I am a Mainframer: Eric Chevalier

By June 23, 2020Blog, I Am A Mainframer

In today’s episode of the “I Am A Mainframer” podcast, Steven Dickens sits down with Eric Chevalier of Phoenix Software. On this podcast, Eric discusses his journey with the mainframe, bringing Linux onto the platform, advice to young people who are just beginning their journey, and where he sees the Mainframe going in the future.

Listen Now!

Steven Dickens: Hello and welcome. My name’s Steven Dickens and you’ve joined us on the I Am A Mainframer podcast from the Linux Foundation. The Open Mainframe Project is a collaborative project under the Linux Foundation collaborative project structure, and the Open Mainframe Project is designed to advocate for Linux and open source on the mainframe. Today, we’ve got Eric Chevalier from Phoenix Software, who’s joining us on the podcast. Hello and welcome, Eric.

Eric Chevalier: Hello. It’s great to be here. I’m honored.

Steven Dickens: The honor’s all mine. The honor’s all mine. So thank you for joining us today. I was really interested by your bio when you sent it over. So I think this is going to be a good podcast for the listeners. But one of the questions I sort of use to get us started, just give us a little bit of background. Give us an introduction, if you would. Tell us a little bit about kind of where you’ve come from and your role at Phoenix Software.

Eric Chevalier: Okay. Well, I’ve been with Phoenix Software for 31 years now, which is really 31 out of about 43 years of experience in the computer business. It’s been a good company to work for. I’m happy to be here. And it’s given me the chance to work on a lot of different projects, on a lot of different platforms. I came in originally to do some mainframe programming and to be sort of a liaison between the mainframe side of our company and PC development, which was just kind of getting started at the time.

And my boss, who’s the owner of the company, felt it would be useful to have someone who had feet in both areas. My original experience was definitely IBM mainframes, but I had sort of bought my first IBM PC back when it first came out. And that got me interested in the alternative platforms, the little systems. So my boss, Fred, felt that I could be useful in being the bridge between the two areas at Phoenix.

Steven Dickens: That’s interesting. A lot of my background is the same. I spent 10 years at HPE settling sort of distributed systems for them. And then I’ve come to, obviously onto the mainframe platform the last 10 years. So, it’s good, as you say, to see that sort of foot in both camps. Maybe that’s one area to go next to then. I mean, you mentioned sort of having that foot in both camps. I see from the bio you sent over, you’ve done some work in Linux. And from what I can tell back from the early days on the Marist distribution, so maybe just give our listeners an insight into your track record and what you’ve been doing on platform over those years.

Eric Chevalier: Well, it’s funny you mentioned the Marist Linux because that was the first distribution we started with on our system. It really was… I had been familiar with Linux by that time because of my work with it on a PC, I love Linux. I kind of really liked the Unix environment. A lot of the tools that are available. And we had been kicking around the idea of why don’t we see what Linux on the mainframe is like and what it can do for us. It was originally a proof of concept. We didn’t have a lot of big-budget to spend on stuff, but we were able to carve out some space on our disks for VM and the Marist distribution, Debian at the time, as I recall was free.

So that was really pleasing to the boss. And it was a successful proof of concept. We got it up and running. And it really was a nice little Linux environment. I think the only problem we had at the time was we didn’t have any of these specialty engines on our processor. And so the Linux was running on the general-purpose processors, and it was consuming, not a terrible amount, but a noticeable amount of CPU.

Steven Dickens: Yeah, it’s interesting. And we’ve made that obvious transition to those specialty processors with the integrated facility for Linux. And now sort of standalone machines that only run Linux with the Linux one branded. But it’s always interesting to chat with the guys, and I have the pleasure of working with a few of them out of the labs in Albany, and who were involved back with the Marist distribution. And we love our friends from Marist who is part of the project. Harry and the team over there, they do such a great job. So I always find it fascinating to listen to people who were kind of breaking new grand and bringing Linux onto the platform back in ’99, 2000. It’s a fascinating period in just the history of what we’ve seen on the evolution of the platform.

Eric Chevalier: I don’t think we have any specific applications in mind when we put that system up. It’s been about 20 years since we put up that Marist. And I think one of the first applications was an instant messaging server. And over time it got to a point where we realized there were other things we could be doing on the Linux. So we upgraded. We got an IFL. We decided to go with Red Hat. We had been using that on our Intel based system, which was our public web server and email server. It seemed to work fine. So we decided to go with them to the mainframe.

We wound up putting more and more applications on our systems. We now have two dedicated… Dedicated in the sense that they are VMs in our environment. Z Linux Three, Z Linux Four. One of them runs our internal primary name server. We still have that instant messaging server. We have an IBM spectrum protect server running on one of the Linux boxes. We have a problem tracking system. The Linux boxes have gone from being a proof of concept in our environment to absolutely essential systems that we couldn’t run without.

