PurePerformance - Dynatrace Perform 2017 Tuesday Part 2

Episode Date: February 7, 2017

We continue coverage of Dynatrace Perform 2017 with interviews from John Delfeld of Ixia, our favorite performance geek Andreas Grabner from Dynatrace and strategic partner Ryan Faulk of Faulk Consulting.

It's time for PerfBites. What the f*** is PerfBites? The fourth square meal of the day. Don't bogart the PerfBites. F*** waffles. Microwave ready. Add nutritional value to your brain. Adam Jackaloo.

It's time for PerfBites with your hosts Mark Tomlinson, James Pulley, and Howard Chorney. PerfBites. Whatever. We are live back at Dynatrace 2017 performance conference. Awesome. Sure are. Excellent. So it's me, James Pulley.

I have my gregarious co-host, Mark Tomlinson. Yes, yes. I'm here for you. And Brian Wilson of Dynatrace. And the Beast Boys. Thank you very much. No, you're not. Don't forget that.

It's not that Brian Wilson. I am. I am him, and I'm going to go collect my money. Okay, good. Good enough. But more important than us geeks on the microphone, we have John Delfield from ICSIA. John, welcome to the podcast. Thank you very much.

Two podcasts. Yeah, wow. Yeah, so everyone who listens to the PerfBytes podcast has heard ICSIA mentioned a number of times related to the network performance engineering side of the house, performance testing, device testing, my fondness for IX Chariot. I love that tool, IX Chariot, yes.

So, Ixia is here at the conference, and let's talk about why you're here, your integration with Dynatrace, what's new and going on with Ixia because we like you guys. You're cool. Great.

Yeah, yeah. Thank you very much. Yeah, we've been coming to perform for the last five years. Wow. And the focus that Ixia has here and particularly on our partnership with Dynatrace is around what we call our visibility products. So, of course, Ixia is very well known for its test products. But when we say visibility, what we mean, it's about products that get packet data from the network to monitoring or security tools with Dynatrace DC ROM.

Okay. Got it. Yeah. So different than the exterior days, my friend. Yes. Yes, I know. Yeah. John, you also mentioned-

I just have a fondness for that whole test thing. It's a good one. We age. I'm dating myself. It doesn't mean you're actually dating yourself. He's looking to me because I'm younger than him. What do you young kids say about aging? I'm aging myself. Is that still the term?

I'm aging myself. You are dating yourself. From the older test side of the products, as you guys have expanded into the visibility products, you've made some acquisitions, you just mentioned. Yes. Things are growing. So bring us up to speed on the new world.

I got into this visibility, this packet visibility business, through two acquisitions. One was a company called Inui Systems. Inui. And then the other one was a company called NetOptics. NetOptics, okay. And so those acquisitions were about three years ago. And so that's a significant part of ICSI's business. Yeah.

How are they different? Like how do they help the expansion? I know they're two totally different parts of the puzzle. Because with ICSI's test products, they're primarily used in the QA lab with large network equipment manufacturers and also in the labs of large enterprises and carriers. But they're not necessarily in the production network. The visibility solutions are deployed with, like, DC ROM, APM tools, IDSs. So they're actually in the production network.

And so that's kind of a – and also they're – I would say they're very widely used by enterprises. Okay. Yeah. Now, one thing I can see the acquisitions leveraging is this whole library of stuff from the test and measurement side of the house, particularly for patterns related to DDoS, virus, and things of that nature that are in the network that are used for testing purposes. And now they can carry that right over into production to be able to observe these things happening in the wild and then cross-pollination back to test and measurement. So I can see a lot of synchronization there.

We've done a lot there, so I'll give you an example. ICSI acquired a company called Breaking Point a few years ago. So Breaking Point was security testing. I think they were in your booth last year. Yes, that's right. We were showing it. That's right. That's right.

So Breaking Point has something called Application Threat Intelligence. And so we have a service that we provide for our test customers so that we can always have the latest malware, the latest DDoS attacks. Right, right. And also all the latest applications, the latest DDoS attacks, and also all the latest applications, the latest version of Facebook and all the application traffic, which our customers use to test their products. Right. We kind of turn that inside out.

And now, just like what you were saying, is we now use that to do filtering. So I only want to see Facebook or I don't want to see Facebook. Or I can derive from an existing DDoS pattern that if something shows up, perhaps I can say, hey, this is suspect. Yes, yes, that's right. So we kind of use all those kind of things to help us then get the right traffic to

different security or monitoring tools where they can then do kind of the final analysis. Yeah, and run regressions on any of those simulations to make sure that it's actually doing what it's supposed to be doing. Because this is a weird thing

in the security side of the house. Again, the redheaded stepchildren of the non-functional world, performance and security. A lot of people, you can't see what you don't see. Yes. So you turn on this,

okay, I guess I'm safe now. Remember an old, old product, Zone Alarm? It's a PC. It was on a PC. I think I had it, yeah. It was a customer, yeah. It would be one of those that would show you somebody tried to get in. The first time I'm like,

why is somebody trying to get into my system? But at the enterprise level, I mean, there's so much going on. You put in, you spend a lot of money on security infrastructure. How do I know it's actually doing what I expect it to be doing? Do some customers even do a mix where they have their, obviously, their firewalls and the other IDS stuff running in prod? Do they use some of this equipment along with live traffic? Oh, the testing equipment, you mean, or the visibility products?

