Discover PerformanceHP Software's community for IT leaders // September 2013
Staying ahead of the game
Predictive analytics was a smart bet for online gaming giant Playtech, which has improved performance and knocked out potential crises.
In the highly competitive and fast-paced world of online gaming, downtime is not an option. And nobody understands this better than Playtech, a leading online gaming software supplier with annual revenue in excess of US $133 million.
“Downtime can cost Playtech and its customers significant amounts of money—as much as hundreds of thousands of dollars an hour,” says Eli Eyal, who leads Playtech’s operational support system team.
The 14-year-old company develops software platforms and content for the online and land-based gaming industries. Its products include casino, casual games, sports betting, live gaming, lottery, bingo, and one of the world’s largest poker networks. Flawless performance and 99.999% uptime are critical—especially during peak usage times such as sports races and other special events. Even minor performance glitches can spoil users’ gaming experiences and disrupt revenue streams.
Eyal joined Playtech in 2010, hired to revamp the company’s monitoring systems. Prior to his arrival, Playtech had used very limited diagnostic tools: technicians had to cut and paste event information, then search web pages for instructions on how to handle the alert. “The monitoring system was very basic, lacking support and the ability to develop new features,” Eyal says.
Using predictive analytics to extend monitoring capabilities
Eyal helped Playtech replace its limited, heavily manual infrastructure monitoring system with HP IT Performance Suite software. Recently, Playtech also started monitoring the applications side, using the game-changing capabilities of HP Service Health Analyzer (SHA). HP SHA—a predictive analytics engine—helps to forecast problems and prioritize issues based on business impact. It analyzes historical data to automatically create real thresholds that uncover potential issues to allow action before service is impaired.
“We were pretty much covered with monitoring on the infrastructure side,” Eyal says, “but we needed a whole new perspective on the applications side of the business. That’s where HP SHA made a huge difference.”
Most monitoring tools are reactive—triggered by arbitrarily defined thresholds. Playtech’s new solution, in contrast, performs statistical analyses on historical data to learn system behavior and set thresholds based on what it discovers.
Recently, Playtech noticed unusual changes in gaming activity in Indonesia. System alerts indicated that the number of games being played every five minutes during daytime hours had dropped by half. Playtech called its customer to investigate and learned that nothing was technologically wrong: one of the countries in the region had temporarily shut down its gambling casinos due to a political situation.
On another occasion, the system immediately informed Playtech of a potential denial of service attack, and the company was able to take fast action. In both of these examples, Playtech was immediately alerted that its normally predicted usage thresholds were out of sync—thus requiring IT’s attention.
With advanced monitoring on both the infrastructure and applications side, Playtech is now able to see user behavior, detect and resolve issues quickly, and minimize—if not fully eliminate—downtime. “We get alerts from places we couldn’t even reach before, because we can now monitor the entire picture—3,000 servers on the infrastructure side, and activity on the entire applications side,” Eyal says.
Eli Eyal will discuss Playtech’s success with predictive analytics at the “Power to Change” free virtual event. Check out the 18 online sessions and register today.
Diving into disruptive technology trends like cloud, mobile, and Big Data, HP’s CEO talks about moving not just IT, but the whole enterprise, into a new era.
Dig into strategic trends with our new Discover Performance Weekly video series, and go backstage at events like RSA.