Steven Dickens: That evolution is what we see in a lot of our clients. I have the pleasure of chatting to clients every day who are on this journey. And a lot of them have got that same path that you’ve gone. They start with the POC, they get comfortable with the platform, and then they’re starting to bring more and more workloads on, just because it makes sense for them. So, I mean, pivoting there a little bit. You talked about being on the pioneering front end of what we were doing with Linux on the platform. And then I look, you’re doing the same thing again with open source. And Zowe on z/OS. So kind of give me that same perspective, sort of 20 years later, going back to a new technology and a new sort of movement on the platform. Do you see the same parallels? Do you see the same sort of pioneering spirit? What’s the experience?

Eric Chevalier: We do. But, it’s interesting, that first Linux was just proof of concept. We just wanted to see if we can run Linux on there. With Zowe, we think that’s going to be an essential complement to our existing products. So we’re starting out with the position that this is not just a proof of concept. This is real systems we’re going to be developing to enhance our (E)JES product specifically. So we’re taking it seriously. I was kind of the Pathfinder on the project. I installed the first version. I think we started with 1.0 or 1.1. We’re now up to 1.10. We just recently started or are going to start using the IBM distribution of Zowe. It has some advantages in that it can be maintained as part of our mainframes standard support system.

Steven Dickens: So tell me a little bit about what Phoenix is doing with related to Zowe. I think that’ll be interesting for the listeners. We’re seeing, and I always find it fascinating to see there’s the open source community around Zowe just churning along, and moving, and trying to get better and better versions out. And then it’s interesting to see the adoption and what people are doing with that, and how it’s going to live and breathe within their own products, and within customer environments. So maybe just give me the link, if you don’t mind, between Zowe and Phoenix. And you mentioned, I think your (E)JES product. Kind of just elaborate a little for the listeners.

Eric Chevalier: Yeah. Our (E)JES product, it started out as a competitor to SDSF for the JES3 environment. And it was developed about 20 years ago, close to 30 years ago. It was something that my colleague, Ed, was working on when I joined Phoenix. And it’s an interface to the spooler system. We have quite a customer base for it. But Zowe looked like an excellent way to be able to provide our (E)JES facilities, not through a green screen, 32-70 terminal, but through web browsers and alternative techniques.

So we have developed a rest API that can be used to interface to our (E)JES system. And in conjunction with that, we’re also developing Zowe CLIs that will be able to access those rest APIs. We have, for instance, a CLI that will let you stream the SYSLOG or the OPERLOG down to your PC. And we figured there are opportunities for being able to do scripting types of activities by people who are not necessarily really mainframe knowledgeable. We give them a tool that lets them use the existing tools on the PC or the Unix environment that they’re familiar with to be able to access mainframe components.

Steven Dickens: So you’re taking Zowe to kind of provide that graphical user interface and that API driven sort of front end, and get that modernized and transform the experience. So you don’t have to do that in the core of the product and take on that? So it’s kind of an abstracted kind of layer on top of the (E)JES product? Is that the right way to think of it?

Eric Chevalier: I think that’s a way of saying it. Yeah. The core product that runs on the mainframe is always going to be the core product and has the primary functionality. But if we can present that, and make that functionality available in user-friendly ways, ways that are user-friendly to the non-mainframe people that are starting to become involved with mainframes, we think that’s a good move.

Steven Dickens: And that’s a fantastic use case, I think. And this is why I was keen to pull on it, to understand. Because I think the more time I’ve spent in the open source community, it’s that innovate where it makes sense and then collaborate where there’s just the opportunity to go and get composable value that you don’t have to go and do the development work. And I think that’s an interesting piece. You guys can focus on the (E)JES kind of product and make that as a fantastic experience for your clients as possible. And then just get the benefit of Zowe that’s happening as a community project.

And you can invest and get involved in Zowe, but you don’t have to take all the work on yourself. I think that’s really interesting, kind of what’s the commercial adoption. You’re selling a product with (E)JES, and there’s an open source community going on. How do those co-exist? I always find that fascinating to see how companies are exploiting that.

Eric Chevalier: We think it’s going to work out well for the (E)JES product.

Steven Dickens: And how are you finding that community experience? What’s been the reaction? I mean, it’s been fascinating for me seeing the Zowe crowd kind of find their feet, get the product out of the door, build a community. You sound like you’ve been in it for a while. How have you seen that growth?

Eric Chevalier: It’s been interesting and there’ve been other people… I’m not the only one in the company who’s involved in the Zowe project. And it’s interesting hearing their experiences, as well as those of my own. My impression is that the Zowe people are really dedicated to their project. They really want to see it succeed. They’re extremely helpful. We’ve had to call on them through their Slack channels on more than one occasion to help get a resolution to problems we’ve encountered. It has been interesting that many of them are really knowledgeable in the areas of open source, but they’re not necessarily that familiar with some of the mainframe cultural issues that come up.