Again, sort of combined. The visibility products, for sure. So it wouldn't be uncommon at all where someone would have taps or virtual taps in the network, span traffic, coming to one of our visibility, we call it a network packet broker. And then from that, you would hook up an IDS. You might have an inline IPS. You would have DC ROM. You'd have an APM product. So that's very, you would have DC ROM, you'd have APM.

So that's very, very common where people then mix it all together. Right. And then each one would like a certain different types of traffic. And so you would then use these different filtering algorithms to send the right traffic to the right tool. But, yeah, that is very commonly done. Okay. Yeah, yeah. So it just goes along with what we see in load testing.

People kind of freak out when you say, I'm running load tests in production. Oh, well, actually, you can deal with a mix of synthetic traffic and real traffic to do that. It's a similar kind of thing with the visibility and actually being able to generate that. That's right. And what we find is that there are customers that use, you know, there are very good customers, long-time customers using our test products. And then they kind of get intrusive that way. And then they realize that, hey, I didn't know you were in this other part of the business. And so there's nice synergies there. And there's also technical synergies, as we discussed.

Yeah, roadmap. And, of course, coming to perform, you're going to find performance geeks that live on both sides of the firewall. Yeah. And I should give a shout-out to a couple people I know over at XE as well, Joe Policelli and Warren Stewart. Great. Oh, Warren, yeah. Excellent. Good, good.

Awesome. So you may not be at liberty to talk about sort of roadmap stuff, but do you have any announcements this week? Is there any new kind of fresh happening? There's nothing brand new. Yeah, I guess there's something that we have announced and we're getting a lot of interest in is in monitoring in the cloud. So, you know, right now it's hard to get the packets in the cloud. So you can deploy your favorite packet base to it. Because the customer doesn't own that infrastructure.

Exactly, that's right. Whereas with on-premise, you can plug it all in. And plus in a virtual machine, you don't have a secure enough clock that maybe you can grab everything out of the virtual machine. Yeah, there are a lot of obstacles. And so we've announced it. We're going to be generally available next quarter, but something we call Cloud Lens. Cloud Lens. It's a SaaS service where you can get access to packets in a highly scalable manner so that we can detect on the client side as the load increases

and then on the monitoring side as the tool instances increase to correspondingly grow. Elastic. Bursts. Yeah. Cool. That we can detect that and then make sure that the right packets get to the right tool. And so that's something that's brand new for us. And, again, it's not

released yet. We're in beta. But there's been tremendous interest in it. And so that has been quite exciting. That sounds really cool, actually. Yeah. That's kind of the trickiest part of the cloud, isn't it? Because one of the things I think about when I'm hearing this conversation is, first of all, how much I ignore network in general. I was a load tester. But it's... So I'm used to being...

SYN, ACK, SYN, ACK, SYN, ACK, come on. But I feel, you know, performance's day has come, and I feel with all the DDoS going on, all the IoT, you know, my Criswell prediction for the future will be that network's going to have its day very, very soon in terms of really, really hitting up in the spotlight. Because with everything so connected, it's just not in the spotlight. No one seems to be paying attention to it, at least not in the more visible way. Now, I think we've had something recently which has highlighted some of the network vulnerabilities that are out there.

Right. The distributed denial of services attack against DynDNS. Right. It was in the news. It took out significant amounts of the Internet. A lot of home security. Its name resolution could not take place. But we haven't really saturated the major backbone links.

If you think about that core fiber that's in the ground or under the ocean, they keep cranking it up on the OC side of the house. OC 192, 384. I guess it's in the 7s now. OC dodecahedron. Yeah, exactly. So the speed keeps increasing. Dodeca OC. They can put more colors on the wire and get higher throughput. And so in that sense, I like seeing fiber to

the premise, which is becoming more and more common. Heck, Mark, you've got fiber to your house. Yes, I do. I'm a performance guy. And thanks to the American taxation system, I'm able to expense a lot of that monthly. But I am enjoying my high... So to everyone listening in America, thank you very much for your contribution. You've got Fios. It's funny what we can do when we all pitch in just a little bit. I get fiber to my house. We get roads.

