>> Narrator: From Chicago, it's The Cube. Covering Veritas Vision Solution Day 2018. Brought to you by Veritas. >> Welcome back to the Windy City everybody. You're watching The Cube, the leader in live tech coverage. We're goin' out to the events, we extract a signal from the noise. We're here at the Veritas Vision Solution Days in Chicago. We were just a few weeks ago we were at the iconic Tavern on the Green in New York City. We're here at the Palmer House Hotel, beautiful hotel right in downtown Chicago near the lake. It's just an awesome venue, it's great to be here. Arista Thurman III is here, he's the principle computer engineer at the Argonne National Labs. Great to see you, thanks for coming on The Cube. >> Yah, good to be here, thanks. >> So tell the audience about Argonne National Labs. What do you guys all about? >> About science, so we're all about the advancement of science. We do a lot of different experiments from technology for batteries and chemistry. The project we're working on is the advanced photon source, which is a light source that's used to collect data in experiments with a photon source. >> OK, so you're an IT practitioner, >> Arista Thurman: That is correct. >> Serving scientists. >> Arista Thurman: Yes. >> What's that like? Is that like an IT guy serving doctors? Are they kind of particular? >> Arista Thurman: A little bit. >> There's some challenges there, but yah it's great. So basically you have a unique customer base, and they have additional requirements. So, it's not like a normal customer base. They're very smart people. They have a lot of demands and needs, and we do our best to provide all the services they require. >> Yah, so given that they're technical people, they may not be IT people but they have an affinity to technology. First of all, it must be hard to BS them, right? (laughter) >> Arista Thurman: No doubt, no doubt. >> They'd cut through that, so you got to be straight with them. And they're probably pretty demanding, right? I mean, they have limited resources and limited time and limited budgets, and they're probably pounding you pretty hard. Is that the case, or are they more forgiving? >> They're great people to work with, but there can be some challenges. I mean, it's unique in the idea that they work on multiple platforms. So it's from Unix to Linux to Mac. Multiple computers in their offices, multiple data requirements. And a lot of things happen without a lot of process and planning. Some things are ad hoc. So, it puts a little bit of strain sometimes on you to try to make everything happen in the amount of time they have. And everything is There's some challenges with regard to how to get things done in a timely fashion when you don't know what's going to happen with some of these experiments. >> I mean I imagine, right? They can probably deal with a lot of uncertain processes because that's kind of their lives, right? You must have to cobble things together for them to get them a solution sometimes, is that the case? >> We do sometimes. I think it's all about getting enough funding and enough resources to take care of all the different experiments. >> Dave Vellante: A balancing act. >> Yah. >> Dave Vellante: Ya so you look after, compute and storage. >> Arista Thurman: Yes. >> Right, so talk about what's happening generally there and then specifically data protection. >> So in general, my primary focus is Linux. Linus administration, Red Hat Linux. And we've seen a lot of data growth over the last five years and we've got projection for more growth as we are planning for an upgrade. So we're going to change our bmine and make it more efficient. Have a better light source and that's all planned in the next two to three years. And so, there's a lot of extra projects on top of our normal workload. We have a lot of equipment that probably needs to be refreshed. There's resources and with IT and any kind of data management things change. So whatever we're doing today, in the next three years we'll be doing something different because things change with regard to CPU speeds, performance of IO networking, storage requirements. All those things are continually growing exponentially. And when scientists want to do more experiments and they get new resources in, it's going to require more resources for us to maintain and keep them operational at the speeds and performance they want. >> Yah, we do hundreds of events with The Cube. We do about 130 events this year, and a lot of them are so-called "big data" orientation. And when you go to those data oriented events, you hear a lot of, sort of the roots of that. Or at least similarities to the scientific technical computing areas and it's sort of evolved into big data. A lot of the disciplines are similar. So, you're talking about a lot of data here. Sometimes it's really fast data, and there's a lot of variety, presumably, in that data. So how much data are we talking about? Is it huge volumes? Maybe you could describe your data environment. >> Primarily we have things broken up into different areas. So we have some block storage, and that provides a lot of our virtual the back-end for our virtualization environments which is either Microsoft or Red Hat RHV. I would estimate that's somewhere in a petabyte range. And then we also have our NAS file systems which spread across multiple environments providing NFS version three and four and also to Windows clients CIFS and some of the Mac clients also utilize that. And that's at about a little less than a petabyte. We also have high performance computing and that's a couple petabytes, at least. And all those numbers are just estimates because we're constantly growing. >> Any given time it's changing. But you're talking about multiple petabytes. So how do you back up, how do you protect multiple petabytes? >> Well I think it has to, it's all about a balancing act 'cause it's hard to back up everything in that same time window. So we have multiple backup environments providing resources for individual platforms. Like for Windows we'd do something a little different than we'd do for Linux. And we have different retention policies. Some environments need to be retained, retention is three years and some is six months, some three months, and so you have to have a system of migrating your storage to faster discs and then layer off the tape for long term retention. It's a challenge that we're constantly fighting with. >> How do you use Veritas? You're a customer obviously? >> Yah, we've been a Veritas customer for many years and we utilize Veritas in our virtualization environments. They kind of help us out with central platform. We've actually explored other things but the most cost effective thing to us at this point has been Veritas. We utilize them to back up primarily our NAS and our black files, our black file systems that provide most of the virtualization. >> Why Veritas? What is it about them that you have an affinity for? There's a zillion other backup software vendors out there, why Veritas? >> I think we have invested a lot in Veritas over the years. Predating my time at Argonne we've been using Veritas. In my previous career, in Sun Microsystems we also had some kind of relationship with Veritas. So it's easy and I think, like I mentioned earlier, we explored other things but it wasn't cost effective to make that kind of change. And it's been a reliable product. It does require work but it has been a reliable product. >> So, you'd mentioned your Linux, Red Hat Linux. >> Arista Thurman: Yes. >> So you saw this IBM announced it's going to buy Red Hat for 34 billion dollars. What were your thoughts when you heard that news? >> I was like, "Wow, what is going to happen now?" I was like, "How is that going to impact us?" Is it going to change our licensing model? Or is it going to be a good thing, or a bad thing? Right now we just don't really know. We're just kind of waiting and seeing. But it's like, OK, I mean that's a big deal. It is a biggest deal certainly from IBM. Their biggest previous deal was I think Cognos at five billion, so this dwarfs that. The deal of course doesn't close probably till the second half of 2019. So it's going to take a while. But look, IBM is known when it buys software companies, saw this with SPSS, you've seen it with other companies that it buys, it often times will change the pricing model. How do you license Red Hat? Do you have an enterprise license agreement? Do you know offhand? >> We do have an agreement with them. >> Dave Vellante: Lock that in. Lock that long term in now before the deal goes down. >> One of my counterparts is in charge of that part of it. So I'm sure we'll be having that conversation shortly. >> Yah, interesting. Well listen, Arista thanks very much for coming on The Cube, really appreciate your insight. >> Thank you. >> It's great to meet you, all right, you're welcome. Thanks for watching everybody, it's a wrap from Chicago. This has been The Cube, Veritas Vision Days. Check out SiliconAngle.com for all the news. TheCube.net is where you'll find these videos and a lot of others. You'll see where The Cube is next. Wikibon.com for all the research. Thanks for the team here, appreciate your help on the ground. We're out from Chicago, this is Dave Vellante. We'll see ya next time.