We had one colleague who was amused when he was talking to some people about JES, and JES2, JES3. And they had pronounced it as Jess. And so, some minor amusement there.

Steven Dickens: You don’t want to get in the middle of JES2 and JES3. I mean, that’s a religious debate right there that… I’m not a z/OS guy, but I know at least that you don’t get into that. That for me is the interesting piece, those two worlds coming together and infusing. I think that’s really fascinating. And that, as a mainframe guy, that’s the most kind of heartwarming part of the Open Mainframe Project. Those two kinds of open source and z/OS kind of coming together, I think. And I see you laughing on the Zoom if people don’t hear this, but I think that for me is the most interesting piece about these two worlds colliding.

Eric Chevalier: Well, I don’t see it as a collision because the two sides are talking to each other. And I think that’s going to be beneficial to the mainframe side. And I think that’s going to be beneficial to the open source side. Instead of the two groups being two distinct, separate, and isolated areas, that they’re talking to each other and working together. I think that’s going to be one of the big benefits of the Zowe project. That even though it’s not necessarily what they maybe had intended, but I think that communication is very helpful.

Steven Dickens: Maybe I’ve seen too many collisions and maybe it’s more now collaboration. It was interesting to see those worlds come together in the first few weeks and months. But I think I kind of feel a fatherly pride having seen that community sort of developing and come together. I was involved before we set it up as a project. And some of those early calls listening to kind of mainframe, z/OS guys, kind of take the baby steps into opensource was… It’s just fantastic a couple of years later where we are with that community. And the number of people that have come into it. How they’ve got a code base out, how they’ve got the conformance project.

The steps they went from sort of the three organizations who kind of gave the first code drops, now into organizations like Phoenix and others kind of coming into that community. It’s been really good to see. So give me maybe a little bit, if you would Eric, on Phoenix. Tell the listeners who maybe aren’t aware of Phoenix, kind of give us a run through what you guys do. We talked about your (E)JES product there briefly. But maybe just give me a sort of whistle-stop tour through what you guys do.

Eric Chevalier: Well, we’ve been an ISV since I guess the late ’70s. Our original product was called Condor. And it was a transaction monitor time-sharing system. Originally I think, developed for the DOS/VS environment. And then poured it over to what was then MVS. Out of that, we developed a product called Falcon, which was a data entry product. And that subsequently spawned a PC version of Falcon, which could talk to the mainframe version, exchange data, and so forth. (E)JES was the next product to come out of our shop. And it was that SDSF for JES3. That was the intention. Although, we are proud of it and we think that it actually has much more functionality than SDSF. For both JES2 and JES3. We’ve had a variety of PC products. In 2000, we acquired a company here in Tulsa, Oklahoma that was data entry. I had data entry software for PCs, and that’s been where I’ve actually been based for about that long.

The latest sort of announcement that we’re involved with is the fact that we are providing a product called JES3+. That IBM has given indications pretty strongly that they were going to… They functionally stabilized JES3. And I guess there’s a lot of people in the JES3 community that really depend on that product. And we basically licensed the source code from IBM and we have actually put out our general availability, released one of the products about a month or two ago. And we intend to provide a completely upwards compatible version of JES3 that customers can continue to use.

Steven Dickens: Okay. That’s fantastic. A really good run through. I wasn’t aware of so much of Phoenix. I’ve kind of known you through your membership of the project of the role, but wasn’t so aware of the product range. Where do you fit into that, for our listeners, Eric? What’s your baby amongst those products?

Eric Chevalier: I’m responsible for a number of things. I was, like I said, the Zowe Pathfinder. I’m still kind of maintaining the mainframe side of it and doing a little work on the CLI. I’ve been involved in supporting the Viking products, which really have nothing to do with the mainframe, but somebody has to keep those customers happy. I’ve sort of a Jack-of-all-trades. I’ve been involved in some projects that didn’t get anywhere. We had some years ago, a product that was going to be a biometrics keyboard recognizer. Could tell who you are by the way you typed. And that never really came out. We were unable to implement that. So some of the stuff I’ve been involved in works really well. Some of it doesn’t.

Steven Dickens: That’s really interesting. That kind of leads me on. I mean, you’ve obviously got a long track record on the platform. We’ve got a lot of younger listeners. And we have the Open Mainframe Project interns come through every summer. What would be your advice to your sort of self, you’re coming out of college, that summer out of college, you’re 22. What would you be recommending to somebody who’s looking to build a career in the mainframe space, given your sort of experience on the platform as you are now?