That's nice. I've got a service provider near me, unfortunately doesn't go to my house, that offers gig fiber to the premise. And that bandwidth is just going to keep increasing. The sad part about that is that going along with that is now the marketing professionals are going to be like,

well, sure, I can put that 40 megabyte image on the front page because you have 10 gig to the house. That's all that matters. John, let me ask you something else just back to kind of the cloud concept as well. My session, and I actually noticed it in a couple of sessions, are talking about sort of skills transition into the new world, into be it a DevOps world or these new gadgets. And you've got a lot of potentially like legacy, I want to say legacy,

but like classic network engineering folks that are familiar with how things are built on premise. Like you say, you go to the cloud and you lose that visibility. Is there something, how do you pull people with sort of classic skills to kind of get on board with this new offer? Well, actually, it's been a big push. So we started over a year ago where all our engineering team was a very talented engineering team. They're all learning public cloud, AWS training, all these things.

So it's becoming just ingrained. And it's exciting for us because in addition to some interesting new things, it opens up a lot of great opportunities for all our businesses because how do I test in the cloud? I mean, it's a nice driver for business as well. Right. And so we're seeing great growth on those areas. IoT, we mentioned.

All those things are just, you've got to test it. Yeah. And so that's very good. And I think also what we're seeing is that in addition to kind of like our traditional markets of selling to large network manufacturers and carriers, we have virtualized a lot of our tools. And as a result, the price points have come down. And so we've always had a solid, very strong business with enterprises, you know, the large financials, those guys. But now as we virtualize our test tools, we're seeing a much broader people embracing testing earlier,

even smaller companies. And so they no longer have to worry about the chassis. They can just go grab the AMI and the license key. That's right. That's right. And so that's good. I mean, that has been good as well. So we've put a large emphasis on virtualization and cloud because, you know, the markets are

changing and like everybody, we're investing in it and it's good opportunities. Are you guys familiar with Thomas McGonigal? Our podcast hasn't aired yet, but he's also been, he's working at F5 now and his big thing is preaching

the concept of network as code. So we had in the whole DevOps cycle, there's been hardware or infrastructure as code. He's getting people to bake in some of these network settings, making your firewalls, making your routes, all that kind of stuff, into your Jenkins builds and pushing it out.

So there's definitely a lot of really, really cool interesting things going on on the network side, for sure. So maybe there's some similar things there for Ixia to consider. Especially, here's all the visibility algorithms and configuration you need for this version of the app to roll out. And boom, all of the insight piece, the net optics piece.

That could be cool if you integrated into CI to do that. I wanted to ask one other question. Nui, again, enterprise-wise, they have some of the latency emulation. I'm an old Shunrei HP guy. But, again, enterprise-wise, they have some of the latency emulation. That's right, yeah. I'm an old Shunrei HP guy. Yeah. So Nui's got some new growth in there. There's a big enterprise.

There's still data center relocation consolidation. That's a booming business. They have to recover data center. So the emulation business is still very good. A lot of that core old Nui IP is now in the Ixia products. Right, right. That's right, absolutely. Cool. Yeah.

And I think there's some virtualization in that product line as well, maybe cloud-based emulations and things like that. Yeah, because the latencies are less controllable. Yeah. And so you have to test it more. Right. So, yeah, absolutely. Different scenarios.

Absolutely. Cool. That's nice to know. Yeah. Awesome. All right. John, thanks very much for joining us, man. This is great to talk to you.

Is there anything else that you want to chat about? We could cover? No. Did we forget to ask you anything? No, no. I think that's, I mean, the, you know, the answer is to be here and perform. And so, you know, thanks to the good folks at Dynatrace. It's been very good. Yeah.

And, yeah, I enjoyed chatting with you guys. No, thanks for coming. There is one additional item. I noticed on your panel you've got a network tap. So have you got a new line of network taps? Oh, yeah, yeah. See, there's always something else. Yeah, net optics.

So we kind of always kind of preach, when you're getting traffic off the network, you can use a span port and you can use a tap. And so a lot of times it's like, well, a span port's free. What's the problem? But as I can see by your faces, you guys realize that, and we agree, that span ports, you get packet loss. It's really not the ideal way to do it.

You still, how many ASICs you got sitting behind there? Yeah, exactly. I mean, if you know the model, it's like, okay, I can span these two because I'm on different ASICs, but you span on the same ASIC, some of those chips can't do it. And you're going to have drops, you're going to tax the switch. So the taps, we come up with different types of taps, smaller taps. So we're seeing nice, you know, that continues to be a very nice business. And the other thing, also, I'm glad you kind of reminded me about what's new.

The other thing that's new is that these network packet brokers, which take input from the taps and spans, then groom, filter, de-dupe, and then send it to the monitoring tool. Oh, yeah, yeah. That's, we do a lot with, that's a very big part of the business. And the trend in that is more intelligence. So what you're seeing is that instead of just doing filtering on layer 2, 3, and 4, you can do layer 7 filtering.

