PurePerformance - Dynatrace Perform 2019 Devops, Containers, Microservices and all that with JP Mogenthal

Episode Date: January 30, 2019

JP Morgenthal, CTO of application services at DXC Technologies talks to us about all the things devops you were afraid to ask.

Coming to you from Dynatrace Perform in Las Vegas, it's Pure Performance! Hey! We're back! Hey, we're back. Live from the Marketplace. Yes, Diatrace Perform 2019. 2019. This is Las Vegas. Las Vegas. Brian Wilson, Mark Tomlinson, and James Pulley. And our next guest.

Our next celebrity guest, J.P. Morgan Paul. From DXC Technologies. Welcome, J.P. Hi, thank you. Welcome. Have you ever been on a podcast before? Many. Really?

Yes. Are we the most unprofessional one you've been on? No. Oh, good. That's good. We can take some feedback. We welcome that. But it's awesome to catch up with you on DevOps, microservices, containers, and everything else that you were afraid to ask.

Lions and tigers and bears. Oh, my. Tell us about it. Actually, it's funny. I just put out a blog with every major name of every major emerging tech, and the last thing was oh my. Somebody else titled it for me. But we tend to go there, don't we, from that, see how the impact of the Wizard of Oz on our social culture. Yes, absolutely. Amazing stuff.

Listen, I love this stuff. I really do. I've been doing technology since the early 80s. What really excites me is how all these things connect. It's rare that we've had such a massive transformation or change going on simultaneously where so many new components have emerged at the same time. Yeah. That need that have been transformational. And it's hard to decide, you know, which ones will I pay attention to?

Which ones can I ignore? Because I can't eat this elephant all in one bite. Right, right. Right? So it's like, you know, you have the microservices around some Greenfield stuff, but also modernization. You have introduction of IoT. You have, you know, container technology,

which is maybe used to deploy your microservice, maybe not. Yeah, you have functions as a service, a lot of Lambda and things of that nature. And then you have the whole Agile and DevOps going on. And somewhere back there is a mainframe. Somewhere back there is a mainframe. And someone behind a curtain. Of course it's now partitioned into Windows

machines, but... Yeah, no worries. Not that that's a problem. We were just even talking yesterday about running a VM in a container. Yes. Inside a burrito, wrapped in a pizza. We were just even talking yesterday about running a VM in a container. Yes. Inside a burrito wrapped in a pizza. Yeah, somebody was talking about that, right? They are now doing virtual machines in the container itself. Well, the container is, let's face it, it's a great packaging scheme, right?

Yeah, exactly. And so, I mean. How do you know what to keep up with, though? The biggest part of the container people don't realize isn't the container technology because the separation of the process has always been, well, has been part of the Linux operating system for a very long time. It's the addition and the packaging of the storage file system that it builds on that keeps everything into that secure model that has really revolutionized all of this, right?

Yeah. And then, of course, you start adding in the networking and the ability to have external volume control. And now you have these little packageable units that can travel. I mean, you can put data in there and send them to somebody. They're like little USB thumb drives, but in a file form. It's a long way from using Ghost to sysprep servers from bare metal. Oh, definitely.

Thank you. And there's still people doing that. Anyone shivering? Rankin-Sane is on bare metal. Kubernetes is on bare metal. Yeah, that works. It works. It's interesting stuff.

But, you know, what's really crazy is how all this is coming at the same time. And I just, you know, I know I come from the integration days, right? I did a lot of SOA, a lot of the integration stuff. And that was complex enough. I mean, people never understood things like Corba and RMI. I actually was one of the first people to do talks on RMI for Sun as an analyst. Sure. And the funny thing was, like, no one ever picked it up.

No one really grokked it. It's like beyond. And then you add the orb from Corba into the RMI. And it's like, okay, you know, I'm completely lost on this thing, right? We were on the other side of that working in the performance testing universe where we're having to test, you know, DCOM, Corba, RMI. There you go. And, you know, all sorts of other different interfaces.

