(upbeat music) >> From Washington, D.C., it's theCUBE, covering ScienceLogic Symposium 2019. Brought to you by ScienceLogic. >> Hi, I'm Stu Miniman, and you're watching theCUBE's exclusive coverage of ScienceLogic Symposium 2019 here at the Ritz Carlton in Washington, D.C. Happy to welcome to the program first-time guest but a long-time customer of ScienceLogic, Karl Fosburg, who's the senior director of systems integration at Hughes. Thanks so much for joining us. >> Thanks for having me. >> Alright, so we're here in D.C., and that's important 'cause first of all, you're based down here, and ScienceLogic is based down here. >> Yup. Bring us back a little bit. You said you'd been a customer a long time as to... maybe give us a little bit of the before picture, if you could. >> Sure, so yeah, we've been a customer for 12 years now, and we picked ScienceLogic for a big list of reasons, actually wrote the RFI itself, and probably 20 pages long. Lots of people came back and gave us responses. ScienceLogic was one of the short-listed candidates that we picked out. We did a bake-off with a couple other vendors, and ScienceLogic was the clear winner. >> All right. So Karl, lets zoom out for a second here >> Okay. and just give us a level set on Hughes, what Hughes is today. You know, I'm familiar with what Hughes was back in the day and there's certain pieces that are no longer there so give us a level set on the company in the business. >> Yeah, sure. So, Hughes is formally known as Hughes Network Systems, were owned by EchoStar Corporation and we're a managed service provider. We have a consumer business where we provide broadband internet to folks that live really out in the countryside and can't get cable, or DSL, or FIOS, things like that. We have about 1.4 million subscribers in our consumer business. We've also launched consumer services in South America, Brazil, Ecuador, Columbia, places like that. Really serving under-served areas for giving them broadband. We also have an enterprise business where we sell to credit card processing, gas and oil, pipelines, fast foods, places like that. >> Okay. So Karl, is it safe to say you use satellites but no longer put them into space? >> We use satellites, that's correct. We contract that out now. Yeah, we are the last remaining Hughes company. >> Yeah. So, service providers are always fascinating to me because we talk about enterprise IT and how fast things are changing. At least for my entire career, when I talk to service providers, change and growth is really just baked into the DNA. >> Yep. I need to move fast. When you talk about scale, it means something very different and living in that complex world, and just give us a little bit about what things are like in 2019 for you. >> Sure, yeah. The scale is always our challenge. Like I like to say, we have sales people too and they're out there selling new products and services constantly. So we needed to be able to grow with those sales. We started out with a couple thousand devices that needed monitor in applications. Now we're up to almost 30 thousand Nox systems that we monitor. Also, we're keeping track of nearly 2 million terminals and the status of them and things like that. So, yeah, scale is super important to us. >> Okay. So, bring us inside, where ScienceLogic fits into your equation. >> Sure. So when we put out our FI to industry years ago, we were trying to replace a whole bunch of different tools. We had other vendor products and things like that. We really wanted to consolidate tools as much as possible into a single platform. Traditional ICNP, SNMP monitoring is how we originally started. Now we have lots and lots of integration with other tools, APM products, different streaming media products. We're integrating more and more with streaming services now in terms of getting data into the platform. So, yeah ... >> Yeah. Karl, I'd love to get your viewpoint. Something that came through to me in the keynote is on the one hand the years like, oh, well AIOps is going to replace things like some of the traditional players here, but then you see onto the stage it's like, oh okay, we're actually going to have integrations with a number of these tools. So yes, there's overlap but it needs to be integrated. How do you look at that as, is this the primary product? Is this a piece of the product? How do data collection between all these various tools go together? Well, that's a great question 'cause that's exactly what we and lots of other folks are grappling with right now. We've got data producers all over the place now, and we're really focused on the data production and high quality data back at the source into a real pub-sub type of architecture of which we believe that ScienceLogic will be both a producer and consumer of that pub-sub architecture, and whether it's the one tool to rule them all or not? Probably not, no ones going to be that, and we've got lots of vendors that purport to be the one tool to rule them all. But really, we're focused on ScienceLogic at this point to be really the focus, especially for our operations folks. We've got 24/7 staff. They use ScienceLogic as their main tool that they go to. So that's really where we want the data to end. That's where we want as much intelligence to end as possible. >> So, I'd be curious... You've been using the tool for a dozen years now. 12 years ago the discussion of data wasn't no where near what it was today. >> Correct. Can you bring us through a little bit of that journey, and you mentioned data a bunch, but how important is that? Where are you in your journey for... There was that maturity model that was put up there, the role of data today, and where do you see it going? >> Well, data is everything today. 