Hyperconvergence? What’s That?

At King Price Insurance, we’re no stranger to innovation. We’re revolutionizing automobile insurance through a one-of-a-kind business model: as your car depreciates, as all cars do, your insurance premiums decrease. After all, why should you pay the same rate month over month to insure a car that’s value is constantly decreasing?

This mindset of disruption is evident in every facet of our business. We don’t like doing things one way just because that’s the way it’s always been done. We’re constantly optimizing and rethinking every process, every technology to see how we can increase value to our customers while decreasing time, effort, and cost for the business.

As the Virtualization and Storage Administrator at King Price Insurance, I’m always looking for ways to speed up our DR, backups, and every other IT infrastructure function while reducing the administration burden and cost. So, when I became fed up with the time it was taking us to perform normal jobs on SQL—indexing and routine operations were taking us up to 14 hours to complete—I started brainstorming ways to reduce this effort and maximize our efficiency.

King Price was growing every day, and the size of the data was starting to cause our banking and collection files jobs, which run on a daily basis, to bleed into our normal working hours. The window of opportunity for a collection job is only so big, so we were facing time constraints on when we could run these jobs. Our growth was literally becoming an IT nightmare. 

We were spending our entire day just trying to keep our environment up and running and stable. Being in South Africa, most of our peers were running on similar systems as us, so at the time, we didn’t realize what other options we had. But we knew we had to make a change.

Hyperconvergence? What’s that?

In business, timing is everything, and I have to credit our success with our eventual vendor with being in the right place at the right time. We were hunting for a new vendor, and as you do, were asking everyone we knew who they used and what they thought of them. To be honest, hyperconvergence technology wasn’t even on our radar. We had been running a traditional SAN stack that utilized multiple solutions from many different vendors. With the capacity and processing ability of these disparate environments, it just couldn’t handle the large workload.  

As luck would have it, we were introduced to a couple of guys at HPE SimpliVity. This was the first time we heard of hyperconvergence, and to be honest, we couldn’t believe the simplicity and speed they said we could accomplish with their technology. They claimed that with HPE SimpliVity we could restore between data centers within seconds. 

In addition to functionality, we did a quite extensive exercise to compare the cost of hyperconvergence with traditional SAN, and even without considering all the bells and whistles of HPE SimpliVity (like deduplication), hyperconvergence came in at or even cheaper than the other options. It sounded like an offer we couldn’t pass up, so 18 months ago, we made the switch.

Life After HPE SimpliVity

The best thing about running our infrastructure on HPE SimpliVity? We can actually do what we’re supposed to do now. There’s never free time in IT, but with the speed of hyperconvergence, the worries and time to do just everyday activities like restore and backup just disappeared. Two years ago, we literally didn’t have time to innovate in this area of our business. But now our guys have the headspace to think more strategically about IT in general, not just focusing purely on the day to day grunt work.

During our initial conversations with HPE SimpliVity, what they told us about their backup and restore capabilities blew us away. They claimed that with hyperconvergence we could restore between data centers within seconds. That kind of assertion had to be put to the test, and not too long ago, we had the opportunity to see exactly what our new infrastructure could do.

We have a file server running on HPE SimpliVity that’s 14 TB, and a couple months back someone mistakenly deleted the assessing folder, without which the assessing team can’t do 85 percent of their work. In less than 20 seconds I was able to restore the data, and in under 10 minutes it was back up and running. With our old legacy system, we would have had to recall disks, get that data back, copy it—all of which would’ve cost us more than four to eight hours. In addition, the SQL job mentioned earlier that took 14 hours before HPE SimpliVity, now completes in 2 hours. 

We now have a second site with a second vCenter to use RapidDR. I’ve tested failing over the whole environment in a fully automated and orchestrated way—we’re talking 25–30 VMs—and that took about 15 minutes. It’s reassuring to know that, in the case of a disaster, I can have everything back up and running quickly. Even in peak-production times, we’d be able to get the system back online in less than 30 minutes. Using our old cloud-based backup solution, recovery time would have been significantly longer. 

We’ve been running HPE SimpliVity for 3.5 years now and I’ve never had a failed backup or failed host. We’ve had zero downtime since the implementation, which was a problem with our legacy solution. That all provides me with a lot of peace of mind.

Another area that HPE SimpliVity has had a major impact is with a business-critical database. It pulls every piece of data in the company: sales numbers, claims being processed, the location of region-specific sales—you name it. It’s a 7 TB database, but the system changes 9 TB of data every week. Even on a system like this, the performance is unbelievable. With our legacy systems, we’d start queries at 7 p.m. and they’d often still be running when we got back in. With HPE SimpliVity, these same queries are done in 4–6 hours. 

Capacity Savings & Deploying Workloads

Right now, we’re running 114VM across the 8 systems with 325TB on our 112TB cluster of 8 nodes with prod only using 32TB Physical and Dev only 29TB—that’s a total saving of 213TB. The dedupe and compression with hyperconvergence is just insane.

Also, deploying new workloads used to take us hours, but with the pre-ready templates, in just a matter of seconds it’s deployed. That’s hours of our time back—and with just one standard process. 

Scaling IT 

King Price is growing like crazy, and we don’t have time to waste when it comes to scaling our infrastructure to keep up. With HPE SimpliVity, it’s as simple as popping a new node into your server cabinet and deploying it. It’s just so simple, and now we’re looking at DR and other active data centers. For anyone considering hyperconvergence, I encourage you to make use of the DR functionality from the outset. Don’t wait. 

Our book of gross written premium, the total amount of premium we collect annually, is growing at impressive rate per year since opening our doors 5 years ago. This is almost directly related to the load and capacity on the environment, including people and processes and technology. We’ve focused mainly on the process and technology, and these investments have paid off big.

In the entire company, out of 24 divisions, IT is the only department that has received quite a substantial growth in our budget for the new financial year. The thinking behind that is that we need to invest in IT because we’ve been able to deliver during our rapid growth, with the help of HPE SimpliVity. As we look to expand King Price internationally over the next five to eight years, it’s vital for us to design a proper active data center across borders, and the time and cost savings from HPE SimpliVity combined with our expanded budget is setting us up to scale easily and efficiently—and with very few personnel resources.

With our old system, we had about four guys managing the admin side of our infrastructure, but with the new system, I’m able to manage everything on my own—in only about a quarter of my time. That resource savings is absolutely vital when growing a business.

One thing we’re really excited about is HPE InfoSight for HPE SimpliVity. I’m looking forward to the insights we’ll get into our environment—the type of insights we’ve never had before. I’ll be able to see exactly when we’re predicted to run out of storage so I can plan ahead. Using InfoSight will be like having an extra set of eyes looking at our system the whole time. I’ll be able to monitor our environment from anywhere in the world—not just locally.

Inherent Risks?

As one of the first companies in South Africa to deploy hyperconvergence technology, we appreciated how forthcoming HPE SimpliVity was in their POC offering. We had questions about if this tech would gain traction worldwide and if it would be able to perform like it was supposed to, but, to be honest, within two weeks of their systems sitting within our environment, our purchase decision was made. From where we are now, it’s definitely been well worth the risk.

Plus, the technology keeps improving with each version of the HPE SimpliVity software they release. For example, now we don’t have to restore an entire VM just to restore one file or folder, and our dedupe and compression has gone up, saving us a ton of storage. This constant iteration on an already amazing system has made me and the entire King Price team huge fans of HPE SimpliVity, and I’m confident that as we grow, they’ll be there to ensure we’re not just running along smoothly, but innovating and improving our processes so that we can continue to give our customers the best insurance experience on the market today.