And it's hard because you've got all that network. You have to understand network and network latency, and then you have to understand the machine CPU. And you have to know where all the bugs are in a DTC. I don't know what that means. Distributed Transaction Coordinator. Something you never wanted to use for Microsoft but couldn't avoid it. This is psychologically an issue for me now.

In DXC, we have the digital transformation centers. Oh, there you go. So you say DTC, my mind goes right to digital transformation center. No, I went somewhere worse. Much more dark. I'm just falling. It's the dark web. That's where SQL DTC went.

So I see companies starting to play with this stuff, but I just can't imagine how they you know, how they can expect to have any level of maturity around the solutions that they're building on this, right? And people, you read about these companies, oh, we've done this. And they make it sound like, you know, they've created this masterful thing using all these technologies. And then, you know, if you look under the hood on this stuff, right? I mean, you probably see it's like Kelo World, right? Because people want to give it the press and shove it up and hype it. But the truth of the matter is, this stuff is so new, you can't know the ins and outs of every one of these technologies and how they work.

I love the recruiters that ask for a decade's worth of experience in a technology that's less than 18 months old. Those are my favorite. Yes, I'm looking for somebody with 15 years of React, JS. Exactly. Yes. Good luck with that one with 15 years of React, JS. Exactly. Yes. Good luck with that one. Do you expect to get paid this year? And they'll find someone with a resume that says that.

Yeah. Well, that's even scarier. Yeah. Yeah. We'll make up anything to get a job. So tell us a little bit more about DXC. Sure. So DXC is one of the largest IT professional services companies in the world.

Cool. We are the product of the merger of CSC and HP Enterprise Services. Okay, yeah. So the group that had acquired EDS and managed that service function. So strong heritage. And not CSC like CSC. This is the corporation. Computer Sciences Corporation.

Computer Sciences Corporation. Yeah, okay, okay. So these are two companies, been around for many years, strong heritage in managed services, strong heritage in safe pair of hands, running businesses, billions of dollars of transactions and systems value, right? Yeah. But the heart of the company during the transformation, the reason why we picked a different name is because we really have embraced the digital era and want to help companies transform. And we've spent a lot of money and time transforming ourself through that merger. Tell us more about that. Well, I'll give you a great example so one of the things that we've done is

that we have uh we're like any other company we have uh system software that we need to run our business right and this isn't you know when you're a professional service organization it's not oracle off the shelf sass yeah right it's some really unique stuff because you got ops you got finance you got you know metering you got some s in there now, you've got service management, all the resourcing, delivery, right? And your delivery might be on site, might be in a delivery center somewhere. So global companies, 70 countries, right? And a highly mobile workforce, right? Highly mobile.

Yeah, yeah. So you put all that together, and clearly there's some software you need to help run that kind of business right and so you know we had a version that we you know chose some pieces from each of the merged companies but it wasn't you know the long-term initiative right we refocused internally and started building a uh a platform to run our business on and in in in doing that, we actually leveraged our internal product engineering teams and CIO teams to work on this platform and use that opportunity. We're about to enter now the eighth what we call build-a-thons. We'll send 200, 300 people to Bangalore, to Manila, Plano, Texas, right? And they will work two weeks straight, sleep, eat, code. That's, you know, and they learn how to do Agile.

We use Agile development tools. We have a CICD pipeline. So they learn what it's like to go through that, to actually deploy their code and into production and release and testing, all using, you know, a common DevOps metaphor. Cool. And so, you know, we've trained maybe, you know, 1,000 to 1,200 people now that used to probably work in old waterfall methods. Yeah.

And how to live in an agile team. Yeah. Right? And sprint hard for two weeks in a large group. Yeah. Sprint hard for two weeks in a large group. And because you think of a professional services company, what's the one thing you can do? Is ever take them out of the field to send them training, right? They never get to do it, right? That's the bench time statistic.