Eric Chevalier: I would have to encourage them to, if they can, to get involved with the Open Mainframe Project. Because if you’re coming out of college now more than likely the mainframe is something that never got a lot of attention in your courses. It’s something that maybe you haven’t had any actual exposure to at all. And you might tend to think that it’s… Why get involved with the mainframe? It’s a dinosaur. It’s dying. But there are a lot of companies that still use it for mission-critical applications. And the idea of expanding its ability to be used by people who come from that non-IBM background. If you get involved with something like Zowe, that’s your chance to get involved in a project that can use the talents you probably have, the skills that you came out of college with, to keep the mainframe dynamic, and advancing and growing.

Steven Dickens: That’s really interesting. I think I have the pleasure at IBM to work with some sort of recent college hires who’ve come in and seeing them go through the journey you just described. I can bring something to this platform with my skills. I don’t have to be a 30 year veteran on the platform to understand it and add value, which is a key sort of message I would say.

Eric Chevalier: And it’s not just Zowe, because IBM, the way they have supported Linux on the mainframe if you have Linux experience, that’s immediately useful in an environment where maybe the only system they have is a mainframe, but it’s a mainframe that’s running Linux. IBM… I’m not sure. What I was about to say might be something that’s under a nondisclosure.

Eric Chevalier: Let’s not get ourselves into trouble, Eric. We’ve done so well so far.

Eric Chevalier: Let’s say that IBM, even in the mainframe environment, recognizes that Linux is important.

Steven Dickens: For sure, for sure. I mean, it’s almost 40% of the platform that we ship now. The capacity on the platform is Linux. It’s a strategic part of what we’re doing. It’s interesting from just my role within IBM, what we’re seeing and some of the strategic deployments with customers that are new to the platform, that never had the main training before, but also clients who were looking to… Who maybe had got a 30 or 40-year history with the platform. Who is getting into Linux and Linux in a big way?

So as we start to think about wrapping up, Eric. One of the questions I always ask of our guests is, I’ve given you a crystal ball, you’ve got the chance to look into it and maybe not pick stock numbers, or stock prices, or lottery numbers, but maybe all you’ve got is the ability to kind of look ahead to the future of the mainframe platform. From your viewpoint, where do you think we’re heading? What does the next sort of three to five years look?

On this podcast, Sebastian discusses his journey with the mainframe, the Open Mainframe Intern Program, becoming an IBM Champion, and where they see the Mainframe going in the future.

Eric Chevalier: Well, I think you’re going to continue to see the mainframe enhanced in ways that facilitate… I’m Tying to think about using some big words here. Facilitate the ability to support things like Linux applications, open source applications, products that 20 years ago, 10 years ago, maybe even five years ago, people would never have thought of running that on the mainframe. So things like Linux. I was under the impression that IBM is working on kind of bringing containers to Linux. I think their goal seems to be whatever you can run on your Intel box, you can run on the mainframe.

Steven Dickens: And that’s interesting. I mean, that’s definitely the vision. We’re working with some of the Linux distro providers. Obviously, Red Hat around what they’re doing with OpenShift and their container platform. We’ve got similar conversations going on with the guys that at [inaudible 00:25:03]. We’ve got [inaudible 00:25:05] as part of the project here and do great collaborations with us. They’re making some pivots within their product portfolio and we’re looking to bring that to the platform with those guys. So I think I see that same vision that you do around what we’re going to see more and more open source commit to this platform.

Eric Chevalier: Mm-hmm (affirmative).

Steven Dickens: For sure. So Eric, any sort of final thoughts, or comments, or things you want to share with the listeners before we start to wrap up? This has been a fantastic few minutes.

Eric Chevalier: I’ve enjoyed the conversation. It’s been fun sort of speculating on wherever the mainframe’s going. It’s been an interesting platform to be working on. I got my mainframe experience in college on an IBM 360 Model 30. That’s what kind of hooked me. And it’s been an interesting road because it’s been a road that always has new things to learn. And that’s one of the things I like about where the mainframe is going, is that the mainframe of today is not at all the mainframe that I was working on 20 years ago. And in what IBM is doing, encouraging some of the things we’ve talked about, it’s going to be continued to be an environment where there could be more things to learn.

Steven Dickens: So Eric, that was a great audition for my job. I couldn’t wrap that podcast up any better. I mean, what a fantastic way for us to finish. It’s been great chatting with you, Eric.

Eric Chevalier: It’s been fun talking to you.

Steven Dickens: I think the listeners are going to really enjoy the show. I think a great perspective. Certainly, go and check out Phoenix Software. So I’ll wrap us up there. My name is Steven Dickens. You’ve been listening to the I Am A Mainframer podcast, brought to you by the Open Mainframe Project. Please click on the links below and subscribe and give us a rating. We’d love to see you on future shows. So thank you very much for joining us and we’ll speak to you next time.