>> From the cube studios in Palo Alto in Boston, connecting with thought leaders all around the world. This is a cube conversation. >> Hi, I'm Stu Miniman And we're digging in with the VMware cloud on AWS update, of course, an important solution set we've been talking about for a couple of years. If you see we've done interviews with some of the VMware and AWS executives, we did a deep dive on some of the technology. And now we get to dig in with one of the users of the technology. Of course, the executive talk about the proof of how many customers have been using it. So happy to welcome to the program I have two guests from PLM insurance. First, sitting right next to me on the screen is BJ Gardner, who's the lead system architect. Next to him is Dave Zeigenfuss, who is a senior systems architect. BJ and Dave, thanks so much for joining us. >> Thank you for having us. >> Thank you for having us. All right. So BJ, just for our audience that does not know, Pennsylvania Lumberman's Mutual Insurance Company, give us a little bit about The company 125 years in history, obvious with the name, it's in the insurance business but help us understand, you know, what your business is and what you and Dave do for the organization. >> Sure, so Pennsylvania Lumbermens has been around for under 125 years, we just celebrated the 100 and 25th year. This February Actually, we are commercial insurance company, property Casualty. And we specialize in the wood niche. So we cover everything from lumber yards to auto fleets that have anything to do with moving wood selling wood. So we're pretty niche, we're pretty specific in our brand. And we're a mutual insurer. We're one of the few if not only one, left that does not offer insurance on mutual space >> Alright, and BJ give us a little bit of snapshot from an IT standpoint, obviously, you're using VMware, cause you're here, talk to us about what data centers and cloud usage looks like for PLM. >> So I've been without Pennsylvania Lumbermens for about 15 years, and we were operating in full on-prem with bare metal servers, and 2007-2008, We started with the VMware product set. And since then we've been moving little by little to the cloud. We have many of our core applications are sitting with vendors in the cloud as of right now, we have a small data center in Philadelphia that is an on prem. And then we have, which we'll talk about, why have cloud data center as a service model with the VMware Certified cloud partner called Faction. And then we also now have our disaster recovery as a cloud product. >> Excellent. Since we're talking about the the VMware cloud on AWS bring us inside a little bit,that DR In this case that you're using. That hybrid model, help tease that out. BJ will Start with you. And I'm sure, Dave will have some color to give after you share. >> Sure, I mean, you know, when you're talking about disaster recovery in general, the need to maintain business continuity, while keeping a lean IT staff and with no extended downtime and data loss is just... it's not an option. You can't afford to be down, you can't afford to lose data. So having a cloud service now for disaster recovery, or at least the concept of that helps us more IT shop, in the way of resources that we just don't have on hand on staff. So, that's, pretty much the biggest goal for us, is to maintain business continuity and you know, with our lean staff at the same time. >> Echoing BJ a bit, having an on prem solution and really, to BJ'S point about our lean staff, It made things quite cumbersome for us with maintaining backups replications and such. There was a lot involved. It was very time consuming. So the handoff to utilizing VMware cloud for our disaster purposes really, really helped that benefit our team as a whole. >> All right, you mentioned your partner on this solution is Faction. Help us understand how you made the decision to go down this path. >> So I can give you a quick... a quick rundown here how Faction came to be. so we're located our corporate is in fellow Philadelphia PA. We occupied two floors in an office building. Our data center was on the one floor we were consolidating. And we moved up to just one single floor. So we basically lost the footprint of the data center. So I went out hunting for co location type vendor, and hooked up with Faction. And yeah, so we've been with Faction for since 2015. We've had their, I call it kind of co lo, plus data centers as a service model, since then, since 2015. And we've been with them doing different initiatives here and there over the years and disaster recovery as a service is now one of them. >> Great, Dave, you've maybe supply a little more color on that piece. >> Yeah, sure. Yeah, the use of VMware cloud with Site Recovery Manager. Again, from a technical standpoint, it was second to none as far as the flexibility it gave us to grow our workloads, to maintain them. Recovery point objective was what really sold me. It allowed us to get extremely granular from a business continuity perspective. And, yeah, I'm a fan. I just, I really like VMware cloud with SRM. It's proven