Len Santalucia: I am a Mainframer

By April 28, 2017 March 29th, 2018 Blog, I Am A Mainframer

Screen Shot 2017-01-17 at 1.49.41 PM

In our fourth episode of our “I Am A Mainframer” Interview Series,  Jeffrey Frey talks with Len Santalucia, CTO and Business Development Manager for Vicom Infinity, about the value proposition the mainframe platform has built around the economy of scale, and the power of Linux on the mainframe and its open environment.

You can listen the full recording and read the transcript of the interview below.

If you’re a mainframe enthusiast or interested in the space, we invite you to check out our new community forum.

Create a profile and post a selfie with your mainframe system, and you will receive an exclusive “I Am A Mainframer” t-shirt.

Jeffrey Frey: Hello. This is Jeff Frey. I’m a retired IBM fellow and prior CTO of IBM’s mainframe platform and very much a mainframe enthusiast. Today,  I’m very pleased to host on the “I Am A Mainframer” interview series, a conversation with my good friend Len Santalucia from Vicom Infinity. This conversation is sponsored by the Open Mainframe Project. As a Linux Foundation project, the Open Mainframe Project is intended to create a mainframe focused, open sourced technical community. It’s also intended to serve as a focal point to the deployment and the use of enterprise Linux in a mainframe compute environment.

The goal of the project is to excite the Linux community around the use of the mainframe and to foster collaboration across the mainframe to develop and exploit shared Linux tool sets, resources, and services in a mainframe environment. In addition, the project seeks to involve the participation of academic institutions to help and assist creating educational programs aimed at developing the mainframe Linux engineers and developers of tomorrow.

Len, it’s such a pleasure to talk to you, my friend, today. I know we share the same passion for Linux on the mainframe. Len, tell us a little bit about Vicom Infinity and the work you do.

VIACOM INFINITY AND THE MAINFRAME

Len Santalucia: My name is Len Santalucia and I am the CTO and business development manager for a IBM premier business partner by the name of Vicom Infinity. We are based out of New York City, and our work is all around the idea of mainframe. We are across all of its operating systems:  z/OS, Linux on Z, z/VM, KVM, z/VSC are the major areas where we focus. We also deal with all of the hardware, z Systems hardware as well as with the associated disk and tape and IBM software and open systems, open-source software. We have been in business since the ’90s. We are considered in the top five IBM premier partners in the world. I was part of the startup of this group, Jeff, way back when it first was starting because I was still employed at IBM at the time. Like you, I retired after 30 plus years from IBM in 2008, and this company asked me to come on board to help them do the job I’m doing for them today. But back in the ’90s when it was just a one or two man operation, I was asked to help support them in the greater New York area. They didn’t forget some of the things I did for them and when it came time to looking for something to do after retirement, they offered me the position. It’s been great because it’s working with a different tier of customers. IBM handles the very large customers. This group handles the next tier below, although we do have some very large customers that normally would fall into the tier level one of type customers. They’re very large, but because of our proximity to New York, it just happens to be that way. That’s a little bit about the company. It’s been doing very well and really focusing a lot around Linux and open-sourced software quite a bit. Still today, even carrying forth my passion like you, Jeff, from our days at IBM.

MAINFRAME AND LINUX DEPLOYMENT

Jeffrey Frey: Yeah. That’s great, Len. I knew some of that history and I wasn’t aware, I don’t think, we knew each other. I can’t remember if we knew each other when you were at IBM. I think we may have met after that. I can’t quite remember, but it’s been a while.  

Why don’t you tell us a little bit about some of the projects you worked on involving some of the use of the mainframe. You said, I know especially in recent times, you’ve been very involved in the open side of this and really very interested and involved in the Linux deployment and use of Linux on the mainframe. Can you tell us some of the projects you’ve been working on?

Len Santalucia: Yes, I can. Back in the 1999, 2000 time frame Linux was brand new to IBM and it was an emerging business. I happened to be fortunate enough to been involved with it back when it was first starting out and watching it evolve into what it has been today especially on the mainframe side of the business, has been so great to experience. First, starting from server consolidation and print serving and web serving, to today, where it’s now become a great platform for Blockchain because of its integrated cryptography and integrated vector processing capabilities, for numerically intensive as well as transactional cryptographically enabled transactions. That’s a mouth full.

And with all the new things to the platform such as with Docker and Chef, these new technologies emerging have been just a great natural fit to the mainframe Linux environment. Now,  Ubuntu from Canonical becoming available, to see that evolve into what it has, Jeff, and really placing it right in the middle of being a great platform for cloud and analytics with Spark. This stuff is just absolutely amazing. It’s never, ever been a better time to be enabling these kinds of workloads on the mainframe in its history. It’s just phenomenally architected platform compared to anything else that’s available.

