Cars.com uses large data to improve Web site performance

Car Sales Web site Cars.com through the Web user clicks and machine generated log files to improve the revenue of the enterprise, and successfully resisted the malicious robot.

E-Commerce has established an association between revenue and the quality of the user experience. For the large data of cars.com, the measurement of the degree of experience has become the main goal of the large data analysis. By analyzing 12 million visits per month, the company has not only optimized the user experience for consumers, but also gained deep operational insight and fraud prevention capabilities.

As a professional web site for car consumers searching, consulting and buying cars, Cars.com's strategy of using banner ads to advertise cars, SUVs and small and medium goods vehicles for all major carmakers has not only increased the revenue of the site, but also earned advertising costs from car sales. A faster user interface allows visitors to spend more time browsing the site and dramatically increases the chances of a visitor buying a car and clicking on a banner ad.

Cars.com's application management team has three goals: maintain high performance, protect content, and track traffic sources for advertisers. In the background of the Web site, robots and crawler traffic are a persistent threat because they can lead to degraded site performance. Some malicious robots will also grab a list of cars and other content, so that spammers can trick consumers into leaking personal information.

Log files control the key to identifying malicious behavior and performance optimizations, but it is tedious and time-consuming to get blog traffic data and analyze it manually. Because there is no real-time reporting, the management team can only reserve space on the server infrastructure to ensure that the page loads faster.

Cars.com eventually chose Splunk to collect, index, search, and analyze large, machine-generated datasets from a large number of different sources in real time. Its core is the Splunk patented machine data Web to organize and analyze the log data. Splunk also includes prefabricated reports to help the management team identify illegal web crawl and robot traffic, differentiating and separating them from normal access traffic. The report can also be a valuable sales tool and powerful support tool to help internal personnel take appropriate measures to address unnecessary traffic.

The return on tangible assets investment is achieved in two ways. First, efficient real-time data collection can save more than 400 hours per year. Second, it can help to adapt to peak traffic. For example, in the 2012 American rugby Super Bowl, detailed performance statistics provide operational insight that helps save about $160,000 in server and management costs.

"Splunk allows us to deal with a lot of problems in the shortest possible time," said Jon Abend, Cars.com's technology operations manager. In addition to Web logs, we can now easily analyze application logs, application servers, middleware components, and system metering logs. Without the ability to manage all types of systems in a balanced way, we may not be able to provide services to a variety of users, such as performance engineers, middleware teams, and search engine marketing teams. ”

From the beginning of 2012, Cars.com's large data environment began managing more than 35TB of data. The Cars.com website will add 2.5 million posts per hour, add 1TB of data each week, and handle more than 750 million consulting information per month. With this large data analysis insight, Cars.com will be able to maintain a faster access to the site. (Grid Network Gundy compilation)

(Responsible editor: The good of the Legacy)