So maybe you say, I don't want to store the YouTube. So you get rid of it. Or I only care about Citrix or SAP, and so you only send that. So it's adding more intelligence, layer 7 intelligence to the filtering. And then in addition to that, what you're seeing, the trend is things like, for example, SSL decryption. We can now build into these network packet brokers. We can do decryption and generate net flow. So these are kind of the – it's kind of like –

So you can recover the entire application infrastructure by which nodes are speaking to which other nodes. Right. And map it out dynamically. And map it out, even to the point of geographies and things like that. So you can say, I only want to send this traffic from the Dayton branch office, only the Citrus traffic. And you can be very, very precise. And as a result, the more precise you are, the more efficient you're using.

And the Layer 7 definition also just makes it more accessible to less geeky network people. It's like, I don't know. I can look at HTTP. I don't have to look at this frame. So again, it makes it more accessible to a wider range of engineers, maybe even newcomers. It makes it easier to get on board. And then they can dig lower

as they get more sophisticated. I think that is the trend, is that people are very busy, things are getting more complicated, so simplicity is king. So how much of this filtering actually takes place at the tap versus at the packet broker level? Today, most of it, not all, but a lot of it is done at the packet broker. Okay. But we are seeing, you know, there's a trend where you also,

because you could, yeah, but a lot of the filters are very, like, passive. You know, you got, it's a simple device. You got a prism and two watches. So if you go that route, that's its beauty. It doesn't have much intelligence. They're very affordable, and they're extremely reliable. And in that situation, you think about putting it on a separate component like a packet broker. What's the urgency of knowing that information in the greater visibility? Am I first line of, like an IPS or an IDS, where it's like I have to know now,

and so I can't put a lot of overhead? Or some of those things could be run with lower priority. The greater context of using the information, I'm thinking. That's right. That's really pretty cool. All we have to do now is layer Davis on top of this for the artificial intelligence side of the house. Had to bring in Davis. Had to bring it in.

No, that would be cool. Is that Davis' real voice that I heard today? Well, it's a computer. Does it have a real voice? But that is the voice. Nice voice. How come I didn't get chosen as the voice of Davis? You know, that would be pretty interesting.

Hello. You know, I bet you if you put Davis on top of that, you could probably avoid a lumber 84 kind of situation. Yeah, we could set filters dynamically based upon what's interesting in the network. Yeah. I just have to get lumber in again. We're going to try every time. 84 lumber. Yes.

84 lumber. You ever shop at 84 lumber? No. Yeah, no, it's fine. A lot of people have it, surprisingly. But the ad was kind of intriguing. I didn't go to see the ad. I actually really want to just call up the guys with the Lode Runner license at Home Depot

and see if they actually just DDoSed their competitor. No, no, we don't want to talk about that. Oh, not that we know anything. There's a guy that knows a guy. Awesome, John. Thanks very much for joining us. Thank you so much. Thank you.

Enjoy the show. Thank you. Look who has just walked by in kind of a drive-by event. Mr. Hotshot, I'm a big star. I can only spend a minute or two here with you all. Andy Grabner. Eric Grabner.

Well, first of all, if you tell people on the air, look, where would they look? Look. And see me. Directly in front of me is Eric Grabner. Listen to this strange Austrian voice. I know, I know. I'm here. Hello, Andy.

How's the day going so far? How did the next session go down there? I mean, your session was amazing. Did you like it? I loved it. I saw it was perfectly. I just needed 15 more minutes. I know.

I know. Honestly, the love between these two makes me really uncomfortable. I'm comfortable with it. You're comfortable with it? With what? But you know, I'm from the South. We're not comfortable with any of that stuff. The love?

Well, there's some special brotherly love here. We sway. Left and right. Actually, we need to do some salsa now. Yes. Oh, no. You can teach me. No, but your session

was amazing. I know we pressured you because we cut off 15 minutes of your original schedule. That was Andy's idea, too. I'm a nice guy. Of course, because I told

the people on the main stage, please run over by 30 minutes. Yeah, exactly. That's what I did. Thanks for that. Yeah, I know. I know. It happens.

No, but it was very well received. The only thing that didn't work out, what you said, you know, you made the reference to a lot of, you know, more female. Yes. More female in engineering, performance engineering, and they typically either have colored hair or a tattoo. And then Anita came on stage right after you, and she said, I don't have a tattoo.

I don't have colored hair. She needs a tattoo. That's what it is. Let's go get it out. It's Vegas. You can do everything in Vegas. They have great tattoo shops here. Here we go. Okay.

Andy, this is your first time stopping by, really, since the start of the show. Give us your top two items that stick out from today so far. Other than Mark's presentation, it could be the introduction. It could be your experience

right here. It could be, hey right here. What if my session was the most interesting? You ran into someone extremely dynamic and they changed your life. Top two items. Go. So I spent, I was in the keynotes. Were you in the keynotes in the morning? Did you get to see it? I was in the broadcast.

