On-Demand Mainframe Environments in the Cloud: Accelerating Mainframe Modernization with PopUp Mainframe

Hi everyone. I'm Guy Currier analyst at FU and Group, and I'm pleased on this text on TV episode to introduce Gary Thornhill. He's the CEO and founder of Popup Mainframe, and we're gonna be talking to him today about popup and about Mainframe.

So, hi Gary. Welcome. Hi.

Thanks for having me. As my name is Gary Thornhill. I am the CEO and founder of Popup Mainframe.

And interestingly, we created Popup Mainframe originally in a, through a DevOps consultancy, actually solving directly customer issues around, uh, addressing the shortage of dev test environments. So that's, so that's how Popup Mainframe was founded, um, addressing a, a particular client problem. And then it wasn't just them that need this, it's, it, it's, there's, there's an issue around the speed of, uh, mainframe change.

So popup mainframe accelerates mainframe change, and it does it more cost efficiently than ever by providing mainframe delivery teams with an immediately available fully functioning mainframe. Our product revolutionizes mainframe delivery by removing the most common place of challenges, which is the bottlenecks caused by the shortage of environments, access issues, and quite often too many teams and too much bureaucracy. So we, we, we hit those, uh, directly head on.

So you're providing mainframe environments as a service and hosted in Azure? Correct. And I know you had a product announcement, we're gonna get to that, but, but it, it's pop up in the sense of, you know, with an Azure account takes you a few minutes and you have a, a mainframe environment, a a z systems mainframe environment available to you.

Is that right? Yes, and it, but it's not just Azure. We can actually run in any cloud.

So any X 86. Yeah, so any X 86 architecture, so that can be on-prem, it can be Linux based, windows based, and in any, any hyperscaler, we can pop up a mainframe in less than 10 minutes and they can be logging onto their very own, um, mainframe, albeit it's running on Linux, not on on, not on Z. Okay.

But it's, it's, it's compatible. It's a, you were talking DevOps. Um, I know that in the mainframe world, um, uh, for many years now, um, there has been an increase in, you know, uh, the cloud integrations or cloud-like behavior, um, for the mainframe, even as it continues to fulfill its core mission of some of the most, um, intense mission critical, um, key applications and, and, and data processing, um, uh, scenarios going on in the world right now.

Right. Um, what, what would you say is the, is the current state of the mainframe like today as far as that progress has gone and what sort of uses it's being put to today? Because it's, it remains a vibrant market.

I mean, yes, and interestingly enough, I think the, the mainframe market and, and IBM are selling mainframes with, with 10 digit growth every year now. Um, so it, it is never been more in demand for, its, its processing and reliability, but Reliability is a big thing. Yeah.

Yeah. It, it's really the only system that has that really true hot failover because of the parallel plex, which has been there for years. Yeah.

But one, what, what is, um, what is the challenge is, is to deliver a speed on the mainframe because, because of the reliability it does, it does to be a bit of a shrine with the way that cha the way that change happens on it. So if you think about, think about the modern enterprise, you've got so many different platforms and they're all interlinked, I think, I think the challenge is for organizations is that the mainframe delivery can be slower and more expensive than other areas of the organization. And, and that is, uh, a big challenge for CIOs because they want the reliability and, and the processing, particularly as AI is, you know, is, is is very rapidly coming on us.

Um, but you know, things still have to happen at pace. Um, and you can't afford to, for projects to take months and months, it's gonna be done quickly with quality. There's a continuing reliability, and it's not just amongst the biggest firms in finance.

There's a continuing reliability in government, actually, there a number of sectors on the mainframe. But what you're highlighting is that because of the nature of the mainframe, the benefits you get have come with a fair number of controls and operational issues that create those challenges of, you know, moving fast. Is that, is that a fair characterization, at least, you know, up to this point?

Yeah, I mean, I think, I think you, you summed it up well, uh, I mean, obviously, uh, the organization wants to have its cake and eat it, right? It wants to have that reliability and speed, but why can't change be done quicker and, and respond immediately to market demands? Why, why shouldn't, uh, a mainframe, why can't, why can't you make change in a mainframe?

Like you can make, um, in the web or applications? You should, you should better do it if the process is there. And this is where the mainframe to me, needs to mo to modernize.

It needs to modernize more around the process and how it's used just as much as the technology needs to advance Well, being able to spin up the equivalent of a mainframe in any X 86 compatible cloud service providers environment, which is what you described at the beginning. That sounds like a promising way to address that challenge. But, but, but first, um, I know you just, uh, conducted a market study on this.

