Why New Relic
Provides an easy-to-use solution for monitoring platform operations, user experience, and business performance
• Drastically reduced time required to identify and resolve performance problems
• Provided means to establish and meet service level agreements
• Delivered complete visibility into all code—regardless of development language
Picture a specialty used-car dealer in San Francisco. His very particular, long-time customer is in New York. And the car she lusts for—a low-mileage, Dakar Yellow BMW M3—is up for auction in Ohio by Manheim, the world’s largest provider of vehicle remarketing services.
The dealer’s chances of making the winning bid, however, are only as good as the stream of real-time data he’s getting through Manheim’s sites. Should a software glitch cause the live auction video to freeze and prevent the bid status from being refreshed, that dealer could potentially end up watching another bidder roll out the door with “his” sweet ride—having paid thousands less for it than his client was willing to spend.
Not an ideal scenario for the dealer, his client, or Manheim, which has spent decades building the respect and trust of its customers. But with an agile team, a dynamic infrastructure environment, and end-to-end performance monitoring, the company works hard to make sure these kinds of issues don’t mar its reputation. By partnering with New Relic to monitor every part of its hybrid environment—from its public cloud environment (Amazon Web Services) and the Docker containers residing there, to its on-premise data centers and auction centers around the globe—it ensures a steady stream of real-time data and consistently high performance across its sites and apps.
More than just an auto wholesaler
Today a part of Atlanta-based Cox Automotive’s diverse portfolio of leading digital marketing, software, financial, wholesale and e-commerce solutions, Manheim employs 20,000 team members who operate more than 110 locations around the globe, registering 7.5 million used vehicles annually through more than 780 weekly sales (as of January 2017).
For Manheim, technology has long represented the key not just to keeping pace but leading the way. Says Manheim’s director of production engineering, Jason Riggins, “With the advent of the Web and sites like eBay, the need to actually go to a location-based auction and purchase a car physically is on the decline. Our response at Manheim has been to disrupt the industry with technology—using it to offer all the data-driven services and real-time information about vehicles that make us more than just an auto wholesaler.”
In addition to the traditional and online auctions, these services include financing, title work, transportation (auto hauling), recovery, auto body repair, dealership management systems, dent repair and automotive reconditioning, automotive re-marketing, and more.
Explains Rick MacConnell, senior director of technology at Manheim, “We are now an inventory services and logistics provider as well as an auto wholesaler. And we’re increasingly relying on the data we gather from the vehicles that pass through our locations to provide additional products (like our industry-standard wholesale pricing index) and services (such as condition grading).”
Enter New Relic—Exit guesswork and manual labor
In 2008, a key component of Manheim’s services—OVE.com, or Online Vehicle Exchange, its heavily trafficked eBay-style online auction site—was having a hard time meeting its performance goals. The company needed a tool that could monitor its native application to pinpoint the problems, and New Relic was the only tool that fit the bill at the time.
For Manheim software developer Kris Kemper, that initial deployment of New Relic transformed his job almost overnight. “I was working on the OVE team at the time,” he says, “and the only way we had to solve performance problems was to go onto the production servers and look at time stamps, logs, and things like that—a difficult and time-consuming way to get information. We could also go through the manual process of trying to reproduce a problem—which again was extremely laborious and not always successful. Either way, we had these really drawn-out cycles where you were operating on data that was not as good as it could have been to come up with fixes that might or might not work.”
Enter New Relic, and all of that guesswork and manual labor disappeared. Says Kemper, “New Relic auto-instrumented itself on our native app underlying our digital platform, and it would say, ‘These are the calls to your controller methods; this is how long they take. This is how many database calls happen; this is how long they take.’ Suddenly, when someone said there was a problem, I could go back and correlate these things with very little effort. Plus, New Relic produces these beautiful graphs to show what's happening, so you’re not just interpreting numbers. You can see spikes and valleys, and you can overlay the graphs in different ways—both of which are very useful.”