No one likes to see MPSO. Right. Utilization down or something. Yeah. But, you know, through this method, and this is what companies want. They want to know, you know, they shouldn't be looking in our box, right, because you're buying a service, right? You go to McDonald's and ask, can I walk around the kitchen? You don't.

Might not be a bad idea, though. It may not be a bad idea. But customers come to us and they're like, well, can I see what's in the box? Can I see how you make the sausage? It's important to me to understand how you make the sausage right i'm really into it's important to me to understand how you make the sausage so we walk them around our dtcs where we do design thinking on one floor and then we take them to a floor where we put out together a backlog and a user case stories we put them through minimal viable products start to incrementally cool they immediately gets value yeah you know from that process right we um so they want to see that we are really embracing and delivering in this way

and that they're not getting Agile on the outside, Waterfall on the inside, right? Yeah. Because they want the opportunity and the value that comes from Agile, which is I want to be able to change something rapidly. I may want to decide that this feature is more important than that feature, and I don't want to wait six months to determine that, you know, this feature is more important than that feature. And, you know, and I don't want to wait six months to determine that. Right. So I want that collaboration and that involvement with my people.

Right. Right. So really interesting stuff. Cool. Cool. And how long from the merger? It's been two and a half years? It's almost two years now.

Almost two years. April will be two years. Yeah. So this has been going on this entire time. Yeah. Very cool. What brings you to Dynatrace? What's interesting about Dynatrace for you?

So Dynatrace partnership actually started when we were at CSC. Okay. We have a very strong relationship with Zurich, which we spoke with. Oh, yeah. I think they were one of the first customers to really introduce us to the Dynatrace product line. We actually have a management service that we had built for them way back around Dynatrace product line. We actually have a management service that we had built for them way back around Dynatrace. And so, you know, we got to know the company. That was when they were really APM. And then they started talking to us about their plans for the future

and where they're going. And we started to realize very quickly that the future you know their future and our future is aligned very nicely 8 p.m is you know is important and you know but there is a number of different companies out there that can do it but but somebody who can actually map out the entire flow of an application infrastructure yeah it's northeast, west, south, right? Up and down. Yeah. You know where everything's connected. You know how it's connected. You know where the problems are at any point in time. And then the AI investment to help us identify root cause analysis. I mean, for a company whose job it is for me to keep meantime to repair very low, I mean, they really have demonstrated an ability that I think is one of the tops in the market.

Yeah, that sounds cool. That sounds very cool. And are there any particular announcements? We've had the AIOps and the sort of open part of AI this week. If you didn't see, we're Global Partner of the Year. Exactly. Oh, congratulations. Well, you didn't mention that.

You buried the lead. Sorry. And then the next day, Rick Sullivan, who is our VP and GM for application app services, app dev section, went on to talk about how we're committed with them in this ACM. That's right. Autonomous cloud management. Right. The autonomous cloud program. Cool.

Very strongly behind it. It aligns nicely with our Bionics program. We have a program that in the marketplace is competitive for other services. We are basically looking to drive automation to reduce costs of delivery. Ours is Bionics. We leverage orchestration intelligence and automation as a way to identify waste and remove it from our customer environments and managed services. Right, right. As well as being your mascot is Steve Austin. Yeah.

Who's the Bionic man. No, it's Bionics with an X. Oh, so close. But the ACM storyline really aligns nicely with what we're doing there. And we're really impressed, highly impressed with the organization as a whole. There's good synergy in the field. Our sales teams are working well together. That's good.

Yeah, I mean, we consider them what we call in-app services, an extreme partner. This is a group that we work closely with. They do sales training with us. We discuss pipeline with them. This is a partner that, you know, it's one of those rare circumstances where it's not Barney paper, but we're actually really driving value for our customers. And our customers love their product. They really do.

I mean, it's not often that you say the name Dynatrace. Oh, we like them. We have some of that. Have you ever introduced Dynatrace cold to an account? Like just brand new in the door? But, you know, I don't live out on the front lines. No, no. Yeah.

