Category: Features

I recently had a chance to sit down with IBM Distinguished Engineer Rosalind Radcliffe about her new book, Mobile to Mainframe DevOps for Dummies. The book is actually quite a bit deeper than your usual “for Dummies” book and an excellent read. Of course you would expect as much coming from Rosalind who has quite a list of well deserved accomplishments and accolades on her CV.

Saleem Padani of IBM has a good post up on LinkedIn announcing the book as well that you may want to check out.

FYI, we will be doing more DevOps Chats in the future. If you would like to be a guest on a future episode please contact editor@devops.com

Above is an embedded audio player to listen to the interview of Rosalind. Below that is the transcript if you would like to follow along or just skip the audio and read the interview. You can download the free e-book copy now by visiting https://ibm.biz/m2mdevops

 

Audio Transcript

Alan: Hi, everyone, this is Alan Shimel, Editor in Chief of devopsy.kinsta.cloud, and we’re kicking off a new feature here at devopsy.kinsta.cloud—DevOps Chats. Each episode will feature me speaking with a thought leader around a particular topic of devops.

Kicking off our series today with a very special guest, Rosalind Radcliffe of IBM, and we’re gonna be talking to Rosalind about a new book which she has just offered called Mobile to Mainframe: DevOps for Dummies. Rosalind, welcome to devopsy.kinsta.cloud.

Rosalind: Thank you.

Alan: Okay, let me briefly give our listeners a little background on Rosalind, and Rosalind, I don’t mean to embarrass you, but you have quite a résumé. Rosalind is a Distinguished Engineer in IBM. She’s Chief Architect for CLM and DevOps, and she is responsible now for driving devops for multi-platform architecture. This includes System z and Power Systems. In addition, she is responsible for the architecture for the Collaborative Management capability for Enterprise solutions, and this includes both UrbanCode Deploy and Rational Team Concert’s support for standard mainframe development activities.

Beyond that, Rosalind is a member of the IBM Academy of Technology and a master inventor. Prior to her current development related responsibilities, she has been responsible for IBM’s SOA Management Strategy and Services, responsible for deploying Systems Management solutions, and she started with ISPF development.

Rosalind, I hope I didn’t embarrass you too much, but once again, welcome to devopsy.kinsta.cloud, and congratulations on this great new addition to the DevOps for Dummies series, Mobile to Mainframe DevOps for Dummies.

Let’s start today’s chat, if you can share with our audience a bit more about you and where you are in your career.

Rosalind: Thank you, and I’m looking forward to this chat. So, a little history on why I wanted to write the DevOps for Dummies book. I started in IBM 28 years ago in ISPF development, as you mentioned, and I’ve been in different roles and different responsibilities along the way, including spending a good amount of time in the Systems Management organization, responsible for actually doing operations like tasks and deploying our Systems Management solutions for clients.

I’ve also spent a number of years in the application development space, providing and building application development tools. And so, as this developed over the years, working in operations, I used to blame development, and then when I moved to development, it was, “Oh, I guess I can’t blame operations.” So it’s deemed appropriate that I would move into this devops space of trying to bring these teams together.

In looking out at the industry today, I see that there are lots of questions, especially in the mainframe space, about, “How does this devops really apply to that environment, and with all the drive for mobile and mainframe together, how do we really make this play as an environment?” And so, that’s how I got here and why I wanted to write this wonderful book.

Alan: So, let’s dig in quickly to the book. You know, I think everyone in our audience is familiar with the “For Dummies”, if you will, franchise, and some of them can be rather fluffy, some are better than others, obviously, depending on authors and so forth. My initial reaction when I read it, Mobile to Mainframe DevOps for Dummies was, “Wow, that’s a pretty heavy topic for a For Dummies book.” Why write this book in the first place and why now?

Rosalind: Well, I think now is a perfect time, with the industry evolving, with the mobile frontier and everybody focusing on mobile, and with mainframes truly being our systems of record. So, everybody needs to connect mobile to mainframe, and if we look at the development environment, most mainframe development shops look remarkably like they did when I started 28 years ago, and that’s not necessarily a good thing.

So what I wanted to do was help explain to executives, senior leaders, and anyone in the industry, how mobile and mainframe play together and, in using the devops transformation helping them work together, they can get more value out of the systems that they currently have.