to be top notch. >> Yeah, maybe follow up on that, you've been a VMware customer for a number of years, you're familiar with the tooling, and everything else like that. So, how long did a solution like this take to roll out? >> So, I will guess so, absolutely, there was a good portion from when we started, so you have to kind of put it in perspective, because we had a data center in Atlanta, Georgia, that was our data recovery site with action. So we had a two fold project, we were going into a contract year, a renewal year. And Faction pitched, the AWS VMware on AWS service. So we were decommissioning a data center at the same time as we were rolling it out. So I'll just give you the quick timeline. So November of 2018, was basically the contract negotiations. We finalized everything kind of in February of 2019. As far as kicking off the call on how we are going to actually do the project. Work began around April of 2019. Faction went ahead and set up the AWS DDC environment in early May. Faction builds out the environment for the rest of May. June, we did some non disruptive load testing on the environment in AWS. We set up the replication recovery group build out throughout the summer of 2019. And then we had a full sign off in September 17th actually 2019 so I'll just kind of highlight though, in that process, that it took roughly about four months to do the full build out testing and the Atlanta data center decommissioning. >> Okay, and PJ after having done this, we've now got DR as a Service, what are the hero numbers? Have you reduced their cost savings loannes, How do you report up? The success or result of what you've done so far? >> Yeah, so speaking to that, so when we did the contract negotiation, in November of 2018, one of the things we realized when we were pitching the cloud disaster recovery as a service model, we saw roughly about a 20% annual savings in moving to this cloud service. So, a breakdown of what kind of the savings is it's pretty much in Atlanta we have some resource costs because we're running basically on a pillow type environment with with Baxter. and then we had a circuitry call, so we had a point to point line that would run out to, actually to New Jersey and then down to Atlanta. So we that cost as well. So we saved basically, we ripped out the point circuit And we got we offloaded some resource costs. So, like I said about a roughly about a 20% cost savings. >> Alright, so that that's some of the hard figures. Dave, bring us inside a little bit operationally, obviously, there's got to be a little bit of changes to how you manage things, automation is, has been hot for years, but even more so when you talk about cloud environments. So, how is this deployment, changed what the workers are doing and beyond that? >> Well, it's it's simplified things quite a bit, just by the partnership with Faction and in conjunction with VMware cloud for our disaster recovery solution, It's offered many benefits. For one, we had a primary engineer who left the company, we found some benefits to not having to fill that staff resource, so that that was also a positive from a money aspect. But as far as the day to day functioning where we go about doing things up, it really took things off my plate, off the rest of the teams plate And just really, really gave us a peace of mind as it pertains to our, our infrastructure and our data being secured. >> All right, well, I want to give you both the final word. What learnings do you have out of this? Any best practices you'd share? Or there's also some updates coming, taking VMware being able to take advantage of the latest bare metal offering from Amazon? I'll let you choose maybe BJ, we'll start with you and wrap with you Dave as to that those final words that you would share with your peers >> Yeah, I'll certainly start it off. I mean, coming from my perspective as kind of the manager of the team here, our goal as a company, our goal as an IT shop, our goal as an operations team, is to ensure the company's technology needs, will be met after, in the event of a disaster. And that is the key. You want to protect itself, you want to protect the data, you want to protect the customers. So, in the case of the cloud for us, is maintaining business continuity while reducing physical footprint and keeping the IT operations lean, like I had stated before. And one of the most important things and this is not just about disaster recovery, but establishing good partnerships with vendors, is absolutely imperative. Because I don't care how big your shop is, and again, we're on the small side, obviously, but you can't you can't do it alone. So you need really good strong partnerships and good relationships with them. >> And I would say, make sure you are Identify your critical business workloads. Know your environment, absolutely. It's imperative. Get it, you have to plan efficiently. And by all means, test, test test test, you can't test the solution enough. So that's really about all I have. >> All right, Well, David, BJ, thank you so much for joining us appreciate you sharing your your journey along and wish you the best of luck with the solution going forward. >> Thank you. >> Thank you for having us. >> And thank you for joining us for this update VMware cloud on AWS, be sure to check out the cube.net for all the rest of the coverage we have both in the VMware and AWS ecosystems. I'm Stu Miniman And thank you as always for watching the cube. (upbeat music)