Um, so why don't you tell us about that? 'cause that seemed to provide some interesting insights into this whole mainframe question. Yeah.

Yes, it was, it, it, it was, um, very insightful and obviously we don't need to tell you guys, but the most important opinion about any market is from the market itself. And we wanted to learn more about some of the specific challenges facing mainframe delivery teams today. That survey is just finished and we're just publishing the results.

Now, What was the most interesting data you got back? Um, Paul, point or two. Let's go through the whole survey.

Yeah, Yeah, no, no. That, that, that would, that would take us a while. So funnily enough, what, and what we've talked about is what, what was reassuring was the absolute overwhelming loyalty to the mainframe that the clients had.

And also there's, there's very much a demand to use, um, different processes like the Linux, IFL, and, but there were a number of, of, uh, challenges raised by most respondents. Um, a good example was that 96% said they had challenges developing and testing on the mainframe, and a third of all responders also said they were in critical need of more mainframe environments. You were talking earlier, you said that pop-up mainframe sort of came out of a, uh, a DevOps engagement that you had.

So this is a really interesting contrast to me. I mean, DevOps is certainly about, you know, uh, rapid builds, uh, rapid promotion, continuous development, you know, that sort of thing. What often gets forgotten is the ops side of it.

Um, and operating mainframes is, uh, is, is a thing of its own really important critical element of it. Um, so, so when you're talking about dev and test on a mainframe, and almost all of the survey respondents said that they were, they're, they're challenged in this, and a significant number of them said they're, they're limited by the number of mainframe environments they have. Um, that really seems to play into this idea of cloud-based mainframe, so to speak, being something valuable provided it's operated correctly.

Provided it's maintained correctly. Yeah, Yeah. Uh, you know, just just to elaborate on that, I think traditionally, I guess where you have, you know, a very high performing, um, piece of hardware, that hardware has a cost.

So environments are generally, um, set in stone, and when there's, and when you need another environment, you are tempted to use an existing one. So you start getting an interwoven ness of applications and different projects happening. Yeah.

And what what tends to happen is, is that you find, you, you, you find environmental bugs and things don't work. So if you look at the distributed world, you'll see that typically in with a cloud, for example, if you, if you've got applications running this year, when you want a new one to do something new, you spin it up, do your work, take it down. That's not really a concept that is, that happens in mainframe, but with popup mainframe, we can do that spin up.

So it's about spinning these environments up very quickly, moving the application on there. And we can do that using, taking it outta GI and modern ways of doing that. You then do your work and testing, you've got this agility around the popup mainframe where you can forward and rewind it, copy it to another one.

You get all this agility, so you are working analogous to in a distributed world. Um, but it, it is, it is, uh, using a fully functioning mainframe. And that's, so that's really the power of the solution.

Is it, is it running on IFL at that point? No, so, so we, we originally started with the, what, what, what was the popup mainframe, which ran on X 86 and, but now the X 86 version is known as popup mainframe on X, and now we've got a popup mainframe on Z. That's the announcement, that's the announcement you just made.

Okay. So you have the X and Z, so X is running on X 86, and so is Z running on IFL? Yeah, Z runs on the, the IFL processor on the physical mainframe, and you can also run it on a Linux one box as well, which is a specific, uh, Linux server, and you can run that as well.

And all of this is only development and test. You can only, you can never run production workload on a popup mainframe. So IBM will, will be pleased that I say that out loud to everybody.

Okay. So, so, uh, I feel like we, we zip right past the lead of this interview because a big reason why we're doing this interview is that announcement. Um, so you now have a popup mainframe, which originated as a way to pop up a mainframe environment for dev test, uh, or dev test, let's say on, on X 86 based cloud.

Yeah. As of now, today on, you've launched the Z version, which is actually running on IFL, if not Linux one. That's cool.

So you have to, in my mind, my, my, my mind is, I'm always thinking like lifecycle and in this case would mentioned DevOps. Many times this is a way of, of these ways exist, but this is like a more agile way of using the DevOps workflows or mainframe development along with your other application development, right? That's how I'm looking at it.

So I'm seeing this nice progression. There's sandboxing, there's dev test, there's these ways to do mainframe development quickly prior to production use, which would be on, you know, a physical mainframe. Yeah.

Is that correct? Yeah. So I think we've, we gave you x we've given you Z, but the thing we missed out is the Y, um, You mean YWHY, don't you?

Yes, yes, yes. Okay. That's a good one.

