Fundamentals: Application Early-Warning System

Oct 10, 2012


Application Early-Warning ­System

Most IT teams are responsible for at least one non-Web-based business application that must be available pretty much 100% of the time, and by “available” we mean it doesn’t take 10 seconds to refresh a screen. Think about a back-end inventory database that salespeople and partners depend on; it’s extraordinary that many IT shops are still in the position of not knowing performance is tanking until the help desk lines light up. And if you can’t proactively monitor mission-critical applications hosted in-house, you’re going to be in worse shape if (or when) they start migrating to IaaS vendors.

It’s relatively easy to scrutinize the storage, database, middleware and application servers that host an application. But while background monitoring of these and other infrastructure metrics has its place in application monitoring, it’s not enough if IT wants to get out of the business of manually hunting down problems and into the business of innovation. We need better early warning of problems, better root-cause analysis, an ability to keep tabs on systems hosted externally and ideally the ability to automatically head off performance glitches.

In this report, we’ll dig into the substantial logistical challenges of using synthetic ­transaction monitoring to inspect applications, beyond websites, and look ahead to what’s happening with cloud vendors. (S5971012)

Research Report