Latest / Technology Now / Achieving 99.999% uptime for mission critical workloads
Transcript
- Aubrey Lovell (00:09): Hey, everyone and welcome back to Technology Now, a weekly show from Hewlett Packard Enterprise where we take what's happening in the world and explore how it's changing the way organizations are using technology. We're your hosts, Aubrey Lovell... Michael Bird (00:21): ... And Michael Bird. Right, so in this episode we are looking at how technology is enabling as close as possible to 100% uptime for the most mission-critical business operations. We'll be looking at how software and hardware are coming together to ensure the absolute pinnacle of reliability and what it means for our organizations. We'll be asking how you keep tech running, come what may, we'll be finding out what happens when things do go wrong, and we'll be looking at what our organizations can learn from the way some of the most reliable systems in the world are engineered. Aubrey Lovell (00:59): Perfect. Well, if you're the kind of person who needs to know why what's going on in the world matters to your organization, this podcast is for you. If you haven't yet, subscribe to your podcast app of choice so you don't miss out. All right, let's get this party started. Michael Bird (01:12): Let's do it. (01:16): Well, you don't need me to tell you that IT downtime is an extremely expensive business. One recent report, which we've linked to in the show notes, found that IT downtime costs enterprise level organizations $9,000 per minute on average. For higher risk industries such as finance, government, healthcare, media, retail and transportation, their average cost of downtime can be over $5 million per hour. Aubrey Lovell (01:42): It gets even worse when it comes to manufacturing. According to a 2022 report from Siemens, the annual cost of downtime is now $129 million per facility among Fortune Global 500 companies, which is up 65% on the last survey in 2019/2020. (02:01): What can be done to mitigate these risks? How can fault tolerance, security, and data-driven insights make downtime less damaging or eliminate it altogether? Michael Bird (02:10): Well, Aubrey, that's a good question, and to answer those questions and more, I recently caught up with Casey Taylor. She's vice president and general manager of HPE NonStop Solutions. Casey, welcome to the show. Thank you so much for joining us. In simple terms, what is HPE NonStop? Casey Taylor (02:30): Well, HPE NonStop is a extremely high availability, fault-tolerant converged system. The easiest way to think about it, is it's hardware and software rolled into one that provides a full stack solution to our customers with the most mission-critical workloads. Michael Bird (02:50): In simple terms, it's very, very, very, very reliable compared to stuff you can just get off the shelf. Casey Taylor (02:56): It is, exactly. I mean, for example, the IDC talk about levels of availability. It's level four in terms of availability, which means that if there is some kind of fault or issue, it is going to get that process back to where it was before, no interruption, and it will pick up back where it left off seamlessly. We would say five-ninths of availability is officially what we're supposed to say, but the reality is it's fault-tolerant. Michael Bird (03:22): Okay, so super reliable, and is that in the sense of hardware, software, power? Is that all integrated? Casey Taylor (03:28): It's all integrated, but really the secret source for NonStop is the software. What's interesting is that really this is a software business within HPE, but obviously we sell it as a converged system. We will obviously sell it with our hardware, help out Crystal, sell some ProLiant, and we have some networking and storage that goes in. But we actually five years ago, disaggregated the software from the hardware, and so customers can choose to buy the converged system, hardware and software, or they can just buy the software license. (03:58): Ultimately, the hardware obviously is important for getting the job done, but what makes it truly fault-tolerant is the IP that we own, which is the NonStop software. Michael Bird (04:07): What does that do is sort of just managing the different hardware components and saying, okay, well this stick of memory has gone. Let's replace it with this stick of memory very, very quickly. Is it that sort of thing? But very obviously much smarter than I'm explaining it. Casey Taylor (04:22): Essentially. I like to explain things in simple terms as well, and I would say that we could think about it is that it's pairs of processing happening, transaction processing. Effectively what we have is a pair of processors working and there's three of them. You could think about it, six-way fault-tolerant. If there is some kind of fault or issue in the active transaction that's happening, the one that is passive will take over seamlessly because it's working at the same time. It is a shared nothing architecture with basically six-way fault tolerance. Michael Bird (04:54): Shared nothing, so I've heard instances where you have separate power coming into the building for the different systems. Separate power, separate networking, separate lines. Casey Taylor (05:03): Exactly, and obviously customers can choose how they install and deploy, and some of that will impact some of those outside factors, natural disasters and whatnot. That's obviously why we encourage our customers that are super mission-critical to have disaster recovery options. But ultimately failing some kind of natural disaster, which brings down the entire grid or that entire building, the