant coming. One of the big things I don’t like is when you like architect badly in the cloud.

[00:35:03.480] – Melissa
I’m just going to put everything in the same availability zone and we’re good to go. So, like, the cloud is like a data center, right. It’s got a little physical constructs and it is a server someplace. So would you put everything in one data center? Probably not. So stuff like that kind of irks me. But the most successful I’ve seen where people have actually taken the time and been choosy about what they send. And a lot of times people aren’t sending the hardest apps first. They’re sending something that’s not as critical, because if it takes us a while to figure out the networking, that’s OK, because no one’s using this much or they’re taking that app that like maybe it is more critical. But this is the app we wrote in-house and we know it really well. So if we want to rejigger for the cloud, we have the skills in-house to do that versus we bought this off the shelf. How are we going to get it in the cloud?Right.

[00:35:49.680] – Ethan
Is that actually pretty common, Melissa, where some of the success stories have developer driven organization or at least a developer driven app that something built in-house so that that process thing.

[00:36:00.210] – Melissa
It’s easy then right. It’s easy to do that versus. Oh, well, we don’t have a guy who wrote this app anymore or whatever.

[00:36:08.910] – Ethan
There’s a process thing here, too, right? So if you’re developing apps in-house as opposed to buying them and consuming them, you have this pipeline and workflow.

[00:36:17.430] – Melissa
It’s a lot easier to do that right. To start changing that and adapt it to the cloud versus. Well, you know, I talk to the software vendor and I can use a Windows VM in AWS to run it, but it’s going to run just like what it does now. It’s not going to change anything. But, hey, that’s still my work, right? Because you got windows in the cloud, analytics in the cloud. I can get anything I want. It’s a cloud.

[00:36:40.620] – Ned
Yeah. AWS is the largest consumer of Windows licenses.

[00:36:44.430] – Melissa
Is it really? I didn’t know that.

[00:36:46.140] – Ned
It certainly is, lotta Windows workloads, which is.

[00:36:50.190] – Melissa
Which is just fine. Like, you don’t have to you don’t have to even and you don’t have to be that developer driven person or organization like, let’s say, Melisa’s gluten free cookie company of one person. Right. If it was me trying to move my whole infrastructure to the cloud, I’m not a developer. Let’s say I’m using a lot of windows and off the shelf apps. Chances are if I take a good look at that app, it’s dependencies, its components and how it communicates, I can still successfully put it in the cloud right? But it might take a little more work.

[00:37:18.420] – Ned
I was going to say, chances are for a lot of that COTS stuff. There’s a software as a service out there that just does it.

[00:37:23.760] – Melissa
That’s true. We haven’t even talked about that yet. Melissa’s gluten free cookie company as a service.

[00:37:30.240] – Ned
Only if it’s trans fat free. I can’t take any grams of that, you know, in my cookies. Are we all right? We’re eating cookies right now. Or biscuits. I hear they’re also called.

[00:37:41.640] – Melissa
Across the pond? They’re biscuits. Right.

[00:37:45.360] – Ned
Well, Melissa, this has been a fascinating conversation into why you hate but maybe don’t actually hate the cloud.

[00:37:51.540] – Melissa
I don’t hate the cloud. I hate bad implementations of the cloud. I hate watching people suffer trying to migrate to the cloud. I hate when people don’t have the opportunity to do it the right way because someone said, let me spin the wheel cloud. We’re going to AWS woo-hoo like. That’s what I don’t like. Of course, I got all the area. I have a really funny story about how I picked my favorite cloud too. So if anybody knows me, I’m very big into rocket launches and space.

[00:38:19.860] – Melissa
So there’s three big cloud providers, right? Microsoft Azure, Amazon Web Services and Google platform. I’ve never been a Microsoft person, I don’t have an Xbox, just not happy, happy Microsoft, so Azure you’re out. Which brings us to AWS and Google cloud platform. This has nothing to do with the cloud. It’s hilarious. This is of my illustrations on how people get crazy over picking their favorite cloud. So we’ve got AWS with Jeff Bezos, who I know he’s not there anymore, but he was and he is a rocket company and he poured billions and billions of dollars and he can’t get a rocket to orbit.

[00:38:55.310] – Melissa
If you give me a billion dollars, I’m pretty sure I could get a rocket to orbit. And then we have Google of Google Cloud Platform. And Google was actually a very early investor in SpaceX. So therefore. Google cloud platform is my favorite cloud

[00:39:11.420] – Ned
Because of SpaceX?

[00:39:12.350] – Melissa

[00:39:13.520] – Ned
It makes perfect sense.