Alan: Absolutely. And just so we can mention, of course, Rosalind, the book is available for download, we’ll have links on it up on devopsy.kinsta.cloud. But, as I mentioned, it’s not a fluffy book—there’s a lot of great content here. I can tell how much time and effort you put into this thing.

But, give our audience a sense—how much, how big an effort was it for you?

Rosalind: It was a significant amount of time in writing. I spent a lot of time on plane flights back and forth to Europe, and that gives you a lot of time when you can’t do anything else, and so I spent a significant amount of that time writing. It’s a relatively short book, it’s only 70-something pages, so it’s not as long as a very long book, but it did give me the opportunity to hopefully give some background, explain more about the current environment, and give people an insight as to what I’ve been seeing with clients and help them move forward.

After you write it, then you have to edit it, and I have to thank the many people that helped me work on this and help provide contributions and pieces of the document, because they helped me put this together, and helped me do a lot of editing as well to make sure that it was complete in the environment.

Alan: Sure. I mean, I’ve been lucky enough to have a chance to see a pre-publication version of it, and I do think—you’re right, it is only 70-some odd pages, but it’s pretty in depth. But it really benefits from the sort of, the For Dummies kind of franchise—you know, the icons they use to help you focus in on what’s really important and giving you Cliffs’ Notes version if you wanted to not read every word. You used the format very well, and I think people will get a lot out of the book.

Let’s dive in a little bit, though, into some of the key points that you’ve brought up in the book. You know, just picking some random ones that I thought were important—in chapter 3, for instance, you speak about why we need to understand why devops is so critical for mainframe environments. And we need to start with the dev side.

But devops is about breaking down silos, it is about cross functions—and nowhere, in my mind anyway, nowhere are these silos more separated than between the mainframe teams, like you said, haven’t changed sometimes in 28 years when you started, and the distributor mobile development teams, who are today all about cloud and virtualization and instant up and instant down.

Can you elaborate a little bit more on this point for our audience?

Rosalind: Yeah, I’ll be happy to. This is one of the key points as to why devops matters more than ever now in the mainframe space—this transition to mobile, this transition to systems of engagement and their dynamic changes, their need for speed, and the need to access these back end systems of record. So now is the time for us to evolve the environment, the mainframe development environment, to modern technologies and modern tools to allow these teams to actually work together.

There is absolutely no reason that mainframe environments need to do development the same old way. They don’t need to be done using waterfall development practices. There’s no reason not to apply the same agile and lean IT practices across the development organization from mobile to mainframe, and using the same tools and the same capabilities so these teams can work together, they can share information, they know what each other’s doing, and you can break down those silos.

One other reason those silos are there is, you know, if your mainframe development team is using ISPF, I’m not sure those mobile guys using X code even want to look at a green screen. So this has to change. The perfect time—bring these teams together, use modern development practices, use modern tools, and this allows the silos to be broken down, allows the teams to work more efficiently together, and helps deal with this variable speed IT issue that we have. Mobile needs to move fast. Those systems of engagement need to be moving.

But the back end systems of record need to be able to define standard APIs. They need to provide those services that the front end needs, and they need to do it fast enough—so they need to do it at the speed of business, and devops helps them do that.

Alan: Excellent. And so, Rosalind, you actually touched on something here, and it’s this variable speed IT—variable speed devops, I’ve heard some analysts call it a bimodal  IT—and it, I mean, you nailed it right on the head. These mobile development teams that are running at the speed of today’s business, tearing it up, putting it up, tearing it down, putting it back up—changing, changing, changing thousands of hundreds of times a day, even, in some cases.

We can’t expect the mainframe systems of record to perhaps move at that pace, but there’s no reason they can’t move faster, especially using modern tools, modern development environments, though I still find something comforting about a green screen. [Laughter]

But, that being said, you know, this was the subject of a whole series of webinars we did here on devopsy.kinsta.cloud on variable speed IT and variable speed DevOps. Can you help our audience understand why it’s so critical for mainframe environments to embrace devops and why now, more than ever, now is the time?

Rosalind: Yes, so with this variable speed IT, and as we talk about it, it’s not that the back end is slower because mainframe development is slower, it’s that, in variable speed IT, you need to move at the speed of business, and if you define services as your back end systems of record, those don’t need to move as fast.