system itself is fault-tolerant. Michael Bird (05:28): What sort of industries need these solutions? I'm guessing finance is probably an obvious one, but are there other ones? Casey Taylor (05:34): Absolutely. I would say probably about 70% of our customer base, hundreds of customers around the world, 70% of them are in the financial services industry. A lot of banks, and if you think about North America for example, about 90% give or take of the credit card transactions that happen on a daily basis, happen on a NonStop system. (05:55): Then we do retail transactions as well, so if you think about some of the major retailers around the world, they're using NonStop systems to give them that fault tolerance that they need because ultimately anytime a system is down and they can't take a payment, they're losing business. Michael Bird (06:13): That's a lot of pressure. Casey Taylor (06:13): A lot of pressure. Michael Bird (06:13): I was going to say. Casey Taylor (06:13): But thankfully, the systems behave very well and there are customers that we've had for decades. I mean we're talking 30 years because this technology is not new. It's been around a while and we have customers that we've had for decades that have never experienced an outage. It's extremely reliable. When I talk to those customers, we had a really interesting session with a bunch of the major banking institutions in the US recently, and really what was echoed around the room is the NonStop systems help them sleep at night. Michael Bird (06:43): Well, that is good, isn't it? That's a good position to be in. You mentioned that NonStop isn't new, you said decades. I mean how long are we talking about? Casey Taylor (06:53): Well, actually this November has been the 50th anniversary of a company that started NonStop, which was Tandem, but we have continued to innovate over those years and so we are continuing to innovate into the future to see how we can keep NonStop going. Michael Bird (07:10): We talked financial services, that's the obvious one. What other industries? Casey Taylor (07:15): Some really interesting industries, some unusual ones, which is really fun. Like a bakery in Asia for example, that just needs really on-time deliveries for some of their baked goods. You wouldn't think that they would need a mission-critical system, but they do. It keeps their business running. Some unusual interesting ones. (07:32): But we have a lot of customers in the manufacturing space and that's actually what I'm here at Discover Barcelona talking about is smart manufacturing with AI with a partner of ours, abatUS. There's some household name automotive companies that turn to HPE NonStop with this partner of ours and they get a full-stack solution to manage their manufacturing operations. Michael Bird (07:53): Right, because manufacturing is again one of those things that if there's downtime, you lose money because you're not creating products that you can get out the door. Casey Taylor (08:00): Exactly. It's ultimately time is money, and if they are not making a vehicle or multiple vehicles, then they're obviously having a massive impact to their production line. We all saw what happened in 2020 and beyond when there was scarcity of resources and people couldn't get vehicles. It's challenging, it holds things up. But specifically for those customers, they really need to have a system that's up. In one of these household name companies that we provide the solution for, they have been running this in all of their factories and more than 16 years and they've never gone down. It's fault-tolerant. Michael Bird (08:39): Is this the concept of the factory of the future at all? Casey Taylor (08:39): Yeah, it's interesting because at HPE we can do a lot for the manufacturing industry. Obviously we have edge connectivity, we have hybrid cloud. From a NonStop perspective, we are partnering with abatUS, and they have a really cool AI-driven manufacturing execution system. It's able to do some very cool predictive things. It takes lots of data input. I mean, there is so much data created in a factory at the edge and they want real-time analytics, but they also want AI that can make real-time decisions to optimize production. (09:10): Those kinds of things can be about using AI to do sequencing. In other words, hey, we have all these custom orders from people around the world that want a blue car, this one wants a red SUV. They need to think about the most efficient sequencing for how they manufacture all of those different vehicles. The system will do that. (09:30): But it does things like, it has sensors that will understand exactly how much torque was used to put a screw in. If they have any quality issues, they can go back and they can have a look and see exactly how much torque was used and therefore they can make that tweak. It's hugely data intensive and we help bring that together with a fault-tolerant system that won't go down and bring AI to help increase quality and productivity. Michael Bird (09:54): It sounds like it's all about those really, really small incremental gains that makes you just a little bit more efficient, which means that you can maybe get things out slightly faster. Is it that sort of thing? Casey Taylor (10:03): Yeah, it is. Antonio talks about AI really being a productivity tool, and you do see that in a smart manufacturing operation, so it's definitely about efficiency gains. The other thing that it does that it helps them with predicting quality issues. Anytime a vehicle is produced, there is sometimes a hundred point quality check that happens, and that's fairly labor-intensive. (10:25): But