>> Announcer: Live from the Bill Graham Auditorium in San Francisco it's theCUBE. Covering Pure Storage Accelerate 2018. Brought to you by Pure Storage. >> Welcome back to theCUBE's coverage of Pure Accelerate 2018. I'm Lisa Martin with Dave Vellante. We're here in San Francisco at the really cool historic Bill Graham Civic Auditorium. We've been here all day talking with lots of great folks, and we're happy to welcome back another Pure customer, Patrick Welch, the network services manager for the Mississippi Department of Revenue. Welcome Patrick. >> Thank you, appreciate it. >> Tell us a little bit about the Department of Revenue. What do you guys do? What kind of information do you collect? >> Okay, we bring in all tax revenue for the state of Mississippi, including vehicle services. We register all the car tags in Mississippi. Income tax, corporate tax, any revenue that's generated in Mississippi comes through us. >> Tax refunds too? Or do you just take, you give? >> We take and give. I have to do it too. (laughing) >> So talk to us about some of the challenges that you had in your environment. I was reading your case study and what you guys are taking in is totalling $7.8 billion a year. As we just identified, some of it's being given back, but what were some of the, what was the infrastructure like to support that before you became a Pure Storage customer? >> We used an internal Mississippi, they're called ITS, they handle all internal infrastructure, that kind of thing. They were using a mixture of Dell, EMC, Compel that type of thing. We use a third party vendor who has an office shelf software package. And they have about 50 to 60 customers in different states and municipalities and countries around the world. In that environment of Dell, EMC, Compel we were about 47th on their list of productive sites. So we were way far down. We were not performing, latency across the board was horrible. The user experience was the worst. If you've ever been on a website and click the button and seen the spinning wheel, we had that in droves. And not just tax payers, but our internal people that worked DOR were not able to work efficiently. We came in and evaluated, and I looked at the infrastructure, and I said my team can do it better. Then when they said, we'll do it better I was like okay now I have to go out and actually do it better. I started researching other companies, and Pure kind of rose to the top of the list. We talked with other customers and partners, kind of how they tackle those type of challenges. We went through a lot of POC process talked with a lot of vendors, things like that. We ended up buying Pure. We are now number three. We went from almost 50 to three. Out of 50, to three. The only two sites that are ahead of us are smaller sites, their transactions aren't nearly as high as ours. >> Okay hang on, how much of that effect could be attributed to the storage infrastructure? Do you have a sense of that? >> 99% >> Really? >> Yeah because before we had, to be fair Pure is all-flash storage, right? And with Compel and EMC or hybrid arrays, at the end of the day, the latency that we saw was due to read and write input being very low. We implemented Pure, through the roof. Storage is not something we would ever look at if we had a problem. We know that that is performing well above capacity. >> Okay I got another follow up. I asked this earlier to another customer, so you're basically comparing an all-flash array to a sort of previous generation hybrid. So it could have been three, four, five, six years old, it could have been 10 years old, so, you had the option obviously of bringing in an all-flash array from the competition. >> We did. >> And you had processes and procedures tied to that, your data protection and you know those products well, but you chose to switch vendors. Why, you could have gotten comparable all-flash, but you chose Pure. Why did you choose that switch and that disruption? What business benefit did that bring you? >> There were several things that led to that. One of the things that we really liked was the proactive support, in terms of every three years they swap out your controller as part of your support and maintenance agreement. Which is huge for us because we don't have a lot of money, our budget is very small for IT, so I can't afford to replace equipment as often as some people can. Their proactive support model, not just in terms of swapping out equipment, but personnel, our sales team that we deal with, our engineering team that we deal with, we're on a personal basis with these people. I have cellphone numbers, I know who to call. We found that out through talking to other customers that, hey you call these guys, they're going