Top Tips on Application Analysis

I was reading Ann Bednarz of NetworkWorld and she posted a very insightful article discussing the importance of network visibility to ensuring optimal application performance.

In the article she spoke with Bojan Simic, research analyst with Aberdeen Group, and discussed how enterprise companies are using tools for network monitoring, troubleshooting, analytics and simulation. Bednarz also presents a list of top tips for monitoring application performance from Aberdeen.

“For companies that want to achieve best-in-class network visibility, Aberdeen recommends these actions:

* Establish baselines for normal network performance.
* Use application response times (rather than waiting for end users to complain) to evaluate network performance.
* Develop capabilities for measuring bandwidth consumption per location and per application.
* Enable remote access to network performance data and remote troubleshooting of network performance.
* Develop capabilities for simulating network performance.”

A lot of these suggestions will come down to the tools you use to monitor network activity and application traffic. Taking the list a step further, I have assembled a list of top tips for application monitoring tools. These are things to look for in any type of network analysis device.

1) Since a large part of performance monitoring is timing, how does your current network analyzer monitor application response time? We covered this in-depth in an earlier post.

2) Long-term application performance monitoring and trending is critical for establishing baselines of normal network performance. First, do you rely on your network analyzer daily, or is it pulled off the shelf only during an emergency? Consider the reporting capabilities of your analysis tools. Do they provide long-term trending and collect information across your enterprise’s network? How easy is it to share this information with IT team members and management?

3) What level of application information are you analyzing? Can you drill-down to the packet-level to analyze the cause of a performance slowdown?

4) Can you track application response time and conversations across mulitple segments or hops on the network? With MultiHop Analysis, your analyzer should automatically correlate the timing of packets between local and remote packet captures. By comparing multiple captures from different ends of the network, you can determine: if the packets reached their destination, delay issues at each hop, and the approximate location of a failure.

5) Does your analysis tool provide a graphical view of system conversations for instant identification of long latency and response times? This feature within the Observer network analyzer is called Connection Dynamics. Retransmissions and lost packets are flagged in red for quick identification. Connection Dynamics is a simple way to easily pinpoint network impediments.

This white paper details important application monitoring features to look for in a network analyzer.