this solution can take into account different inputs like for example, the day of the week, the temperature outside, the color of the car, and what model of car it is, and then it will spit out a report to say, these are the top 10 or 12 items that we think that as a system we know predictively are going to potentially have issues so that they can look at those first, rather than having to look for a needle in a haystack. Michael Bird (10:52): How are you innovating into the future, 2025? I presume you are not standing still. Casey Taylor (10:59): We are not, no, and you're right. We have to think about ways to build on that storied past over the last five decades and take NonStop into the future. Ultimately it started out as a monolithic converged system, and as I mentioned to you now, we have desegregated that software and so what we want to move it towards is software in the cloud or assess offer, but what does that really mean? It's really about making sure that we are giving our customers choice and flexibility. Aubrey Lovell (11:26): Thanks Michael and Casey. It's absolutely fascinating to hear how all of this tech and innovation comes together, and we'll have more from you in a moment, so don't go anywhere. (11:38): Well, now it's time for Today I Learned, the part of the show where we take a look at something happening in the world that we think you should know about. All right, Michael. You're teed up. Michael Bird (11:48): Oh, well. I'm glad you said teed up. I would say I'm teethed up because it's a slightly quirky one for me this week, Aubrey, because let's talk teeth or rather dentures because researchers in the UK have come up with a novel way of keeping false teeth attached. (12:03): Now usually glue is used to keep them in place or suction based off creating a seal of saliva, but as countless online videos can attest, neither option is foolproof. Add to that, the fact that dental glue isn't especially hygienic and isn't widely available in the developing world causing massive problems for many of their over 300 million users. (12:24): Instead, dental scientists in London have come up with a novel solution replicating the suckers of octopi using 3D printing. It's a field called biomimetics, essentially mimicking natural developments in human manufacturing. Octopus tentacles are covered in dozens of flexible suckers, which create an incredibly strong negative pressure seal on whatever they want to cling to, even the slipperiest of surfaces. (12:52): Even in early trials, the improved denture design was twice as reliable, whilst also being easier and more hygienic to fit. What's more, it's a cheap solution that lasts as long as the teeth do. It's hoped the technology could help millions of people worldwide with their dental hygiene and their nutrition and overall health. Pretty cool. Aubrey Lovell (13:13): I can't decide if I'm fascinated or I'm also freaked out. Michael Bird (13:16): Yeah. Aubrey Lovell (13:17): It's pretty cool though. Michael Bird (13:18): It's a great story to get one's teeth stuck into. Aubrey Lovell (13:22): I see what you did there. Very nice. Very cool. Michael Bird (13:27): Right then, Aubrey, it is now time to return to my chat with Casey Taylor to talk about resilient IT infrastructure. (13:37): Now, whether we've got this on tape, I don't know, but you talked about how some customers want to put their stuff in the cloud, which I find fascinating. Presumably your crown jewels, the stuff that you care about the most, putting it in the public cloud feels really quite scary. Casey Taylor (13:49): Yes, I would say that's true. I mean, I think that there are many customers that say, yeah, these are our crown jewels. They are staying on prem with us, and so they don't want to move them to a public cloud, and you can understand why. (14:00): But there are actually maybe less mission-critical businesses that still want the uptime and that fault tolerance that would be comfortable going into the public cloud. Manufacturing may be that good use case. Manufacturing, we talked about automotive, but any other kind of discrete manufacturing where a customer puts in a specific customized order. (14:20): In Europe, for example, kitchen cabinets is a really big one, and maybe they don't need to worry too much about putting that into the public cloud. There are different levels of customers and we are trying to look at how to broaden the customer base for NonStop. Michael Bird (14:34): What keeps you up at night going into 2025, work-related? Casey Taylor (14:39): Well, not the NonStop system, so that's good. What keeps me up at night really is thinking about how we can stay relevant with our customers. I feel I have a duty to this business. It's very special technology that's been around a long time and it does such a critical thing for our customers. I mean, it really powers the world around us. For me, it's about making sure that we are innovating, we have a roadmap that is going to meet customer needs into the future, and that we can think about how we bring new customers onto the platform. Michael Bird (15:07): I mean point of responsibility to some extent. Like you say, it is powering a good chunk of the transactions happening around the world. There's this legacy, it's been around for so long. Casey Taylor (15:18): It has, and I will say too that the people in the business units, the sales team, and even our customers that come along to our user group meetings. There's such a great community around NonStop business and you really don't have a set of customer base anywhere else that love this product so much. I mean, we get some of the