[00:39:14.360] – Melissa
It doesn’t matter. It doesn’t matter. Just pick the cloud that meets your requirements and works best for you. Really, it doesn’t matter which one you pick, it’s just someone else’s computer.

[00:39:23.420] – Ned
All right, well, to sum up the whole episode, what are your key takeaways about it?

[00:39:28.790] – Melissa
First and foremost, if you’re faced with an adoption of migration to the cloud, ask why, why you’re doing it? Are you doing it because someone spun the wheel a cloud and said, Alibaba? Like, who knows? Right. What is your motivation? What are the business drivers? And then take the time to do it right. Right? That’s my number two: take the time to do it right. If you can’t do that intermediary step, if you’re a big VMware shop use a VMware cloud solution.

[00:39:54.020] – Melissa
At least get the stuff. Release the pressure from your data center if you’re out of space or whatever, and then take your time and do it right. And then cloud might not be the answer for every application right now. And that’s OK. Just because everybody else is going all in on the cloud or everybody else is moving to the cloud. If it doesn’t make sense for you and it doesn’t meet your requirements, it’s OK not to go yet or it’s OK not to go all in yet.

[00:40:16.700] – Ethan
Does that mean ROI to you as far as why it might not be the answer or just. It could be anything.

[00:40:21.950] – Melissa
It could be anything. Right. ROI I think we talk about the Cloud isn’t a charity, so maybe you’re not saving money, but, you know, if it just doesn’t fit or you can’t do it right. Right now. But, you know, in six months we’re starting a new infrastructure refresh plan. So maybe in six months we start our planning or we plan now to move in a year or something like that. That’s OK too. You don’t have to go to the cloud right this very second.

[00:40:45.200] – Ned
It’ll be there when you’re ready.

[00:40:46.280] – Melissa
It’ll be there. It’s not going anywhere. It’s not going anywhere.

[00:40:49.760] – Ned
All right. Well, Melissa why don’t you tell the good folks out there where they can find more of your interesting thoughts and ramblings.

[00:40:56.150] – Melissa
My Interesting thoughts and ramblings? Best place to start is probably my website You can also find me on YouTube, Twitter and a lot of random different places on the Internet. So start with my website, And go from there.

[00:41:10.250] – Ned
Awesome. That’s a good place to find it. That’s V-m-i-s-s dot net. So definitely check it out. In fact, I think you have a I hate cloud rant on there right now.

[00:41:18.830] – Melissa
I do, actually, as a matter of fact. And the number of people that didn’t actually read it and just came at me was astounding. And I said to them, go read it and then come back and tell me that I’m an idiot. And I was like, no, actually, I kind of agree with you.

[00:41:34.160] – Ned
Wait are you saying people just read the title and don’t read whole thing?

[00:41:38.840] – Melissa
Oh grr grr, what do you mean you hate the cloud? Go read the blog post.

[00:41:41.510] – Ned
Oh, dear. Well Melissa Palmer, thank you so much for being a guest today on Day Two Cloud.

[00:41:46.400] – Melissa
Thank you for having me so much. And as you can tell, after listening, we talk a lot. I don’t necessarily 100 percent hate the cloud all the time. I just want everybody to be successful in their cloud endeavors, that’s all.

[00:41:57.710] – Ned
You’re a cloud therapist. That’s what you are.

[00:42:00.280] – Melissa
OK, sure.

[00:42:02.470] – Ned
All right. And hey, listeners out there. Virtual high fives to you for tuning in and checking out what Melissa had to say about the cloud. You know, if you got suggestions for future shows or guests you want us to have on the show, we want to hear about it. You can hit either of us up on Twitter at Day Two Cloud show. Or you can fill out the form on my fancy website, Ned in the cloud dot com.

[00:42:23.420] – Ned
You know, if you happen to be a vendor of some kind and you’ve got a way cool cloud product you want to share with an audience of I.T. professionals, you could become a Day Two Cloud sponsor. You’ll reach several thousand listeners, all of whom have problems to solve. Maybe your product fixes their problem. We’re not going to know unless you tell them about your amazing solution and you can find out more about this at packet pushers dot net slash sponsorship

[00:42:48.150] – Ned
Until next time just remember, cloud is what happens while IT is making other plans.

More from this show

Day Two Cloud 174: Building Kubernetes Clusters

On today's Day Two Cloud podcast we walk through how to build a Kubernetes cluster to support a container-based application. We cover issues such as what constitutes a minimum viable cluster, rolling your own vs. Kubernetes-as-a-service, managing multiple...

Episode 104