12 years ago we were grappling with things like naming conventions and simple firewall rules and whatnot. Those days are long, long past. Now, the data quality and the pipeline is what we're focused on right now 'cause like Dave said in the keynote, "Garbage in, garbage out". We're really really focused on trying to get good quality data by focusing on the source of the data. As opposed to fixing it after it's been moved into whatever platform it ends up in. So we're using proper scheme of management and trying to bake-day the governance into the actual engineered products, and if it's not governed data then you don't get to look at it. And that's really our focus. We're an engineering company at heart so we actually write most of our own software. So we're kind of in control of our own destiny there, and we're really focused on pushing that back because we think the benefits in the long run are going to be worth that investment to get clean data all the way back to the source. >> Yeah. So Karl, one of the big shifts I've seen in the last few years... When you talked about managing and monitoring, I used to as the administrator or controller, used to be able to go and touch all of those pieces. Today, there's more and more some of those pieces I need to manage not just the stuff that's in my environment or my hosted environment, but outside of my environment and doing public clouds. >> Yep. >> Bring us up to speed as to where does Cloud fit? What's your Cloud strategy? >> Sure. We're actually launching some of our first applications in GCP right now. So we're working with our Google partners in this particular case to integrate the data that they can collect natively in their systems, bring it back in as actionable events into ScienceLogic platform, while keeping the vast majority of the data native to their platform. No need to bring back application specific data unless we're actually going to do something with it, or if we need to cross-correlate it with other information. The data sources live in our data centers, not in GCP. So we need to combine it with information we know about, our on-prem equipment, plus the applications running there. So that's the data we'll bring back to cross correlate. >> How do you decide what lives where, and where does ScienceLogic fit in the whole discussion? >> Yeah, that's a good question. What lives where... We kind of go back to license models and cost models. We're pretty good sticklers about focus on doing proper upfront analysis to make sure we don't end up with some six or seven figure bill at the end of the year from a Cloud provider. We also tend to do a lot of stuff on-prem because a lot of our systems have to run in one of our data centers. If you've ever driven past our building you'll see these large large dish's antennas outside. A lot of our equipment has to be within milliseconds or microseconds even of those dishes. So we actually have a large data center presence kind of scattered around the country and around the world. So, we have the compute resources to do it ourselves. >> Yeah, and even I would think edge computing something that plays into what you're doing. What do you see as some of the main challenges as the kind of footprint for what you're doing and things to spread out more? >> Yeah. Keeping, let's say pet projects, and shadow IP projects, keeping them in check is a really big focus right now, and also with DevOps sort of the "I'll do everything, I'm going to be my own IP department" philosophy is a new challenge that we're facing. So integrating with what the DevOps guys are building into our overall monitoring strategy, that's when a new challenge has really creeped up or it last, lets say six months or a year. >> Okay. Is there an intersection between your use of the ScienceLogic in the DevOps team yet? >> Not a big one yet. I think we're still learning DevOps at this point. I consider it a lifestyle change, not really a thing that you go get. So, I think we're still kind of early adoption for DevOps, and really only greenfield projects at this point in time. >> Okay. How about the term of the show is AIOps, so what's your act in the AIOps? Where do things like machine learning and automation fit into your environment? >> Yeah. We actually have quite a few used cases where we really think that machine learning is going to help us a lot. Cross-correlation