News

Survey Shows Application Performance Monitoring Needs IT’s Attention

IT still stuck reacting to problems received by help desks

• By James E. Powell
• 01/30/2009

While many IT organizations effectively monitor their infrastructure -- from servers and storage to networks and peripherals -- application performance monitoring remains elusive. According to a new study, most IT operations groups are not sufficiently proactive to anticipate issues before the help desk is contacted.

To learn more about this gap, Compuware Corporation commissioned a survey by Forrester Consulting of 389 technology decision-makers actively involved in application performance management (APM). Compuware wanted to know about the effectiveness of their application performance monitoring, what technologies they used, and what technologies they employed.

The survey revealed what Forrester called “a major paradox” -- 87 percent of respondents said they use end-user “experience monitoring tools for at least some business critical applications” but almost two-thirds of these respondents (64 percent) acknowledge that it’s not until users contact the help desk that they know of problems with performance or availability.

Forrester says that “most of the enterprises consider application management tools to be their end-user experience monitors and have not yet taken advantage of new technologies which really do monitor the true end-user experience, including the effect of network, servers, databases, and applications.”

In fact, the survey concludes that IT still is reactive, not proactive -- IT hasn’t come out of firefighting mode. The survey also points that IT still doesn’t understand the business impact of performance and availability issues so they can’t assign “realistic” priorities to solve problems. Furthermore, “even if the fault domain of issues can be identified, itself a challenge for most IT organizations, there is a lack of diagnostic tools and processes to drill-down and analyze application performance effectively.”

Part of the problem is due to application development teams being separated from IT operations, with both teams working “under different success parameters,” according to the report. IT’s understanding of application performance depends on a cross-domain perspective, but IT operations “is often still organized by domain, so there is nobody really responsible for this cross-domain perspective.”

In addition, IT personnel don’t have enough knowledge about the applications themselves. Including application developers or software vendors in solving application problems “is time-consuming and often unrealistic,” so they typically try to determine which infrastructure element is faulty rather than “considering the code or data design of the application itself.” Complicating the issue: IT often cannot replicate the problem, sometimes because there isn’t a problem. “The classic scenario still prevails: a group of server-, network-, database- and application-managers all claiming that their component is not at fault while the end user is adamant that something is wrong.”

The survey points out that “IT organizations that do not invest in application performance management tools that consolidate data across technology silos, including the multitude of applications that run in parallel, will not be able to deliver a proactive, holistic, business-oriented service to their customers.” Forrester recommends that “IT organizations must employ technologies like end user experience monitoring as well as re-organize to better reflect business priorities and interests.”