Jeffrey Frey: Yeah. Let’s dig into that just a little bit, Len. You and I have been working on the evolution of Linux on the mainframe for quite some time. The value proposition of Linux on the mainframe, has always been to take the open standard and open environment, the open software and that whole ecosystem that Linux represents and all the skills that come with it and marry that to arguably the most secure, robust, reliable piece of hardware in the industry. I think your point about all of this stuff coming together and kind of coming of age together, I think, not only has dramatic improvements in the mainframe’s ability to host all of this Linux-based software and the open environment, but I think Linux has grown up too over time to be recognized now as an enterprise class platform. It’s kind of grown up in search of a platform that can deliver the right security and the right availability for enterprise computing. I think you may be right. Maybe the time has really come now where the marriage of these two notions is really come of age.

THE VALUE PROPOSITION OF THE MAINFRAME

Len Santalucia: I totally agree with you, Jeff. Again going back to that early stages, where it was amazing to see a Linux operating system maybe handle two to four cores at a time and be able to manage them. And now they can run on hundreds of cores on the same processor and actually exploit its virtualization technology,  if it’s either KVM open-like virtualization or the virtualization on the mainframe known as z/VM, which really exploits the architecture because they’re both z/VM and the hardware are code developed and they’re special firmware or microcode as it’s called in the mainframe world developed for this virtualization technology. It just allows the Z systems mainframe to run literally at 100% utilization, at all times if you let it, and be able to handle so many diverse, concurrent workloads simultaneously that you could never fathom doing that on any other architectures. You’d always have to leave some kind of head room. You can never run it at full utilization like that.

And you know this even better than I do, Jeff, having been the CTO of the actual hardware development. You know that system has so many level of cache and it can keep feeding that processor, it can run at that full 100% because of this design. There’s just nothing else like it. You know what I’m saying? It’s just amazing. Docker, another one. Docker can run several hundreds. And they’ve actually documented it now to the point where having one million Docker containers on a single mainframe, single Z mainframe. Where are you going to find that any place else?  Nowhere!

Jeffrey Frey: Yeah. People can look at that on the surface and say, “Okay, that’s pretty impressive. What’s the value in that? Why can’t I just have multiple machines? Why can’t I have just a kind of huge scale out kind of environment?” There’s always been kind of a central value proposition, the mainframe built around economy of scale. And it’s so good at sharing that the efficiency that you can get, which ultimately lowers complexity and lowers cost of the overall IT environment it’s just superior and second to none. When you have an architecture that is tailor-made, its purpose in life is to be able to run lots of work concurrently and share those resources very efficiently, then you couple that with historic superior levels of virtualization built right into the hardware, there’s another aspect of why it’s just so well suited for a Linux environment, so well suited for Linux and a cloud environment and a highly virtualized environment. I’m really hoping that people really start to really appreciate the value proposition of the platform and the open environment.

Len Santalucia: I totally agree. There are tools to help customers actually see the value of it, specially for their environment, that enable folks like myself to gather statistics from where they might running it today and showing what value it can bring in doing it on the z/VM. Just recently, Jeff, for one of the financial customers on Wall Street. Known them from days I was a kid, I’d see on the street, supporting them. They asked us to look at their database environment and what it can mean in saving … because they’re really running out of floor space, power, and cooling and that kind of stuff’s pretty scarce on the streets of Manhattan and the surrounding community. It’s very costly. When they heard, after doing this analysis that they could actually consolidate 39 to one. What do I mean by that? 39 cores of what they’re running today in their x86 environment down to one core at mainframe, which we affectionately call an integrated facility for Linux, IFL, they were just totally blown away. It helped them further consolidate where they were on x86 and they found some errors of their ways. But still even after they did some of that consolidation, they still could get to thirty-ish, high twenty-ish cores per one IFL core on the mainframe. It still was a very great story. And then everything else falls into place. The power, the cooling, the floor space, the software licenses, the networking that’s eliminated because it’s all communicating in the same system.

One of the largest police forces in the nation is one of our customers, and they consolidated. It wasn’t the consolidations of the cores that made the difference as much as when it got rid of the network, because the network could be inside the mainframe virtually, they were able to eliminate all those connection points and routers that go along with that. All areas for failure, failure points. And it just made their arrest system and their court system so much more effective and efficient. It’s always available.

Jeffrey Frey: Yeah, not only points of failure but points vulnerability in terms of security systems.

Len Santalucia: Absolutely.

Jeffrey Frey: There’s just all kinds of advantage to that. You said 39 to 1. I don’t want the listeners to get the impression that, to leave with us claiming that these mainframe stores are 39 times more powerful than an Intel core. It’s really about the overall system design.

