The ability for people to be able to buy food and essential supplies, in-store IT and data infrastructure must be kept running. Jeff Miller, Director of IT at Jerry’s Foods, a national chain with fifty grocery, retail, hardware, and liquor stores, knows this reality very well. Miller explains that each Jerry’s Foods franchise includes many systems (point of sale, loyalty/rewards, security video, etc.). He stressed:
“If just one store’s systems went down, we could potentially lose $60,000-$70,000 depending upon the time of day. Should this happen across the board, then the impact would be much worse. Each franchise needs to connect to the Electronic Benefits Transfer (EBT), in real-time, verifying funding for purchases. Without EBT access, we can lose tons of groceries as customers, unable to purchase what they need, walk out of the store, or leave carts full of perishable products, needing an employee to restock them.”
Grocery stores do not take such a hit well, as they have thin profit margins. Also, these markets have less flexibility when choosing technologies. Miller explained, “Wholesalers normally dictate the kind of system a grocery store has to use.” Combine these two factors, and data virtualization becomes an attractive solution, simplifying and consolidating IT infrastructure at each store.
Over time, Miller found that managing data and IT technologies on a store-by-store basis started to become costly, unreliable, and difficult to operate and control. How could Jerry’s Foods upgrade and allocate its IT resources, including data virtualization, to ensure system reliability and functionality across the board?
Starting with Data Virtualization
Jeff Miller has followed virtualization technology back to its beginnings. In the mid-1990s, Miller put in storage area networks (SANs) with VMware software — cutting-edge technology at the time — for Schwan’s Food Company. VMware leverages multiple virtual environments on one computer setup, saving the need to purchase numerous machines or equipment. Virtualization at Schwan’s Foods saved IT maintenance time and costs, too, making this innovation even more attractive. For the time, VMware was a great solution.
No surprise, when Miller started at Jerry’s Foods, he virtualized “everything right away” using VMware to save costs. But, after several years, this solution was no longer viable. VMware changed its pricing model when it was bought by EMC. He remarked:
“I saw a non-ending rat race where upgrading the VMware meant purchasing new hardware systems too. It looked like this pattern would continue indefinitely. I confirmed my fears about a month ago as VMware announced changes to its licensing contracts. I wanted out of that whole game.”
At the same time, problems with Jerry’s Foods’ video systems involving store security became even more apparent. The DVR and its data storage became “overtaxed, overworked, and out of date,” said Miller. The security system along with the data virtualization infrastructure needed an update to a better solution.
Bolstering Data Virtualization with Edge Computing
Six or seven years ago, Miller saw a possible approach to his problem. He met Scale Computing’s founders Jeff Ready and Jason Collier, who astounded him with the simplicity and efficiency of their HC3 Edge products. These appliances empowered remote data processing and remediation at each source by using edge computing. Freed from needing a centralized data center or hub, each grocery IT network could run independently.
Scale Computing’s edge computing nodes promised a more reliable solution with minimal hands-on maintenance. It provided a server, storage, and virtualization product in one. Should a remote location experience an IT infrastructure glitch, a backup could turn on within a minute and systems could be brought up right away. Knowing that each store had self-healing systems that performed very well, in addition to being able to adapt to newer technologies, enticed Jeff Miller further.
Implementing Data Integration at the Right Time
Even though he was impressed, Miller delayed pulling the trigger with Scale Computing. At that time:
“Scale Computing had a bigger deal and a traditional one. It was too much. I kept bugging Scale for a smaller solution, just on a bunch of PCs. I was looking for the right size although I always wanted to bring Scale’s solution to keep connectivity working in a cost-effective manner.”
Miller found an opportunity to justify getting Scale Computing’s system about a year and a half ago. He needed to deal with the stores’ old dying video systems right then. Scale Computing had a viable upgrade, and Miller tagged along with a point of sale system and everything else too – a suite of HC3 Edge products.
He used a Florida store to test some small HC3 Edge devices. The test went well and deployed quickly. Miller started installing in store after store. The security video system improved immensely, with backup capabilities. Also, Jerry’s Foods recorded a 50 percent decrease in cost per instance and a 50 percent decrease in IT infrastructure maintenance time.
Success started to snowball. Miller originally planned for a three-and-a-half-year timeline to migrate all machines to Scale Computing. Now, he said, “We will be able to get all our systems swapped out by another year.”
Sticking to the Plan in the Middle of Success
Successful implementation and migration may not seem like an issue to some, but it can throw a wrench in migration plans. In moving forward more quickly than anticipated, Jerry’s Foods experienced a little pain. The global management piece of the network did not get set up before all the other nodes fell into place. As a result, a combination of Scale Computing’s product flexibility and excellent service got off track by all the success with HC3 Edge implementations.
First, Miller and Scale Computing needed to solve problems with a piece of switching infrastructure not initially compatible with Scale Computing’s systems. Miller and the vendor fixed this issue, getting everything integrated and working. He said:
“I don’t have to think about infrastructure issues as much. Scale Computing has a system that upgrades readily, by juggling machines around to improve one node and then juggling around to optimize the other. The Scale Computing solution was working very well.”
Also, Scale Computing’s support made the implementation easy. Its robust program to implement systems and hands-on customer service made the migration happen seamlessly. Miller observed when offering feedback to Scale Computing:
“After customers install five or six systems, Scale Computing does not need to hold clients’ hands. The process works well enough. A customer does not need the sixteen emails and set up calls and all the rest. We got it.”
Both success factors just increased demand by Jerry Food’s store managers for Scale Computing HC3 Edge. He tried to keep up. In addition to swapping the Florida test for production appliances, last March he requested another six sets of nodes, including one for Jerry’s Foods’ headquarters.
Just as he would substitute 40 systems with HC3 Edge, others store requests wanted to swap out thirty and get rid of another system. Jerry’s Foods and Scale Computing would have to arrange for another node. He realized he was moving the implementation too fast and reset the timeframe. He advises sticking to the migration plan and anticipating success. “If the migration goes sideways, then you can fall back and slow down the transition to the upgraded technology.”
Adding Redundancy and Reduction in IT Resources
Miller strongly suggests improving legacy virtualization with an affordable system, leading to minimal downtime and working more effectively. He thinks that the redundancy and reduction in IT resources from the HC3 solution provides motivation for other retailers to follow suit.
Redundant systems, not quite as common in the typical grocery store, act as a fail-safe at each store level. Where one Jerry’s Foods franchise operates “as a little $40 or $50 million business,” having that redundancy puts these businesses and Jeff Miller at ease.
Many stores, Miller believes, focus on the customer-facing IT applications. But considering IT infrastructure turns out to be valuable in keeping retail going. He added:
“Where you don’t have to babysit IT infrastructure in the store because you know it will keep running, even with a node down, resources can be redeployed elsewhere. You can take your staff and redirect or retrain them for a more worthwhile project or in another place.”
Miller concluded that edge computing remains untapped among many retail stores. He suggested, “maybe some will be willing to plan a new implementation when they see five-year cost savings at other places, doing a VMware/whomever combo.” Retailers may not put IT in-store improvements at the top of their list. But the marketplace will require an upgrade of in-store IT resources to stay in business.
Image used under license from Shutterstock.com