That's I Go ahead. So Y yes, YI just gave YY but maybe I'm wrong. What's the why?

So the reason why we, we also moved to z to, to, sorry, I, I slipped then z um, was because the thing about a popup mainframe, it uses hardware emulation. So it's, you are running exactly the same code base, um, on a popup mainframe. Then you, as you are on a, on a real physical mainframe, okay, you can match it to the, in IBM terms, it's called A PTF, which is exact fix to a particular subsystem or what have you.

Um, but some of the community and clients do not want any of their mainframe out running on X 86. So the, the mainframe is the mainframe, and in some organizations, no data should ever leave the mainframe. So for whatever reason, so they would not, they would not buy a pop-up mainframe on X.

So we bought out, um, popup mainframe on Z is to emulate the physical mainframe, but also on the mainframe. So you get, and, and, and here the client can have, can have both. Their mainframe still stays on the mainframe, but they get all the benefits of the virtual environments.

They can, um, take a mainframe, save it to disk, bring it up later, um, put another copy into another, onto another processor for whatever reason. Um, and so, and there's also a scenario where clients can have both X and Z. So you could, you could build up your goal copy on your IFL, and then you could send copies of that to a hyperscaler cloud.

So you could have almost like lpar, fully functional mult multiple users on these large tests, on these powerful test environments, but you could spin up copies for other teams doing smaller bits of work or in feature teams in the cloud. So you could have you, and you can really benefit from these lower cost, um, dev test environments. So, so what would you say to people who, uh, who are really, like, been doing this sort of work for years just just on the hardware, right?

What would you say to them if, if they express concern to you about security or, or, or, you know, uh, um, management, uh, data protection compliance? Yeah, like that sort of thing. Because I think that although, you know, CIOs certainly and CTOs and IT shops have embraced the cloud, especially in sort of hybrid form as the capabilities have gotten better, I think, um, and that includes the mainstream folks.

I think, um, there's a certain, um, there's a certain, uh, uh, guarded nature about using the cloud for these, these workloads even in, um, pre-production environments mostly 'cause of, of, of data control. So what would you say to that? Yeah, so it's, it's an interesting question and, and a lot of it is perception.

Um, I think with, with, with the cloud. So you know, you, when you look at a pipe popup mainframe, you're securing it a number of wa of ways you are securing it. You can secure it exactly the same as your physical mainframe with, with the same tooling and the underlying security as, as well as, you know, infrastructure, the firewalls and what have you.

So you could, you could, you can actually secure it at two levels. You can secure it around the cloud security and Linux security, and then you're securing the mainframe as well, running on Linux. So you can get multiple levels, uh, of security on that and probably make it even more secure than, than say, a physical mainframe, because you, you've got, as I said, multiple levels of security, one on top of each other.

So, you know, I, again, I think quite, quite often there is decisions around security are quite often that, that are patterns more so than what does the organization need? And okay, we've got a way of doing something different. What is the best way of securing it?

So you can look at it differently. Um, but you know, there are some organizations that just say, you know, that's not secure having a, a, a popup running on, on the cloud for, for whatever reason. And, and quite often it's not, it's not being explored.

It's just a, a blanket statement. Uh, and I Think, I still think just Gary, I still think there's a fair amount of, uh, misunderstanding of, of security and the cloud. Uh, yes, there, there, there have been issues and, and I, I wouldn't want to name names here, but certain of the biggest cloud providers have, um, maybe enabled capabilities and services, but not forced them or required them.

So you see continued breaches. But what I'm leading up to is the fact that that truly some of the best security systems in the world, and again, I'm not gonna name names, have been implemented, are available and in place in these cloud environments. Yeah.

And, and with the expertise certainly and and the right configuration work, um, you, you know, you, you are quite able to achieve better security than, than, than in most cases you do natively in your own shop 'cause of what's available. And I don't think that's fully understood even today. It sounds like, um, you might be taking advantage of a lot of those capabilities.

Is that, is that true? Yeah, I mean, uh, you're exactly right. So, you know, we, we, what what we can say to clients is, is i, i, is that you can, um, you can run your popup in a variety of different ways.

You can pretty much secure it exactly the way that you secure your physical mainframe today, the exact security controls. But, you know, would be that, would that be the right thing to do? Um, and you can run pop up mainframes, depending, so what, you know, the, the first thing that we always do is make sure that the data is, is compliant.