What we need is to define those as services so the front end system can use them, so these systems of engagement can call back into the function. So the need to define these services based on the existing legacy code means development teams need to be able to respond at the speed of business. They need to be able to move faster. And the old ways of doing things—you know, making a change and testing it for three months—just isn’t gonna work. We need to apply the principles of automation, automated testing, virtual services, and interface testing, and bring all of those concepts to allow you to build those services based on the existing business function into the environment so that they can then be called by the systems of engagement.

Companies have significant assets sitting on the mainframe systems. Mainframes aren’t going away. They’re gonna be here forever. They provide the qualities of service needed for large scale transaction performance systems. And so, with mobile, large scale transaction performance systems—that’s perfect. Mobile backed with mainframe is a perfect environment. And companies with the assets that they’ve built up, if they can service-ify them fast enough, they can provide disruptive technology today.

And that’s what the companies need to do. They need to take advantage of the assets they have, providing services using the new development practices, and then it all fits together and they get the business value at the speed they need to. And you know, I helped build ISPF, so I do like that green screen, too—but it’s not really the right development platform for developers today.

Alan: Yeah. But what about—look, you know, people make fun of COBOL developers, but there’s still a lot of jobs out there for COBOL developers.

What about the shortage of experienced mainframe IT skills in the marketplace, Rosalind?

Rosalind: Yeah, so that’s one of the points I make in the book. Developers can write any language. You know, get a good developer, they can write COBOL—and there’s nothing wrong with COBOL. COBOL serves its purpose. PL/I serves its purpose. And some people will cringe as I say this—Assembler serves its purpose. But you know, I’m not gonna write it using ISPF.

If I’m a kid coming out of college and you give me a modern IDE that has all of the language syntax in it, it has code rules, it has the ability to build unit tests for me—I can write COBOL. Not a problem. It’s the—give me an ISPF screen, and I don’t even know how to use it, much less how to do COBOL.

Alan: Understood, understood. Let me—we’re running low on time, so let me turn to some other questions I had prepared, if it’s okay. In the book, you talked about some of the best practices—devops best practices on System z which, for our listeners who may not be IBM aware, System z is the IBM mainframe platform.

Rosalind: Correct.

Alan: And there are several good ones that you have highlighted in the book. Are there specific devops practices that you would recommend for System z to get started down the path of continuous delivery, which some people think is the ultimate aim of devops, or the ultimate goal?

Rosalind: Yeah, so there are a number of different practices, but a few call out based on the way many mainframe organizations are working today, and it has to do with where they are today. How do they improve the process from where they are to get into more continuous delivery practices?

And before someone says something, I’m gonna say—continuous delivery does not necessarily imply continuously delivering into production. It means continuous delivery into a pre-production, et cetera, environment. You may still need to stage those based on an outage window, especially for systems of record, because you can’t afford to have a transaction change mid-flight.

So there are some—you know, continuous delivery within reason. But if we think about those things, one of the key practice is modern development tools to break down the silos. And one of the others that is absolutely critical is automated testing. There is no way to get to continuous delivery without automating testing, and since you’re building services out of your back end systems, services are a perfect thing to use automated testing against.

So, building up automated testing based on the service interface, being able to test the function without having to have users go in and test the business function—just run a bunch of automated tasks and make sure you’re picking the right automated task based on the code that you’re changing, but automated testing is probably the biggest change that is needed for mainframe environments and, in many ways, the most valuable. We look at many shops and they’re spending months running lots and lots and lots of manual tests that they hope will find the problems.

Alan: Got it, got it. Rosalind, we only have time for maybe another question or two, but I have one very IBM specific one, one a little bit more general.

So recently, the new z13 launch really put the mainframe front and center as the core element for IT infrastructure.  Is this the rebirth of the mainframe? How do we see this new mainframe fitting into this seemingly ever changing IT landscape? More importantly, what do existing mainframe clients do to prepare for this new age of mobile to mainframe?

Rosalind: So, the z13 has lots of additional capability and functions along with the updated compilers to increase the speed to allow you to provide more function faster, COBOL improvements using the new system hardware, Java improvements to be able to do more Java development on the mainframe, as well as significant memory availability changes.