I think I was very impressed and heads off from somebody that is doing two live demos on main stage. Alois did a demo on Davis. So our artificial intelligence, our assistant, he did a demo asking Davis, you know, what's going on? How is the system doing? And do we need to freak out because we just made a deployment that didn't work? So that was nice.

And they fixed it. And the thing is, that was the second part of the demo. We have an auto-healing feature in there. So basically he had a website and he crashed it. And then Davis said, well, it's crashed, but guess what? I just fixed it for you because I rolled back. So he demoed that live. Rolled back or rolled forward?

He rolled back in this case. It doesn't matter. He could do either. That's true. He was a little bit under stress I believe. That's why he wrote the password four times wrong when he tried to log into Ansible Tower to run his scripts. But eventually the crowd applauded when he fixed it finally. Because somebody had already, he had a simpler password.

Yeah. And already since last night, he got hacked into on his Amazon instance. Yeah. So that's why. That's nice. Yeah. He wanted to make the password easy. And within 30 minutes, he said his EC2 instance was hacked.

That's why he came up with a more complex password, which was a little hard for him to remember. Which took Davis a longer time to guess the password. Oh, probably. Yeah. You know what? I use phonetic passwords. Oh.

Like, uh, uh, Coney's Madam. Coney's Madam? What matter? Coney's Madam. Now, here's the scoop. Coney's Madam. Yeah. I could just say, Andy, my password is Coney's madam. Now, if you and I have worked out the idea

of the elite, the elite, like E's are threes, O's are zeros, the fifth character is always... You come up with a cipher, but they're phonetically pronounceable, so then right in front of a customer, you can just say, hey, what's the password

on system three? Oh, it's Coney's madam. Yeah, but then I know I'm going to probably do C0 and 3Z. But you don't know where the dollar sign is always going to fall, where the X line is. I know, but you just eliminated a lot of my work. Anyhow. Now, this is not security bytes. Maybe we need to start a security podcast. So we have Davis.

So Davis is really cool. The other thing that was really cool, Visual Complete. So we introduced the new metric, Visual Complete. I missed that part. What is that? Visual Complete is just a new way to measure when is the web page actually really visually complete and not DOM-ready, document-ready, or fully loaded. Render.

So we basically take Visually Complete as a timestamp, let's say two seconds. But then we also look at the speed index and basically say, how long does it take to actually really load the page? What's the speed index? How long does it take for the page to be visually complete? Because it can be that you have four seconds of nothing, then everything is here. Then you have this curve.

I know I'm drawing a curve now. Like this. Yes, so Andy is doing a hand-drawn diagram of the page compositing CSS process and final rendering of the page. And it's like, okay, all of this process could take four seconds, and then boom, it shows up. Exactly, and the idea is

if you draw this line, right, if you draw the line, and then you can it shows up. Exactly, and the idea is if you draw this line, right, if you draw the line, and then you can basically figure out if you – I'm not a mathematician, but basically you can figure out this – what's it called? You can calculate the – Arc? Not the arc, the surface. No, what's the surface called?

Surface area. Surface area. I'm not – I'm so bad at math, whatever. It's just the area. It's the area. You can basically calculate that. So if it is like white, white, white, white, white, and then goes up, but you actually have a lot of white space in here.

But if it goes like this, you see a lot of content right up front, but then it slowly loads. You have a smaller surface area. So you're looking at an efficiency curve. That's what it is. It's the efficiency curve loading the page. The area of an efficiency curve. That's what it is. It's the efficiency curve loading the page. The area of an efficiency curve.

And as the curve shifts from right to left, the page rendering speed to the end user is effectively improving. Exactly. That's what we put into an index. We have an easy number now where we can say, hey, what is it? Is it loading? Is it white all the time until the end, and it's there after three seconds? Or you see already 80% in the beginning, and then it takes a while until it's visually complete. So this is just going back for our users

in terms of perceived render time, which is the metric I was looking for before. This seems to be quite an improvement on it because it's just not a number. It's now taken into account how that is. How the page is loaded. We have the full film strip. What do you know from web page test. We have the full film strip and look at that. And we show you

in Dynatrace now every single stage when the page is loaded, all the different events, including visual complete, how we ended up at visual complete, and show you the speed index, which is basically the calculation I tried to explain and drawing in mid-air. So let's draw it back to the customer. What's driving this visually complete statistic?

What is driving? Versus render complete? Versus, hey, the page loads and it takes nine seconds. Well, single page apps, right? Souders, man. If you load an HTML page, it's loaded within 100 milliseconds, but it's blank because afterwards it loads all the stuff. So it takes maybe five seconds.