highest CSAT scores because it does what it says on the tin. It just doesn't stop, and so they love that. Michael Bird (15:42): If an organization, if they have an environment that is fundamentally like bog-standard, what's the first step of their journey into figuring out? Casey Taylor (15:49): Well, one of the things that we're trying to do to get new customers onto the platform is get away from talking about the platform itself. As a company, we are trying to shift our focus from selling speeds and feeds and performance stats, and shift to talking to customers that want to focus on business outcomes, productivity, what it can do for them and their business. (16:15): You will not have any downtime. This is going to give you AI driven efficiencies and that full solution. Really it's about pivoting to think about business outcomes, and so that's really how I think customers should be thinking is, what are they trying to achieve? (16:30): One really good example of that is a very large customer that we have that takes reservations. They have a reservation system to book their product and they also take credit card payments. They use NonStop and they have for many years, and they are thinking about how do they get a business model where they could pay per transaction, for example. Because that gives them the cloud economics, but they have the fault tolerance and the peace of mind of having a dedicated system. (16:57): In fact, we've just moved it into a colo that now we are managing it. It's a full GreenLake managed service on NonStop, and so at the moment we have a flex course, so we can say, okay, if you're having a really peak period, then you can flex up and pay for that. But their CIO is saying, okay, that's great, but I'd really like to take it to the next step, which is really predictable pricing per transaction. (17:19): Those are the things that I think customers need to think about is how serious are they about uptime and how much of a business impact does it have if their system goes down? Then how do they want to consume it and what is the business outcome that they're looking for? Michael Bird (17:32): Amazing. Now the last question we ask every guest is why should organizations care? Why should they care about reducing their IT downtime? Why should they care about NonStop? Why should they care? Casey Taylor (17:43): They should care because I think these days the public, governments have a high expectation for uptime. All we need to look at is in the last six, 12 months, some of the really visible major IT outages that have brought down industries. I think that they need to care about uptime to protect their brand reputation and to keep their commitments to their end customers. Michael Bird (18:06): Amazing. Casey, thank you so much for your time. It's really appreciated. Casey Taylor (18:09): Thanks for having me. Michael Bird (18:09): It's been a pleasure talking to you. Casey Taylor (18:11): Likewise. Aubrey Lovell (18:12): Thanks so much, Michael, and it's been great to hear from Casey as well. You can find more on the topics discussed in today's episode in the show notes. Michael Bird (18:22): Well, we are getting towards the end of the show, which means that it's time for this week in history, a look at monumental events in the world of business and technology, which has changed our lives. Aubrey, what was last week's clue? Aubrey Lovell (18:35): The clue last week was it's 1967 and 1982, surgeons put their hearts and souls into these procedures. Michael Bird (18:45): Oh, lovely. Aubrey Lovell (18:45): I think you know where this is going. Michael Bird (18:47): Yes, I think I know. This is one of the easier ones, I'd say. Aubrey Lovell (18:50): Well, it's definitely a pair this week. This week in 1967, Dr. Christiaan Barnard and a team of 20 surgeons performed the first ever human heart transplant on a 54-year-old South African businessman. The man's diseased heart was replaced with a healthy heart of a 25-year-old woman who had died in a car crash. Unfortunately, the patient's body rejected the heart, but he did live for 18 days after the procedure, which has since come a long way. (19:19): Then exactly 15 years later to the day in 1982, doctors in Utah fitted the first permanent artificial heart to a volunteer called Barney Clark. Ahead of the operation, he commented, "It may not work that well for me. I'll do it for the next patient." In the end, he survived for 112 days. Man, what pioneers. Both amazing stories. Michael Bird (19:43): Absolutely. Now, the clue for next week is it's 1954, and this invention was a true gem. Any ideas, Aubrey? Aubrey Lovell (19:51): I don't, but I feel like it has to do with something like a girl's best friend. When I hear a gem, it has to mean diamonds. I don't know. We'll have to see. Michael Bird (20:00): Well, that brings us to the end of Technology Now for this week. A huge thank you to our guest, Casey Taylor, vice president and general manager of HPE NonStop Solutions. To you, of course, thank you so much for joining us. Aubrey Lovell (20:12): Technology Now is hosted by Michael Bird and myself, Aubrey Lovell, and this episode was produced by Sam Datta-Paulin and Alicia Kempson Taylor with production support from Harry Morton, Zoe Anderson, Alison Paisley, and Alyssa Mitri. Our social editorial team is Rebecca Wissinger, Judy Ann Goldman, Katie Guarino, and our social media designers are Alejandra Garcia, Carlos Alberto Suarez, and Ambar Maldonado. Michael Bird (20:35): Technology Now is a Lower Street production for Hewlett-Packard Enterprise, and we'll see you the same time, the same place next week. Cheers.