Why New Relic
Proactively identifies performance and customer experience issues with microservices-based platform.
- Uncovers the root cause of performance issues
- Visualizes the interdependencies and associated performance of more than 50 microservices
- Alerts engineering staff about changes to application health
From reactive to proactive
Since its founding, Lyst has enjoyed astounding growth. Today, the platform sees 60 million shoppers a year—and they expect a fast, easy shopping experience.
However, when Lyst first began, it didn’t have a way to monitor application performance for its main platform. “We relied on parsing website logs and trying to identify requests that were taking too long,” says Serko. “We didn’t have a way to be more proactive about identifying and fixing performance issues.”
Then the company turned to the New Relic Digital Intelligence Platform for end-to-end visibility into the performance, customer experience, and business success of its fashion search engine platform. “As soon as we began using New Relic, all of our performance issues were suddenly very visible,” says Serko. “We could then go down the list and mitigate or fix them to improve performance and customer experience.”
“As soon as we began using New Relic, all of our performance issues were suddenly very visible. We could then go down the list and mitigate or fix them to improve performance and customer experience.”
From monolithic to microservices
Soon New Relic would become even more beneficial to the Lyst team. “About two years ago, we began shifting away from our large monolithic application to a microservices architecture,” says Serko. “Instead of one big code base, we now have 50 microservices, with new ones being added as we develop new features.”
For example, one major microservice stores all of the Lyst product data, powering the search function on the website. “You can imagine the heavy load on this microservice because it needs to return all the product information back,” says Serko. Other examples are the search and recommendations microservices, “all of which chain up nicely together,” according to Serko.
While the microservices approach brings advantages for Lyst such as faster development cycles and greater fault isolation, it also has the potential to make the environment more complex to track and understand. That’s where New Relic has been essential. “As we started splitting out features into microservices, New Relic helped us by not only tracking each one of them separately, but New Relic’s service map also lets us visualize the different services together and monitor performance across the entire environment.”
Catching problems before customers do
Today, Lyst relies on New Relic to identify potential performance issues before they impact the global customer base. “We are heavy users of New Relic’s alerting,” says Serko. “We have alerts set up to track most of our servers, each microservice, and key transactions. We want to know the minute anything goes berserk.”
Lyst also uses New Relic Synthetics to simulate user behavior and catch problems before customers do. For instance, the company simulates customer sign-ups and sign-ins to make sure they are running correctly. If not, an instant alert notifies the team of a potential problem. “Our frontend engineers use New Relic Synthetics to help them improve response times and spot issues with the customer experience,” says Serko.
Focusing on the mobile experience
Next up for the Lyst engineers is optimizing the mobile experience. Lyst offers both a responsive website optimized for mobile devices and a mobile application for its customers. “We want to deliver the best mobile customer experience possible,” says Serko. Lyst’s engineers can rely on New Relic to help identify ways to achieve their goal, because with online commerce, good performance is always in fashion.