“By shifting to a DevOps strategy, we’ve eliminated the wall between operations and development. We now have what we call ‘capability teams’ that focus on a particular application or service from development through production. The result is that we’ve been able to greatly accelerate product releases and eliminate the surprises in their deployment. Not surprisingly, New Relic has played a big role in that.”
Continual application delivery for continuous improvement
While Manheim introduced New Relic through the development team, its use quickly spread and became central to the company’s shift to a DevOps approach to IT. Explains Riggins, “By shifting to a DevOps strategy, we’ve eliminated the wall between operations and development. We now have what we call ‘capability teams’ that focus on a particular application or service from development through production. The result is that we’ve been able to greatly accelerate product releases and eliminate the surprises in their deployment. Not surprisingly, New Relic has had a big role in that.”
Indeed, New Relic is now implemented on every system and application that Manheim deploys. Says Larry Green, manager of side operations for Manheim, “The evolving role of operations in DevOps culture is to help teams figure out what’s wrong with their applications, and then let them figure out how to fix them. New Relic is the main tool we use to accomplish that. Without New Relic, the time required to resolve production issues would be increased.”
“Before we used New Relic (and its plug-ins), it would take 10 to 12 hours to resolve situations that we can now deal with in 30 or 45 minutes. The difference now is that we can save the hours and hours of figuring out why the system is breaking and move right to fixing it.”
Going, Going, Gone: Providing data at the pace of commerce
More than eight years into Manheim’s use of New Relic, senior director of technology MacConnell relies on the tool just as much as ever. “Life without New Relic would be very difficult,” he says “We would probably be cobbling together our own tools to make sense of everything that's going on in production. And at the speed we need to operate, that's simply not feasible. We're here to help transform the wholesale automotive industry, not to get really good at building monitoring tools.”
Thankfully for Manheim, New Relic has done that instead, and the result is a suite of tools the company has put to use to dramatically accelerate problem identification and resolution. Says Green, “Pretty much all of our troubleshooting now requires big data so that we can see how the app was performing before the problem and what's different about it now. Before we used New Relic (and its plug-ins), it would take 10 to 12 hours to resolve situations that we can now deal with in 30 or 45 minutes. The difference now is that we can save the hours and hours of figuring out why the system is breaking and move right to fixing it.”
This is powerful stuff, and it also helps in another key area: ensuring that Manheim meets its service level agreements (SLAs). In fact, before deploying New Relic, Manheim hadn’t even established SLAs, since it had no real tool to monitor platform performance. Explains McConnell, “A couple of years ago customers were letting us know that our search platform was slow to deliver pages. This led to us establishing an SLA that we were able to achieve by defining it as an Apdex in New Relic. We were then able to use the data we gathered from New Relic to demonstrate to our leadership that we were meeting that SLA. Today, we’ve set SLAs for each application so that we can ensure that data flows through our system in as close to real time as possible—an absolute necessity for an auction business in which the pace of sales is incredibly fast.”
Better communications for better business and happier customers
Perhaps best of all—at least for director of project engineering Riggins—New Relic has provided Manheim with complete visibility into all of its code, regardless of language. “One of the things we've tried to do in moving to a DevOps culture is to allow our development team to use the most appropriate language or technology for their use case,” he says. “Because New Relic is not bound to a specific language, we can install it anywhere. That in turn means we can deliver technology offerings across business units while still providing a single simplified view in New Relic. This is enormously helpful in diagnosing issues and making better business decisions.”
With the number of New Relic users at Manheim now numbering in the hundreds—and the company now using everything from APM to New Relic Browser, New Relic Insights, and New Relic Synthetics—the automotive wholesaler plans to continue to use the solutions to achieve the business transformation essential to its success. Says Riggins, “New Relic has been vital to improving communication between the teams that form the core of our business. And that can only improve our business and increase our ability to better serve customers.”