So we have lots of masking strategies just about user data itself, and then you're looking at access on the popup mainframe, but you can, you can use all the modern software and approaches for secur securing on the mainframe, and then you can put a cloud S-S-L-V-P-N layer, you know, and you can just keep on adding to this. And of course, always the weakest point of any security, uh, is someone's own access itself. And that's not going away, is it?

Yeah. So, but we always, when, when, when we start working with clients, we always try and understand what the, what they, you know, what are they doing today, and we can help them improve that, um, nine times out of 10, um, by adding in these different layers. Um, so, so, so that the way that they use it is, is secure.

Um, but um, it's, it's having, it's involving the right experts at all levels, um, to meet, to meet the requirements. Well, yeah, and you're, you're addressing that question, um, with your, with your product and with your, your announcements of today. So that's this nice, um, dual option of the X and the, the Z additions.

Um, one more emulation, the other one, you know, direct IFL or, or, or LinuxONE. Um, both have been for, uh, a quicker, more responsive way to work towards the full production environment that's gonna be on your mainframe or mainframes. Um, I think everybody should check that out.

Um, who, um, is interested in, um, and working on mainframes today? Um, the, the study, I actually, if I remember right, the survey that you did, um, uh, we might be discussing in more depth on a webinar coming up, um, also a Textron uh, in, um, in, in a month or so, I think, um, June 18th. Uh, so, um, I would encourage everyone also to look out for that.

Um, uh, Gary, is there, is there any other notes and bits you might wanna mention about the announcement today or about that study to, you know, keep everybody, uh, excited and interested? The webinar will show that the, the mainframe is a fanta fantastic production engine. So whatever apps you're building, whatever Xeros box you're on, whatever data you're using, I imagine there are some bottlenecks a day you wish weren't there.

And we can show technicians, PMs, and change managers how mainframe dev test can be approached differently. We're here to discuss those issues and present an approach which we think will be a game changer for mainframe delivery teams. But in order, in order to meet with the times you've got these new technologies, but it's about how we improve delivering change on the mainframe.

Um, there's a continuous need and desire to modernize not just what the business process the mainframe runs, how quickly they can be built and, and, and you can get both benefits, um, from X and z. I mean, certainly with, with z what is interesting, so that, that and just of your previous point, they're both, they're both emulation, but um, on x, on X 86, um, you don't get the full benefits of the mainframe processor. So you are emulating the way the CPU and the disc io, whereas with z it, the, those Native, those are direct Processor, yes, it goes direct native, um, but the IO is emulated, so it, it does work slightly differently, but it's more closer to, uh, the real physical mainframe and, and performance is significantly, um, improved.

Um, but they both help clients align with how, you know, how, how you make change in the distributed world and splitting things up and down on demand. And it also has a very important using pop-up mainframe, um, massively, um, IM improves your, the sustainability side of things because Yeah, I was just thinking about efficiency when you were talking about that. Yeah, go on.

Yeah, so I mean, it's one of, one of our clients literally save half of its zero costs and, and reduced its carbon footprint by half by just switching off for 12 hours a day when they weren't being used. So that's not a concept you have on a physical mainframe and there's always things running not normal. Yeah.

But you can just, um, with a popup mainframe, two things. First of all, you can switch it off when you're not using it. Um, so it doesn't sit there and just, just, um, just using electricity.

And the second thing is, is that because we have this unique forward and rewind capability, which we'll touch upon is that, um, you can take a checkpoint at any point in time, um, and come back to that. So just the efficiency gain you get from not having lots of bodies, um, making changes and clearing down logs and resetting tables, DB two tables or other things, instead, one person can go into, uh, a gui, take a rewind, and then you're ready to repeat a cycle of testing. Now that is a huge efficiency gain.

Um, and mainframe is famously, uh, renowned for being the most sustainable platform anyway. 'cause you've got so many processes running on one piece of tin add, popup mainframe to the equation where you can reduce the amount of head count required to do, to do work and at speed has a huge environmental, um, saving for it. And I think that's very important, particularly the advent of ai, which is, um, gobbling up so many resources.

Um, we've gotta always have a, have an eye on, you know, how, how we can be sustainable by the use of it. Well, I look forward to hearing more about that then. Hopefully I'll be on that webinar myself.

This is a fascinating topic and definitely an area of interest and pursuit for many kinds of enterprise that around the world running some of the most important applications, uh, being run right now globally. Uh, so Gary, thank you for joining us. Um, thank you for joining me.

Um, thank you for talking about popup mainframes and what's going on in the mainframe market, uh, and mainframe development right now. I'm Guy Currier analyst with the Future Group, and it's been my pleasure to present this to you and I'll see you next time.