Len Santalucia: Absolutely.

Jeffrey Frey: The reason you can keep those level of consolidation is not because necessarily the cores are that much faster, but because we do such a good job at sharing resources and caching in doing memory management, and managing working sets and paging. The IO system of a system z and how parallelized that IO structure is,  and how offloaded the IO is off of the central processing unit. It’s all about the overall system design of a mainframe, It’s just built to do that kind of work. It’s almost as if the more work you put on it, the better it runs. That’s why people are surprised when you find that you can do this kind of consolidation and get these kinds of efficiencies.

Len Santalucia: That’s true. A claim like 39 to 1 is definitely not the norm. But it’s not too far fetched to see 20 and 25 to 1 on the average. That’s why we always perform an analysis that really works with the exact numbers that the client gives us so we can definitely tell them specific to them. The reason why I think it was so different with this large number here, with 39 to 1, was they had designed a system where they because of the availability issues that they had many servers were just remaining on idle just to be there just in case they were needed. When you have that kind of idle time, it can inflate the number to something like 39 to 1. But when they did some of that homework, like I was talking to you about before, where they readjusted things and found out they could maybe not have so many idles, it still turned out to be in the high twenties to one no matter what they tried to save, or correct the error of their ways, as I mentioned.

Nothing beats a good study to do the homework and make sure the real numbers are looked at. But, look at the numbers about I just mentioned a moment ago with Docker images, a million on one machine. You can’t argue about that. And that was no consolidation. That’s just really pumping them up. They really were running them. When you couple that with the kind of disk storage you can have on the mainframe and it’s IO capability, the separate IO sub-system just brings so much to the table when you do things like database serving compared to the other platforms that don’t have that kind of underlying subsystem for IO. They have to do IO and the application all on the same processor. The mainframe has another value add: its multiple sub-systems for IO and applications and math, all different kinds of functions.

Jeffrey Frey: Yeah. I said that when I would talk to customers about the mainframe and tell them that there were literally hundreds more processors in the box than they had imagined because there are hundreds of processors dedicated just to doing IO.

Len Santalucia: Yes, yes.

GOVERNING THE OPEN MAINFRAME PROJECT

Jeffrey Frey: Let me move on here, Len. I know this has been great. I’m fearful that we might be running out of time. You serve as the chairperson as the Open Mainframe Project governing board. Is that correct?

Len Santalucia: Yes. That’s correct. I’m going on the second year of doing that and it’s been very exciting to … When I was selected and voted in, I was very nicely surprised that I had that kind of support from the team. I really enjoyed doing it because I think that the Open Mainframe Project is a very powerful organization, especially since it is run by the Linux Foundation which is the home of the inventor of Linux itself, Linus, if you think about that.

There are many times I tell customers, you get a person maybe like me, who has all this passion. I come in and I talk to them and to their groups outside the mainframe in their organization. You can tell that some people are looking at you like they don’t believe you or they’re testing you or what have you. But if you bring an organization like the Open Mainframe Project into the same discussion, things change. Jeff, you and I both have kids. You know sometimes when you talk to your kids, sometimes you can talk to them, pardon the pun, ‘til you’re blue in the face and they don’t listen. But maybe you have your brother or your sister or grandfather talk to them. They say the exact same thing, but for some reason or another, they listen to them.

It’s the old saying, “It takes a village.” Right? If you leverage something like the Open Mainframe Project, which is based on the Linux Foundation premise, and they’re saying the same thing as you do, it’s really pretty magical what I see happening. It’s really just really furthers the cause of Linux and open sourced software better than anything else I’ve seen, especially at this stage of the game, of how mature Linux is on the platform and open source on the platform. But now the remaining people who are, that dug their heels in and have resisted up until now, they’re the toughest ones. And if you leverage something like this Open Mainframe Project to come in and talk on behalf of yourself or anybody else, they’re going to carry much more weight and credibility than anything else that’s been done before.

And what’s great is the Linux Foundation people are very nice about helping out all the way up to the first lieutenant’s reporting right into Linus, himself. They’ve been very helpful. We should let people know. These are the kind of things, I think, we as members of the Linux Foundation Open Mainframe Project need to convey to everyone or how to leverage it to their advantage.

FUTURE CHALLENGES WITH THE MAINFRAME

Jeffrey Frey: Great, great. Len, let me close with a little bit of a provocative question for you. What do you think are the biggest challenges for the mainframe going forward? You and I, we’ve had history with this thing. If we look into the future, what challenges remain and what do you think the biggest challenges are?