So, with the combination of the system, you can look at it as the best Linux environment that you can use for your cloud environment, so you can have—use Linux on System z to provide your Linux environment for your front end applications, and then, looking at z/OS, you can start to take advantage of this new memory that’s available to increase and improve your performance from a caching standpoint, numerous other things from a performance standpoint with the new compilers, and the mobile connectivity into the system. If we think about the fact that that could drive huge numbers of transactions—could drive huge numbers of transactions—we get the mainframe as the perfect platform to handle that variable load.

So you never know—you know, I guess if you’re a retailer you know if it’s Thanksgiving or near Christmas it spikes, but many times, you don’t know where the spike’s gonna come, so it’s important to have that capacity on demand, to have that variability, and so the system plays very well.

I think there are a lot of clients and a lot of organizations that are taking a new look at the mainframe of the latest z13 because of its capability, because of its enhancements from a memory standpoint, and because they think—you know, reality. People are realizing the mainframe is not going anywhere, and it really is the best platform for those high transaction systems and allows you to consolidate things together in a Linux platform. So it’s a great advantage that customers were already using we they have, and it’s something other customers are looking at to do, “You know, I can take advantage of that as well.”

Alan: Okay. Rosalind, again—I’m sorry to keep saying this, but I’m trying to keep it short for our listeners, and relevant, but there’s just so much here. So you’re constantly out talking to clients that utilizing mainframes, and kind of living this devops and mobile to mainframe way of doing business—can you give me two or three quick examples of companies that are, that you think are doing it right?

Rosalind: So, if we look at the industry and we look at the publicly available references, Nationwide is one of them that’s been out talking and describing their devops transformation and how they’re exploiting the fact that they’ve brought their teams together to provide value faster in the environment. We can look at Fidelity, who’s talked about their improved deployment capability and their improvements overall in their environment. Or, one of the examples I like that Walmart and their example of using services on the mainframe to actually provide infrastructure services to all their distributed applications from a caching service.

And so there are numbers of good examples that talk externally, and then there are a whole bunch more examples who consider their transformation to devops a business advantage. They see that, by transforming to devops, they’re getting a leg up on their competition, and so they’re not as willing to let us use their names—but there are lots of wonderful examples of people who have transitioned and who are getting the value out of it, including IBM itself. We have numbers of examples that we provide to show our development process significantly improving and breaking down the silos within our own organizations and improving our ability to deliver.

Alan: Excellent. Rosalind, I think this was very helpful, as is the book—a great guide for, as you mentioned earlier, both executive managers and practitioners.

What about you, Rosalind? You’ve had such a successful and storied career—what’s next? What’s next for Rosalind?

Rosalind: Yeah, so the next thing next week, actually, I’ll share. So, next week, we have one of the major user conferences or the major user conference happening in Orlando. And, as part of SHARE, obviously, we talk about mainframes and development activity as well as lots of different topics, and I’m looking forward to that in particular, because we’re doing a series on mainframe and mobile. More specifically, we’re doing a discussion about hooking Bluemix, our development app platform as a service along with IBM devops services to the back ends, and helping show the relevance of Bluemix to mainframe environment through a series of sessions as well as showing System z as the new cloud disruptor, and there are numerous sessions by other clients like Walmart who will be speaking.

And so it’s a great opportunity for me to, first to hear stories from our clients as well as to provide some additional sessions for our clients and show what we’re doing, and it will be my first opportunity to have hard copies of this book and do a book signing for any of the client who are there.

Alan: That’s what I was gonna say—there’ll be plenty of copies of the book there available and signed by yourself.

Rosalind: Absolutely.

Alan: Great. Okay, we’re gonna—I think we’re gonna call it a wrap here, Rosalind, and thank you so much for your time. I know you’re busy and I appreciate it. Rosalind Radcliffe, author of Mobile to Mainframe DevOps for Dummies, available now for download or if you’re attending SHARE in Orlando next week, you can get an autographed copy. Rosalind, thank you so much for your gracious time and spending some time with us today. We hope to—you know, best of luck with the great book, highly recommended, and we hope to have you back here soon.

Rosalind: Thank you for having me, and I’m happy to come back and talk to your audience whenever is appropriate.

Alan: Okay. Rosalind Radcliffe, IBM, and author of Mobile to Mainframe DevOps for Dummies. This is Alan Shimel for devopsy.kinsta.cloud, and thank you for listening into this DevOps Chat, and join us again soon for another one. Have a great day.

That’s a wrap on this DevOps Chat. If you would like to be interviewed for a DevOps Chat please contact me at editor@devops.com