Home > News > Poor application performance tools hurting business

Poor application performance tools hurting business

Published:23-July-2009

Application performance issues getting more complex


Lack of effective root-cause analysis tools that pinpoint and diagnose application performance issues is leaving some IT shops held responsible for creating problems that have a high impact on their organisations.

Although some companies have invested in application performance analysis software to tackle the issue, some of the tools cannot deal with the level of application complexity that stems from use of latest technologies such as virtualisation and service-oriented architectures, a latest survey has suggested.

New approaches to APM are needed that help enterprises deal with the unique challenges presented by frameworks, cloud computing and other complex application architectures, the company behind the study said.

Roger Andrews EMEA VP at DynaTrace told us, “APM tools were borne out the need to manage Java and .Net applications, which was beyond the reach of traditional existing systems management software approaches.” He explained that although APM products produced by Wily Technologies, Mercury Interactive, Precise Software and the like were highly successful, the level of innovation in the market basically ground to a near halt after a spate of high-profile acquisitions in 2005 and 2006.

“The game has moved on considerably since,” he said. The rapid pace of change in application technologies and development processes is rendering these older generations of APM tools increasingly obsolete.

The company offers APM tools that are tailored to the needs of development, test and production and which generate so-called PurePath data. Effectively the system uses transaction tracing technology which follows and captures individual transactions down to the code level across different tiers and technologies to provide software engineers or any other interested party with root cause diagnostic information.

The company has added dashboard technology, which allows views to be built to present the PurePath data in a way that makes sense for the person reviewing it. For a developer that might be a view of an SQL statement, whereas someone in operations might want to see response times or memory allocation or usage.

“Using PurePath we can capture all the environmental details and trace an application transaction all the way through, from start to finish” Andreas Grabner, Technology Strategist at dynaTrace explained. “Up to 40% of a developer’s time can be spent on fixes, and by recording and capturing information in real-time, rather than having to reproduce a problem in a test environment, our software can reduce the mean time to fix by up to 90%,” he claimed.

In an average project, DynaTrace reckons that nearly half of the organisations it surveyed are spending at least 20% or more of their time replicating and fixing application performance and availability issues.

The company also found that nearly one-third of performance issues are found in production, where they are not only more expensive to fix but have the greatest business impact.

Research has shown that issues found later in the lifecycle are anywhere between 10 times and 100 times more expensive to resolve than those found in development.

Share this article:

Your opinion

Login to post comments.

True, today’s SOA and virtualization initiatives increase complexity and risk exponentially. The biggest challenge is of course that even the most radical of strategic IT refits do not rip out and replace everything; so performance management solutions that effectively police legacy, current and future applications as a continuum are imperative. It is precisely for this reason that the position of the certain established web application performance management vendors is enhanced. The last time I checked SOA and virtualized environments were still predicated upon JAVA and .NET. The shift toward JAVA and .NET applications, and the emergence of web services, message and integration brokers, Enterprise Service Buses (ESB), Portal Servers etc is all a part of an evolutionary trend; so innovation that builds upon acknowledged successful APM solutions is what is required. APM is about managing every aspect of your customer serving environment from end-to-end all the time without introducing undue stress, load or latency. Cross process transaction tracing is interesting but hardly unique. In my view the real innovation in APM is in tracking web services into/out of the cloud as Business Services, monitoring individual end-user experiences and transactions in real-time, identifying and triaging problems across the entire legacy to V-Cloud continuum, and managing incidents by business impact. Just thought a bit of balance was required here.
Kobi Korsah - 03:46:53 27/07/2009
I totally agree that the APM market has been stagnant for a while and that now, for the first time in many years, we’re seeing a burst of innovative companies with new approaches reinvigorating the market. It makes for exciting times! Where I disagree is that APM was born out of the need to ‘fix’ Java and .Net. In my opinion APM came about because large enterprises weren’t able to get the end-to-end metrics they needed to manage their IT - Java and .Net were just small piece of this. The article talks about ‘the new world’ but the fact is that most big enterprises became web enabled using legacy applications and if they want to compete effectively the challenge that led to the birth of APM remains in place today. In order to address this challenge, IT Directors need to monitor and manage the whole infrastructure, not just one small piece of the jigsaw.
Colin Rowland - 09:29:11 30/07/2009

Newsletter Subscriptions