And the dynamic responsive design. Exactly. And so from an ops monitor perspective, you know, it's done in 100 milliseconds. But users complain it takes five seconds, ten seconds. With visual complete, you know exactly when it is visually complete. That's the key thing. And also it works. I think you easily see the difference between the different devices.

So this goes along with this roar in the market of, I need to measure rendering. Exactly. I need to measure what my end user really sees. And it's not cheap. So we met with Ixia here. We were talking to John about a lot of classic network analysis, SYNAC activity, breakdown of SYNAC activity.

You can't interpret what's happening visually from any of that. So you combine what you know at layer two and layer three and then start taking a look at what's happening visually. Exactly. So this was the second thing that I liked. And the third thing, I know you asked me for two. I want to give you three. The third one was my session.

The third one was yours, obviously. This is a bonus item. It's a bonus item. From Andy. It's not a. This is a bonus item. It's a bonus item. From Andy. It's not a **** item, but a bonus item. What? You're live on the air, you know. Whoa, did you just, what?

I know, I know. Come on. Come on, let's get back on topic, Andy. The last session I attended was the DevOps transformation within DynamoTrace, how we went from six months waterfall releases to one hour code deployments. Yeah, thanks. And it was mainly a cultural change, but not only development side, but also business side, because the business was demanding it, but also was impacted by it later on

because we were speeding up so fast and cranking out new releases, and business was like, ooh, we're changing all the time. I would imagine we're going to have that as a topic of enormous discussion next week in New Zealand and the following week in Australia. Accelerate innovation. That's what it is.

And we're doing a fist bump now for people that don't see. A visual fist bump. And it's an Obama fist bump because it's got the finger flare. So since you were in that, so I met Jeppe. Jeppe is it? Jeppe Kaur? Yeah, yeah, yeah. Jeppe, he was over here.

But now you need to bring Anita by because I will. Yeah, definitely. I will find her. We had her on our show talking about the practical part of that component. And she needs to get a tattoo. Exactly.

And the tattoo for Anita, a UFO for Yeppe. If you see him again, I owe him a UFO. Yeah, he was looking for you. Yep. I need to run now because of the session to run.

We should get some lunch before they pack it up. Thank you very much for your impressions. Thank you. And with that, I think we're going to get some lunch before they pack it up. Thank you very much for your impressions. Thank you. And with that, we're going to drop off and have a little bite to eat. We'll be back on the air certainly at 3 p.m., which is a little bit about two hours from now, local. And we may hop on before then if we encounter a really dynamic and interesting character.

I'm going to go talk to Pivotal. All right. Thank you. Bye. Welcome back. This is James Pulley, live from Dynatrace Perform 2017. And I'm sitting here with Ryan Folk of Folk Consulting. Hey, James.

Good to see you again, bud. So, Ryan, you and I are old performance hacks. Sure. Not to mention the fact we have similar haircuts. We've had similar haircuts in the past. Yes. I know why I like Dynatrace stuff. Sure.

Same reason why I like Splunk stuff. Sure. And testing tools thatace stuff. Sure. Same reason why I like Splunk stuff. Sure. And testing tools that destroy servers. Sure. So I want you to tell me why you're here at Dynatrace Perform 2017. What has been your impression so far? Sure. And, of course, we're going to get to some cool Las Vegas stories.

Sure. Okay. So this is my first Perform with full consulting's presence, if you will. We've been a partner of Dynatrace for three years. Okay. You know, obviously with, you know, my company headquartered in Michigan. Dynatrace, you know, in Detroit, of course, right? Big presence in Detroit.

And you put on your bulletproof jacket. No, come on, dude. It's not. Downtown Detroit is lovely. I spent the last weekend there myself. My wife and I, restaurant and bar hopping. It was wonderful. You should come.

Come try that out. And you're alive still today. And I am here. Jeez Louise. It's not Chicago. Come on. No. Oh, okay.

I'll grant you that. It's not Chicago. So, yeah, this is our first perform. We have three people here, two engineers, specifically around the APM solution. You want to call them out? Well, so I'll call one out because I know he won't mind, but Tom Pavik, who I think you met last night at the kickoff, right? So, yeah, and Tom's a former Dynatrace person. He's also dealt with some of the other competitor solutions, of course, to Dynatrace

and the APM, DPM, if you want to call it that, arena. But his heart lieth within Dynatrace to some degree, of course. And so we're really here, just like with anything, right? It's a great networking opportunity. You get to see a lot of the product roadmaps and things that are very important. Obviously, when you're speaking to customers, you don't want to talk about what's out. You want to talk about what's coming and how that ultimately would affect their businesses. And then ultimately, what levels of maturity can you not only bring with the current solution, but where those solutions are going?

And I was talking with another partner last night, and they legitimately asked me this question. How many people are really doing some type of DevOps model? Your partner in crime there, Mark, had a nice presentation today on what DevOps really means and how did we really get to this whole DevOps thing. Versus DevOps. DevOps, yeah. That's right.