is a big area for us. We have lots of information, but figuring it out, feeding like the APMs and Cisco ACI software defined networking, and those bits of information all into one product, we've been challenging ScienceLogic on this for quite a while. It's like, okay, you guys know about everything now. Tell us something that we didn't know before, and that's kind of where we're at, and seeing the announcements from this morning was really encouraging that we're finally see the horizon at this point. >> Yeah. If you can, (mumble), but how has ScienceLogic been doing on the roadmap? What helps between ScienceLogic and your vendor ecosystem out there? What more could they be doing to make your life easier? >> Yeah, that's a good question. So, if you would ask me that a year ago I probably wouldn't have been as encouraged as I am today. It was a challenge and they're engineering company, we're an engineering company. Sometimes you have to focus on foundation, and it's not cool, it's not sexy, it's not shiny, but you have to do it. And I think they've been focused a lot on their foundational aspects of the product which will actually enable doing things like machine learning. There's no point in doing machine learning if you have bad data or if you have a platform that doesn't support very very fast queries, and the graph QL database. We think that we're going to use that extensively and through the API, not even through the UIs. So, I think foundation is important. I think they focused on it for the last couple of years. I think we're finally going to start to see the benefits of it. Both single factor sort of machine learning, anomaly detection, but we really want to see it on a cross domain. I want to be able to see in ScienceLogic impacted by in our full stack environment. >> Yeah. I'd expect you probably had some visibility into what was coming up in the Big Ben release. Is there anything that jumped out at you, or that you're ready to use day one? >> The automations, for sure. We'll use that definitely day one. The way they've gone through and really made it a lot easier to use. You don't have to be a python developer anymore to actually get a lot of benefits out of the product. So I can turn that over to some of our junior engineers to actually handle those things, and we get a lot more sophisticated with them now. Primarily we used to focus on, "oh, let's send an email" type of thing. Now we can actually execute back-end actions without having to have a programmer to do it. So that right away we're going to use out of box. >> Okay. And in that forward looking piece, without breaking any visibility you have into their roadmap, what would you like to see more? >> I'd like to see more getting performance data into their real scalable, laterally scalable back end. And that's certainly an area that I'd love to see as much progress as fast as possible on. Also the Pub-Sub subscribing to streams coming out of our Kafka cluster. We want that to be in the product as soon as possible 'cause we really believe that that's where the majority of our data of the future is going to come from. Also, new applications, they come and go. Docker containers spin up, spin down. So the state of something is no longer fixed and we need to be able to integrate with Kubernetes and our open shift platform to be able to know, "Well what should be running right now?" So, those are the things that are on our roadmap that we need out of the product as soon as possible. >> Yeah. So it definitely came to me that ScienceLogic's listening. Are they moving fast enough for you? >> No. No ones ever moved fast enough. So, yeah, they're moving so that's good, but yeah, I could use it today if they had it. >> All right. Karl, last thing, you've been to a few of the ScienceLogic events in the past. You've been to other industry shows, what's special about the show? What brings you and your team to ScienceLogic symposiums? >> One of the things that ScienceLogic does a really good job is they bring a lot of resources here, and actual resources that actually know stuff. It's not just telling me, "Oh, that shiny new object is going to be in the platform at some indeterminate time in the future." It's the actual engineers, people writing code, product managers, things like that. So having access directly to the people who actually do the platform updates and changes is super valuable. The new sensor where we can touch and feel, take attires on new things has been excellent this year. So I think that's probably the thing, just quick access to all the resources. We have a bit of an advantage, we're only 45 minutes up the road. We can come down here as need be to visit their headquarters but having everyone here at one time is great. >> All right. Well Karl Forsberb, really appreciate you sharing your history and experience in future direction as to where things are going on your end. >> All right. >> I'm Stu Miniman. We'll be back with lots more coverage here from ScienceLogic 2019. Thanks for watching theCube. (upbeat music)