It's exciting. And it's not a story that's emerged. But, yeah, we probably have. Certainly in Europe, I think, or Australia, we definitely have had, you know, we brought this in as a concept. And they're like, oh, that's cool. You know, it's part and parcel of our apps services automation, service automation offering. Okay. Right. offering. Our service automation offering is about detect,

correct, and then improve. We walk customers through stage. They are our detect arm. We try to leverage as much as possible Dynatrace as the

main tool for the detect. You get some clients who are like, I'm already on X, I'm already on Y. And you can't help that. We've got to live with that. But when we show them what Dynatrace does, they also are like, oh, this is cool stuff. How do I get this?

My other stuff can't do that, maybe. Do you leverage any of the other integrations via the API or plugins into the pipelines based on not just what you can see in the browser integrations via the API or plug-ins into the pipelines, you know, based on not just what you can see in the browser version of Dynatrace, but all the back-end integrations that can be done? Yeah. We're also one of the largest ServiceNow integrators.

Oh, yes. Yeah. Really? We're fruition partners. That's a huge piece. We acquired CSC, and so now we have a huge ServiceNow practice. Wow, yeah. And so, as you know, they're peanut butter and chocolate, Dietrace and ServiceNow. This is a second reference to Reese's peanut butter cups this week.

Well, I think there was a few out here yesterday. They put those on us. Oh, man. Maybe it got in everyone's mind. Maybe it got people talking about it. But, yes, so they really are. And so we leverage that integration. But also, we're going to be heavily leveraging their APIs as part of our automation platform that I was talking about earlier that we run our business on. We're going to be leveraging a big part of the ACM as well. of helping our customers understand how that APM works and how you're shifting left

based on moving the testing earlier on in the process. And then, of course, that gets into a lot of leveraging their APIs and their tagging architecture. Yeah, yeah. Awesome. Very cool, very cool. Are there any... I was mentioning the free dev tools or the session replay.

We've been kind of digging in the announcements this week. We really haven't gotten into session replay, which I think is very, very cool. I think that that's a whole other area of ops. And it really gets down to trying to troubleshoot some things. So you really got to get to the point where you're using Dynatrace effectively be you know into the the l3 uh oh sure right yeah and so i and i think it's coming it's just not there yet i i think they're a little early on on what that thing does and how it works now on the network side that you have something like net witness which does that and for packet relay and stuff like that people love it so it's just

a matter of time to catch up and understand you understand the ability of this product to be able to help you understand exactly what's going on in your apps. Those bugs that you just, I give up. Why won't this work? We all have them built over time, right? And then you get way back and you go, oh, look at that. Okay, I missed that one completely. What about the developer tools, free for life? I mean, that seems like very disruptive from a market perspective with great benefits. Well, you know, we're not an end user customer in a typical way. So, I think that's great for, you know, their direct customers. But, you know, with us, it's, you know, phenomenal.

We have such a different relationship in how we go to market and how we work together. I have an out-of-the-blue question, kind of. We mentioned all these funny old technologies just because funny old technologies are funny. You mean like Microsoft Bob? Microsoft Bob. Clippy. Clippy. Clippy, yeah.

But honestly, like the title of our live episode is like DevOps, microservices, containers, and all that jazz. You seem like you're in a spot where you get to play in the sandbox and witness all of this stuff both internally and with your customers. It's exciting. Yeah. So my question is, there's so much new stuff. What are the things that keep companies from just kind of big hug and bring all this stuff together? Go Slack. The biggest issue we've continually seen with businesses is that there's no Slack in the schedule, right?

Oh, yeah. There's just no time. You need people. People need time to sit and learn, right? Where are these developers all coming from that are available and they're working on their own time? They're doing stuff. I do it myself, right? Udemy courses. Udacity courses, right?