You know, he said if you're really doing DevOps, one of the slides that he brought out today, he said one of the things that if you're really truly doing DevOps, it probably means that you're in an office that has microbeers, people have many tattoos and lots of piercings. Yes. Which is why they don't trust me. Yeah. So, you know, I, I, I don't agree with, I, so it's funny because, you know, a lot of, uh, bleeding edge companies that are out there that, um, work in a lot of incubators. Like we, we do, we, we meet with a fair amount of customers that work

in, um, what I would mature incubator type of technologies. I can think of about a dozen in Atlanta, for instance. I can think of four in Chicago that we've worked with or talked to that are employing and successfully really doing a DevOps. Whereas you see a lot of the major Fortune 500 companies really kind of struggle with how they implement it across what ultimately be a vast IT organization, right? Yes. That being said, as we spoke about last night, there are some customers that I know that are doing DevOps. And they are integrating multiple tools that allow them to get to what I would consider to be continuous delivery in its truest sense in what you see in these conferences,

whether it be product-specific conferences like Performiz, of course, to Dynatrace, but also some of the ones that are just holistic, like SDP Econ, for instance, and some of the other testing or ERP. DevOps world. Yes, exactly, right? Yeah. So I'm really interested to see, you know, who's really adopting some of these terminologies and who's just using Jenkins to use Jenkins, right, or something along those lines.

Oh, you mean resume development model. I hate to say it, but it does exist for us. Well well for sure i i'd like to give it a little bit more credit than that because you know there are some there are you know once again we're we're we're uh for instance uh i had a i had a company asked me the other day how many companies do you know that are really using containers and i said i could name off uh six right now. Yeah. And they said, well, we still haven't a clue how to do that. And I'm like, okay, that's not bad. You know, you have to have the business case, right? And is it, you know, at the end of the day, it comes down to, most of the time, it comes down to cost.

Are we seeing the ultimate cost savings? Do we have a two-, three-year plan where we see that ramp down in cost? Are we seeing the ultimate cost savings? Do we have a two, three-year plan where we see that ramp down in cost? Because ultimately, at the end of the day, very similar once again to Mark's presentation today, a lot of this is based on can we do it cheaper, better, faster? And cheaper is related to cost. That's why it's the first part of cheaper, better, faster, right? Yes. So once again, I kind of spun around here a little bit on you a little bit, but we're here at Dynatrace to, once again, get a lot of the roadmaps. I'm here to meet with a lot

of the leadership here for our ability to support Dynatrace as a partner. We were not a silver or gold sponsor this year, but I would suspect we will be next year. And yeah, that's why we're here. So outside of roadmaps, give me the top two things that come to mind when I say Dynatrace Perform 2017. Boom, boom. These things jump to the top. For me? Yes. Oh, so it was basically the meetings that I had with their leadership today and some current business that we have and how we help them grow that business, when you come to Dynatrace or you go to other, other conferences

where you're partners with the, you know, a company like mine is partners with those companies in, in the partner or the channel type of meetings, they want to say, how do we help you grow your business? It's not real the majority of the time, but as a partner or channel manager, you have to say that because you need your partners to help grow your ecosystem. Most partners would love to say, hey, through this company, I was able to grow my company by a revenue of X. But sometimes that's not real. The opportunities come when they come and you've just got to be in the right position to make them. But at the same time, specific companies like Dynatrace, you know, when they say, you know, we want to go to market with these things, sometimes they have the internal staff to do that.

And sometimes they really truly do need to rely on their partners. And that's what I'm here to try to see some of that roadmap and then ultimately help deliver on some of those services. Okay. So you've got the number one item, which is really VAR focus, which is your vision as head of full consulting. Any technology thing that jumps out at you and it's like, hey, that's cool. I got to have that. So from a that's cool perspective, I think Dynatrace is a that's cool product. Okay. From a profileriler diagnostics view of the, of the it infrastructure and, and, you know, the, the, the data, the statistics that you can get out, whether it's application monitoring,

the data center ROM or synthetic monitoring, things of that nature in terms of their, your, their core competencies of what you can be, for instance, certified in. Right. Yeah. Um, those, you know, just the, uh, uh uh i hate to use the term best of breed because you know there's many different ways to skin a cat as you and i both know um from the performance testing space for instance right there's many different ways to do many different things yes right um but uh just the ability the access to information and how it companies don't realize what they really have access to with the investments that they make. And helping them realize that is really what moves my needle, if you will.

So I'm going to put you on the spot on Davis. So Davis, the artificial intelligence agent, watching for performance issues in production. We're old timetime performance guys. You used to have to hire us to find some of these issues. What do you think Davis is? What are the implications of Davis, one, from a customer-facing perspective, and two, from a channel perspective? Boy, that's a great question. So why don't we elaborate a little bit more on the customer perspective, right? Because the majority of people that are probably listening to this podcast are customers out there, for instance, right? Sure. So let's dive into that, I guess, right?