to be there for you. Coming from not having that before, we knew that the people we had before, were not going to perform that same level of service. Even if we went to their all-flash product, we were going to have the same support, that we had had before, which was not good. >> And you didn't have that previously because, why? You weren't like a big bank or you just didn't spend enough? >> Because you're a number and in our business, we didn't spend near enough money to be considered. That's a theory of mine, I'm not sure exactly what the actual issue was, but it felt like we were not big enough to get that kind of attention. >> You're the little guy. Pure makes you feel like you're the big guy. >> We think we're doing okay. We have six arrays now, so were not tiny tiny, but we're not also we're not Citibank. But I've never felt any different than a Citibank type customer with Pure Accelerate. >> You're in two years you said? >> A little over two years, yeah. >> You've had enough experience to, you know when you first buy something, you go on Amazon you see the reviews this is great, you wonder if it's still great two years in. >> Patrick: Oh absolutely. >> You would still give a five star rating? >> Oh absolutely, I've done a case study, customers call me and I'm happy to talk about Pure to anybody. I have a lot of friends in state government, I try to head them off from making bad decisions. I'm like if you like your job, you want to keep your job, buy this. >> It's interesting to me, now one of the things that the customers tell us is they love a lot about Pure, but they really like the simplicity. You mentioned Compellent before, Compellent, in its day, was known for simplicity, compared to the old main frame storage. It's interesting to note how technology has changed in whatever 10, 12 years, comments? >> Yeah Compellent was a great product. Back in the day when it came time to evaluate products, they had not performed along the same track as a company like Pure, which consistently innovates its products. If this is again about feeling like the big guy, even though you're a small guy, they keep us in the loop of what they're bringing down the pipe, and it really makes us feel like we're invested in that ecosystem, and we know exactly how they're transforming, how they're going to develop their business going forward. It helps keep us as a happy partner. >> So it's, from what I'm hearing, Patrick, better experience all around, very happy. Did it save you any time? Are you able to now do things differently, add more value to your organization as a result of bringing in Pure? I wonder if you can talk about that. >> Oh absolutely, we spent a good chunk of time troubleshooting issues directly related to storage before whether it was storage creep where we had too much data versus the capacity of the array, or the input output problems in terms of IO, latency those types of issues. We don't see any of that anymore. So that frees our engineers up to work on other problems in the environment. >> What workloads are you running on Flashdeck? >> Mostly production sequel, high sequel workloads mostly. >> You mentioned the dreaded spinning color wheel or whatever kind of computer we're running, and that was affecting not just employees, but also Mississippi citizens. Problem gone? >> The problem is gone from the aspect of our side of things, now this is Mississippi so you still got a lot of rural customers who are still on some dial up internet, so we can't solve that problem for them, but in terms of our side of the fence, we know they're not going to see any latency because of us. We're delivering the application as best you can. Like I said, we're number three in the list of their sites, and we came 44 spots down. >> How quickly in the last couple of years alone? >> Patrick: Immediately, yeah. >> You have to wear a neck brace from the whiplash. >> Yeah we put it in and I'm just crossing my fingers, 'cause if I told them I could do this, and we're 45th, what did we really solve? We didn't solve the problem really, but we came from that high up to all the way down to three, it like felt my team had accomplished something really great. >> And pretty dramatic improvements to your database. I was reading the case study, within the context of your IT transformation, that you improved database transaction performance by as much as 20X. Big, also data reduction rates. So I want to get your perspective on the impact of TCO, and why that's so important for a public agency. >> A lot of things go into TCO. I think user experience is one of those things, downtime for the state. The biggest cost we had was not really something you could see before because our system went down all the time due to not being able to meet the requirements of the taxpayers and the people that work at the Department of Revenue. We don't have that problem anymore. We would spend days of downtime before, that's revenue lost for us. So TCO in that instance is kind of hard to calculate, but I know that the number is big. I know we've saved a lot of time and money. >> Why not just forget