Huge opportunity, right? Training programs like the academy that they have online, right? You put yourself through it when you have 20 minutes, 30 minutes, right? You watch a few episodes and you pick it up, right? And that's kind of the way it goes. Those people, by the way, are rare and they're starting to see benefit in their own careers. But when you go out there for a company like ours where we're highly dependent on

the human knowledgeable resources, there is not an abundance of microservices engineers and architects. There is not an abundance of microservices, engineers, and architects. There are not an abundance of container experts. Right. There are a bunch of people out there that claim they are. Yeah. But real hands-on. Somebody can walk in and look at your system and go, yeah, let me tell you exactly what's going to happen here.

Yeah. They don't exist yet. Well, they really haven't. And the ones that do are working for startups that are focusing on delivering tools around this, right? Right. So, you know, Pivotal right there, it's right across from us, right? Cloud Foundry, Pivotal, PCF developers, extremely hard to find the industry. Yeah, yeah, yeah.

I mean, very rare breed. Yeah. Because it requires you to understand not only how to write code, anybody could write the code, but how to build a cloud-native application. Sure. And in some ways be full-stack knowledgeable, right? In that context, yeah. And then you have on the operations side, which is also a big aspect of this for us, the site reliability engineering thinking.

And as described by the person who created this SRE team, SRE is what would happen if you told developers to go build an operations organization. Thank you, Google. Yeah. I actually, I get it. Yeah. You want that automation. You want that. So, you know, the funny thing is, way back in 2010, you know, the Ghent had just really recently happened.

And I found myself into the DevOps community through my network. And I got to have some talks with Patrick Dubois. Oh, yeah. Yeah. You know, and I got his story and what led to all this. You know, again, for him, why did you do it? And he was like, well, I work in the middle of the dev and the ops teams. Doing both jobs.

Yeah, he works over, he was in state government in Belgium. And he said, one, I was highly impressed by the rigor that some of the dev teams were demonstrating following this Agile manifesto. And I didn't understand why the ops team didn't have a similar thing. Why didn't they have an Agile manifesto for all their development work and their scripts? I mean, Lord knows, we've been scripting for years. Anybody know what set an awk is? I mean, this is not new stuff, people, okay? If you think this was all created in the last two years, you're wrong.

These utilities did not write themselves. Exactly, all right? Lex, yak, awk, set, all right? If you don't know what those are, then go out and learn history, all right? Because this wasn't done last year or two years ago. But what he appropriately recognized, and you've got to appreciate this, is there was no rigor around it. Or structure. If something had to be done, you had to call Jim or Susie who wrote the stuff. Because nobody else knew how to kick it off or what it did.

And nobody knew how to change it. But now, all of a a sudden we have infrastructure as code we have this stuff defined as you know yaml and it goes into the git and it's part of the testing process now yeah so he's brought about a really cool change in the industry by really just saying i learned something cool over here it's we're in the same industry it's not even like i learned some automotive and i'm playing into finance this. I learned this in other half of IT and you guys aren't doing this. Right? So I think that's pretty cool.

To me, that's the most exciting part of this DevOps story is watching what Patrick saw really start to be embraced and go, people go, yeah, I got no good answer why we're not. And that makes sense. Let's get rid of the heroes. And then you have Project Phoenix. Oh, yeah. I got no good answer why we're not. And that makes sense. Let's get rid of the heroes. And then you have Project Phoenix. Oh, yeah. What was his name? Chuck? Carl?

Steve? No, was it? Trey? Trey, something like that. Yeah, the hero guy, right? We all know who I'm talking about, right? Close the door. Don't talk to him.

Put pizza under it. Exactly. He doesn't wear shoes and doesn't shower. Wait a minute, that's it. Exactly. Who doesn't wear shoes and doesn't shower. Wait a minute, that's me. Exactly. But honestly, all of those things, like you say, bringing about change, there's still a target customer

that's a large enterprise that hits a tipping point where they're like, we need to start adopting these things in transformation. So what does that look like when you see a customer that's like, this is kind of how we guide them through a transformation? Well, we've had to go and develop an enablement. And the enablement, actually, we have some really great enablement programs that came over with HP Enterprise Services into the merger.