Len Santalucia: It has nothing to do with the technology because if people look at the technology and understand it, they’re going to say, “Wow, this thing’s pretty powerful.” But then they’re going to say on the other side of their mouth, especially those that are the naysayers, “Well, but it’s still a mainframe.” People equate the mainframe as a more costly platform, which is just not true when you take into the total cost of operation. And the other thing that it has a challenge with, especially today, is you and I use our cell phones, our iPhones, our droids, all of our smartphones and tablets to access just about anything and everything.

The people that are coming up through the ranks today, don’t even get a chance to see the mainframe because, as I tell students when I’m talking through the IBM Academic Initiative to a lot of these colleges these days, the mainframe is invisible to you. That’s the word I say. It’s invisible. But when I tell them, I ask them very simple questions and it hits them right between the eyes. “Did any of you go to an ATM this morning or this afternoon,” if I’m there in the morning or what have you at the school. They say, “Oh yeah.” I said, “What was the name of the bank?” Some of them say Wells Fargo, some of them say Citi, some of them say other, JP Morgan, Chase. I say, “Well, you just used a mainframe.” “No,” is their comment. And I go, “Yes, you did and you didn’t even know it. Did any of you make a reservation for a trip to go home and see your family on spring break? What airline did you use?” Every one of them named airlines that has mainframe behind the scenes. They didn’t realize that. Insurance claims. “Did any of you have a car accident? Do you use Geico? They have a mainframe behind the scenes.”

Once they started getting that and then they start understanding the career potential is and the challenges are compared to just writing these apps that you touch and it does a little something on your phone,  And they say, “Wow, this could be a heck of a career.” And the kind of money that they hear that they could make. The long answer to a very simple question is the challenges are making it, making it visible as opposed to invisible, and showing the true value beyond the myths. And they’re still believing. I got people that think about the mainframe as it was pictured in some of the old time movies, because they haven’t seen it anywhere else. Those are really some of the key things. Otherwise, there isn’t anything that it can’t run. Anything that runs on Linux can run on it., and then the ISVs that are dragging their heels that are not making their software yet available on Linux on the mainframe as opposed to their Linux on x86 because they don’t know better or they don’t have access to it. Now, IBM’s done a great job there by making a Linux One Cloud available to anybody in the world that wants to try and or develop on it at no charge. A lot of things are being done a lot differently.

Back when we were first starting out, you never would have gotten access to a mainframe for free. You know that. Now look what you can do. All these things, it’s going to change, and it’s going to change for the better. You have to have patience and persistence and perseverance. And it’s already very obvious how great it has advanced. Just these last couple years with the Linux Foundation doing its small part. Hope I answered you question. I know it was a long answer.

Jeffrey Frey: No, yeah. No, you covered a lot there. I know that perception has been, historically, kind of a problem here. The invisibility is a good way to put that. Access has been a problem in the past. As you point out, a lot of these problems are being worked and a lot of these obstacles are being removed. I think the other thing is, you mentioned that if people see opportunity in working in this environment because the companies value what the mainframe brings. There’s this tension between high-volume and premier quality service and value. I think that the more it’s recognized that the mainframe is basically the IT backbone of a lot of the economies of the world and it’s dependent upon to run the world in terms of its commerce, that people might, there’s value in that. And if you have skills in that and if you can contribute to that, it’s got to be worthy of reward in terms of your career, in terms of the opportunity that it provides.

HOW ESSENTIAL IS THE MAINFRAME?

Len Santalucia: Put it in a very simple way. Turn off Google today, the world would keep on going. People would be mad that they have to go to a library again instead of going to a terminal or their iPhone to look something up, but they still can find the information someway, somehow.

You turn off the mainframes, the world will come to a complete screeching halt. It would be utter chaos. People couldn’t access their bank accounts. Planes would come crashing into the ground. It’d just be a disaster. The world would come to almost a complete end. That’s how everybody depends on mainframe. The analogy that people should think about very closely.

Jeffrey Frey : Yeah. Well, Len, my friend, it’s been great talking with you. I think we’ve covered what we needed to cover. I really appreciate you being on the call. I’ll look forward to seeing you in one of our upcoming meetings.

Len Santalucia: Yeah, same here, Jeff. And special thanks to our sponsor, the Linux Foundation and Serena especially for getting us together here today. Very much appreciate it. I always enjoy this conversation. I know we had a short period of time, but I could talk about this for a long time. It’s my passion and hopefully it passes on to other. I know it’s your passion, Jeff. You passed that passion on to me back in your days at IBM, so it was very catchy.

Jeffrey Frey: Well, thank you, Len. It’s been my pleasure. I’ll see you soon, my friend.

Len Santalucia: Yes, take care, Jeff. We’ll see you at the next meeting. Take care. Bye everybody.

Jeffrey Frey: Bye bye.