all this IT stuff, and throw everything into the cloud. I know as an IT pro, them might be fighting words, but it's talked about in the industry all the time. Why the decision to stay on Pram, and was that discussed? >> We definitely look at the cloud, we definitely have Azure workloads that are in testing right now. Unfortunately it's not just as simple as us saying okay let's go to the cloud, 'cause if it was up to me, with limited funding and that type of thing, I would love to move workloads into the cloud. Where it was applicable. The problem for us is IRS. We have a lot of IRS regulations around cloud. So the core infrastructure that we have, has to remain on premise. There's some things that we can do, but the regulations are a mile long. So we have to make sure that we always stay in compliance with the IRS. That limits our mobility a little bit in the cloud, but we're getting there slowly but surely. I feel like in the next 60 years we'll be there. I joke, but everything we do, we have to go through compliance measures, and we have to make sure we're checking all the boxes. There's one thing you don't want to have, and that's the IRS to write you up for non-compliance. If you're attacked or hit by some vector afterwards, then you're on the hook. You weren't in compliance that's why you were vulnerable. We just have to be very careful, but we're definitely interested. And we'll look into the future with the cloud. >> A lot of talk at this show every show we go to about artificial intelligence, machine intelligence. What do you make of it? How does it apply to your organization? Can you use it? Do you plan on using machine intelligence, whether it's fraud detection or tax evasion, et cetera? What's the state of AI in your world? >> I'd say infancy, but we know that due to the fact that the state hasn't kept up in terms of pay and that type of thing with the private industry. We're going to have to rely on artificial intelligence and automation and things like that to remain ahead of the curve in terms of compliance, performance all the metrics we've talked about. You have to have either a very talented and well paid staff or you're going to have to leverage these types of technologies to stay ahead of the game. >> So you have made some big impacts from an IT transformation perspective we talked about a minute ago. Where are you on this journey of digital transformation? What does that digital transformation mean to the Mississippi Department of Revenue? And what stage would you say you're at? >> We're getting there. Like I said before some of Mississippi is still very rural, for the first time ever, we had more online returns processed than mail. Believe it or not, Mississippians still like to mail their returns in. A lot of that is rural location, internet access that type of thing. We're getting there slowly but surely. I feel like in the next five years, we'll be probably 75% to 80% online refund based. I hope anyway, I hope we're still not at 50%. It's a slow crawl, but we're getting there. We do things a little slower than most people, but we get there eventually. >> You're friendlier down in Mississippi. >> We are definitely, you got to have something. >> You do, so in terms of next steps, you've solved the performance challenges, you're kind of on this road to digital transformation. How have you improved the efficiency of your IT team? >> Say that one more time. >> How have you improved the efficiency within network services? >> I think most of it comes down to not having to worry about the equipment and the environment. We have more time to focus on each other, the tasks we have in front of us. Before it was tackling issues that we knew were related to either vendor or product or storage or server. And now we're focused on expanding the skill set of the current staff. It allows us to leverage things like cloud and automation. We didn't have time to look at that stuff before. So when you ask me where we at with automation, we're still in the infancy because before all we did was fight issues related to previous vendors, previous products, that kind of thing. And this, while it's not a magic bullet, we still have, you're always going to have challenges it frees us up to be able to work on those types of-- >> Dave: Close to firefighting and whack-a-mole. >> That's all we did before. This guy is fighting this problem, he's fighting this one, then they don't get time to learn and grow as employees and as people. >> So automation is big priority, what kind of other fun projects you working on? Or techs that you're researching that get you excited? >> So right now we've deployed both of our major applications using Pure. Our big projects are kind of done. Now we're leveraging towards disaster recovery, modern day DR, BCDR, business continuity that type of thing. How do we recover in case of a disaster? That's kind of where my focus lays right now, to make sure the Department of Revenue, if we are affected by some type of disaster, that we're ready for the taxpayers of Mississippi to come up and running in a sister site and be ready