And we have Greenbelt, Whitebelt, and Yellowbelt Dojo. We have a Blackbelt Dojo. That really walked people through the history as it's explained to me because I'm a CSC-er. Yeah. But the history as it was explained to me was it actually originated with Gary Groover, who's well-known in the DevOps industry, when he was on the HP printer team.

That's right. And the whole, hey, we've got to put our quality software in the printer. Yep. So these practices originated all the way back to then, the dojo concepts and training people on how to do this all originated from that. Cool. And we have that content updated for today's DevOps world. And again, where you're saying just there's no slack, there's no room in the schedule

for these companies when they're already flying at 600 miles an hour. So one of the first things we do with our customers that are interested in help is we do a value stream mapping. And during that value stream mapping, we identify bottlenecks and constraints, areas where there's latency and areas where there are holdups. And then we target those immediately. And by targeting that waste and that latency, we begin to introduce Slack into the process. Yeah. And then you can use that Slack effectively.

Yeah, to re-enable and start more transformation activity. And then there is a tipping point where eventually... Then there's that tipping point, right. Yeah, exactly. I'm with you now. Yeah, that makes sense. So I used to do a lot of public speaking on DevOps. Sure.

And with the get starting point, I was at that point in the conference talking phase. And so my last slide, how would you get started? If you had one thing you should do this year, what would it be? And it was always introduce Slack into this. Find out how to introduce Slack into your schedule. Yeah. And then come up with a strategy to use it to create more Slack and get you to the next step. Because without that Slack, there's no time.

I mean, if you're always putting out the fires, you never get out of the cycle. It's a vortex. It was the guy in the Phoenix Project who was just drinking himself to death and going to quit. Well, he wasn't going to quit. I don't remember. He was just the hero. Everybody had a call him to do everything. It was the other guy who was pulling his hair out. Remember? It was like,

let's go out and just have some one-on-one time. The interesting thing about the Slack time you're talking about. We've had on our podcast a few times Donovan Brown from Microsoft, VSTS team, and he's a big proponent of saying no matter what you're doing, find 15 minutes in your day, not outside of your work hours but during your work hours, to watch one of those class videos, to read one of those articles, to do it because his whole point, and again, he's coming from the luxury working in Microsoft, but I think it's kind of realistic for anybody to say if something is going to fall apart because of 15 minutes, it's going to fall apart with that 15 minutes too. Yeah, people can't be 100% utilized. Right, so finding that time and just making it a little bit work, and as you're saying,

20 minutes here, 15 minutes here, 15 minutes here, you can start getting into a cycle of getting to those places, right? And it would, yeah, obviously a lot of people are going to have to do some of this stuff after hours. Right. But there's also, as part of one of the themes was today is let's make a better life for ourselves, you know, and all this. So take that 15 minutes, end of your day, beginning of your day, watch a bit and.

Or script something. Yeah. Right. If you find yourself typing the same thing 20 times, script it. Yeah. One of my colleagues, Andy Grabner, he actually got it from Wilson Marr, who did a what's the future of a performance tester kind of thing, or tips to keep yourself relevant in performance testing,

was find something you do every day and learn how to automate it. Isn't it the guy with the four-day work week who hired some team people in India to basically automate his entire job. Yeah, exactly. I'm just going to have them do it to pay them out of my own salary. Right. And then I'm not working anymore.

That works. I like it. It's really nice to catch up with you. Thank you. Appreciate you guys having me on. How can people tune in? Twitter handle, et cetera. Oh, for me?

Yeah, yeah. Twitter handle is at JPMorgenthau, M-O-R-G-E-N-T-H-A-U. I have a website where I do some blogging, JPMorgenthau.com, and of course, DXC.com. DXC.com. Excellent. Very cool. Thanks for joining us, and thanks for enjoying the rest of the show. Thank you very much, guys.

All right. Bye-bye. Bye.