Application performance management tool selection: Monitoring vs. managing

Last week, while doing some background research for our application performance management tutorial, I got into a discussion with Julie Craig, a senior analyst at Enterprise Management Associates about the difference between application monitoring and managing. While some tools can do both (i.e., troubleshooting and user experience monitoring), selecting the right tool for the job often depends on who’s using the tool.

Application performance metrics provide information about things like response time, quality of experience and user experience, Craig said. Application management products typically go “deeper” into the technology underlying the application, so that they can do the following:

• Detect performance problems
• Detect availability problems
• Correlate information from multiple underlying technologies to indicate overall application health.

“When application problems occur, [application management tools] isolate potential root cause to specific infrastructure elements or types,” Craig said. “For example, a performance problem might be traceable to poor database performance. This could be due to multiple factors, including poorly written SQL calls within application code, failure or potential failure of servers or network connections, too many users or transactions creating bandwidth bottlenecks, or potentially hundreds of other factors (or a combination of several). Application management solutions help IT teams track down problem source and hopefully minimize the amount of time that cross-functional technology engineers have to spend on isolating root cause and fixing the problem.”

Taking responsibility for app performance
According to David Langlais, the director of product management at BMC Software, senior IT staff care about whether end users call in to complain about an application performance issue. “For senior staff, seeing the inside of a JVM [Java virtual machine] is not that important.” These technologists care about technology’s results.

But still, someone has to get under the hood and take charge of the problem. With all of the different pieces of infrastructure involved in a multi-tier application, when an end user says, “There’s a problem,” every single person in an IT department could potentially get involved — from application programmers to network or database managers.

“The responsibility of performance management is distributed, but organizations don’t really like to work that way,” Jasmine Noel of Ptak Noel & Associates said. “What I’m seeing is the folks managing the application servers are assuming the broader role of Web applications manager, and they become the one throat to choke.”

Application performance management tools list
Craig emailed me the following list of vendors in the application performance space. The first list includes examples of tools that help senior IT managers get an end-user view of application performance. The second set lists application management tools for server managers looking for a deeper analysis. Many tools overlap both categories:

Application monitoring tools

• OpNet
• CA (Wily)
• Quest (Foglight)
• Oracle (Empirix)
• OpTier
• Citrix Edgesight
• BMC Proactivenet
• Coradiant (TrueSight)
• HP Mercury (now HP Business Technology Optimization Software)
• ASG Application Management
• This group also includes hosted services like Application Performance, Gomez, and Keynote.

Application availability, troubleshooting, root-cause analysis:

• Compuware Application Delivery Management
• Quest (Foglight)
• CA (Wily)
• BMC Proactivenet
• HP Mercury (now HP Business Technology Optimization Software)
• ASG Application Management
• OpTier
• IBM (Tivoli)
• Symphoniq
• PacketTrap
• eG Innovations

In your IT department, who’s in charge of application performance? Do you know of other application performance tools we should add to the above list? Leave your feedback in the comments.