So this gets back to ultimately my... So there are other products out there from different competitors that sort of do what Davis essentially does, right? But in my opinion, tell me if I'm going down the path that you're ultimately asking the question for, right? Okay. So some of the things that are difficult for companies to identify, and we're obviously dealing with this right now between you and I, I'm another customer, of course, that does not own the Dynatrace model, but one of the things that we were trying to determine was identifying performance requirements.

Yes, and ultimately pattern analysis comes into play. That's right. And artificial intelligence agents are really, really wired for pattern analysis. Right. And hopefully identify new patterns as they develop. Hey, we're identifying this thing. Right. And for the most part, you're hiring guys like us or firms like us today because we have an extensive pattern library of what we've seen and we can bring to the table and what we can do.

And it's absolutely amazing how many customers have no idea what's happening. They think, okay, we're going to call our IT ops teams and say, what's our hit ratios on this infrastructure? But they absolutely have no idea what that fundamentally means across the entire infrastructure itself. And whether that hit ratio is appropriate. And quite frankly, what is that hit ratio over a period of time? Yeah. Right. And so identifying those requirements is where the sexiness, if you will, of that is really there. So once again, it's all about data. It's leveraging the data that you have, you know, feature, use feature, feature, don't use feature. I say that

all the time in front of customers. You bought this stuff and there are features that you obviously would use. But there's some things that just just, you know, in some cases don't make sense to what fits for, you know, every customer is different, right? Just because you have the use case for one customer doesn't mean that it's going to be one for one to another customer. Correct. Right. So some of that's identifying that thing. But once again, it's the awesomeness, if you will, for that is just the access to that data. Now, I see an opportunity on the VAR side for Davis as well. Sure.

Now, some might look at this and say, hey, this is going to dramatically impact my ability to deliver services because, well, Davis is going to be doing it and I'm not going to be doing it. I see that we can educate Davis with our own patterns that we're aware of. Maybe we can license it downstream through Davis as an identification method, and then Davis can carry it forward wherever it's used. We get a small couple pennies every time the analytical piece is used. The iTunes model. The iTunes or drink umbrella model, however it works. So I do see an opportunity there to recapture a lot of lore that's in the field and centralize it. Has Dynatrace fully investigated that at this point?

I'm not sure. I think it's not clear. I think sometimes when you have some of the products they have, it takes people, whether it's employees that are using the solution at these customers that are buying the product and using the product, and some of the VAR partners coming in to help people mature that product in their organization. There are all kinds of methodologies that can come about. Once again, just because one customer does a use case doesn't mean that those use cases are the same or they're the most mature. Correct. Yeah. Well, excellent.

Well, is there anything you're looking forward to through the rest of the conference? You have particular items that you've picked out to go see other than roadmap? Because everyone says roadmap. No, for sure. And that's really, like I said once again, that's the most important thing for me in most of these conferences. I think just once again, most of my team is going to consider that are here. They're going to make sure that they're up on their certs.

So we checked that box earlier today. And then I really enjoy walking around and just talking to all the partners that are here, especially the silver and the golds that have some basically use cases that they're ultimately trying to get people in front of. Those types of things, as you and I both know, we run into opportunities all the time where a customer has a need. And if it marries up very well with, you know, I'm not adverse at all with working with partners if it basically fundamentally gives the best course of action for a specific customer. And you and I work on many of those opportunities all the time.

That's true. I've used your team. You use my team. That's right. It've used your team. You use my team. That's right. It's a nice cross-pollinating relationship. So that's really, you know, as much as I really appreciated Mark's presentation today, I get more knowledge from those types of discussions than I do most of the events that are obviously preplanned. Okay.

Excellent. Yeah. Well, any closing thoughts before we... Well, for everyone that's here, I hope you enjoy the rest of the conference. I, myself, have to catch a noon flight to L.A. tomorrow. But, yeah, once again, the people that are here, they'll be here all week. And, you know, listen, I'm glad I ran into you. Yeah.

So, ladies and gentlemen, Mr. Ryan Folk. Thank you. Of Folk Consulting. Thank you for sitting down with us and giving us your impressions of the conference. Yeah, James, thank you for your time. And I'm entered to win a drone now. Yes, I need your business card for the drone drawing. Of course.

That's fine. Mark, thank you. Thank you, everyone. We'll be back online in 30 minutes with Heinrich Rexed of Neotis. Excuse me, I've just been updated. 3.30. This just in. This just in.

Excuse me. This just in. The Heinrich Rexed interview has been updated to 3.30 p.m. local time. Thank you very much for tuning in for this update. Have a good afternoon, everyone. ... ...

... ... ... Bye.