to go. >> Okay that's a combination of infrastructure, probably going to use snapshots, remote replication, but there's also got to be a software component as well. What are you thinking about whether if you don't have a specific vendor product, but just architecturally what are you thinking about? >> So we absolutely right now leverage Zerto with Pure. Which is a very good combination, they work very well together and we have a co-low facility, it's about 200 miles north of us. We'd like to get more geographically diverse as budget frees up and that kind of thing, maybe move out into the Colorados or something like that. But our sister site, all of our data is replicated using Zerto. We're on, I believe, every 15 seconds we're tracking journal history. In the event of a disaster, and we've test fail overs. 'Cause you've got RPO and RTO. Real time objective and recovery point objective. It's important for us to be under 10 minutes, in terms of how quickly we can recover the environment. It's a real time objective. The last time we did a test fail over, we were about four minutes. So our business has completely transformed. Before if we had a disaster, we would be lucky to have data available to us number one and within three to five days. Now we are being able to turn around and operate in another location within minutes. >> And your RPO you said was 15 minutes, did I hear that right? >> Recovery point objectives, that is 15 seconds. Recovery points are every 15 seconds. Our recovery times, the total time it takes us to come back up and running, we hope to be under 10 and we got it around four. Now that depends on a lot of different things. Every situation is not the same. >> Very tight RPO. >> Patrick: Oh yeah, absolutely. >> 'Cause you're moving money, I guess. >> We're moving money. And it's very important that we stay up at all times. Obviously there is going to be a little bit of downtime, but we want to minimize that as much as we can. >> Patrick last question before we wrap here, this is your first time at Pure Storage Accelerate. A whole bunch of announcements this morning, anything that you've heard that excites you for expanding this foundation that you have with Flashtech? >> A lot of the stuff we talked about around automation and that kind of thing. We're definitely interested in how Pure is going to evolve to the cloud because we know you all we be ahead of us I say you all, so you all will be ahead of us whenever we do get ready, and that's another big benefit for us. We know that when we get ready to transition to the cloud, you guys are going to have your ducks in a row, and be ready for us to do that. >> You all as in Pure? We all aren't Pure. >> You know what I meant. >> We're the blue guys. >> It's real exciting to hear about automation, And where they're going with the cloud, and storage as a service and that type of thing is very neat. I love reading about and hearing about that stuff, we can't always be there like I said because of compliance issues, but as we can, we will if it makes sense for us. >> How important is it to you, I was asking a couple of the Pure execs what their thoughts were on staying independent. You see a lot of storage companies get bought, they get consolidated. EMC, 20 plus billion they got acquired. How important is it to you as a customer to have a company like Pure be an independent storage company? >> I mean, it's enormous. I can give you an example. We were a SimpliVity customer so HP bought SimpliVity, our experience before the merger, fantastic. We would give them very high marks in every category. After the merger, not so much. Support dropped off for us after SimpliVity was bought by HP. For us it's huge that Pure is, now that's not to say, we know that this is a business, and that things may happen, but we hope that if they don't stay independent, somebody that has the same level of focus and effort and determination and support keeps that going. >> We hope so too, we love the competition on theCUBE. We love the growth that drives innovation. Pure seems to be leading the way. We talked about this earlier, what they're doing with NVME a lot of good marketing, but still they're throwing down the gauntlet. What they've done with Evergreen. Obviously first with AllFlash or at least early on with AllFlash, so got a leader. >> That's what you worry about too, the Evergreen type things are the things you worry about going away. If they get bought by somebody, is that the first casualty? That's the kind of things that happen to companies when they get bought. We do love the fact that they are independent, but we know it's a business at the end of the day. But hopefully that remains the same. >> Keep that feedback coming, I'm sure they appreciate that. And Patrick thanks so much for stopping by theCUBE and sharing the impact that you guys are making at the Mississippi Department of Revenue. >> Sure, thanks for having me, appreciate it. >> We want to thank you for watching theCUBE, I'm Lisa Martin with Dave Vellante from Pure Accelerate 2018. Stick around we'll be right back with our next guest.