Keep the application performance blues at bay

Aman Chandra

Familiar with that ill-fated click leading to an eternally rotating ‘hour-glass’ mouse icon with an “application not responding” message? What would it lead you to think - “Our network is becoming worse by the day. It’s time the IT team upgraded the network or added bandwidth to make it work.”

IT team or the network administrator informs you that the application just barreled a huge volume of data down the network when all that required to process the task was 20% of the actual. The question arises about what is to be fixed, network bandwidth or the amount of data the application sends?

An application’s “structural quality” could be the cause. The problem may not be so profound if the application in question is related to some routine admin tasks such as a HR work processing app. This can cause some end user dissatisfaction at the most.

But what if it is a customer facing web application? Imagine the sales lost with persisting problems, especially during the peak shopping periods. To put this into numbers as per a study by research firm Aberdeen, application performance can impact up to 9% of total corporate revenue.

So, how to address this “structural quality” issue? The applications are complex today. There are multiple components that make up an organization’s IT: multiple technologies and devices, legacy components married to latest technology platforms that are linked to a variety of databases and networks. These applications run in different environments (cloud and data center) accessed from multiple devices, making it almost impossible to identify the root cause if a problem arises.

Complexities further grow as businesses deal with application exigencies in a disorganized manner. Take a look at some of the possible causes for your applications’ non-performance:

What then, is the solution? The first step is to quantify the quality level of an application. There are scientific methods to do this. One is to identify and define application health parameters such as performance, security, resilience, scalability and maintainability. The next step is to set an acceptable performance threshold based on the business environment. Together, these parameters constitute an index termed as “Structural Quality”.

Once the structural quality analysis is done, which is mainly static in nature, dynamic analysis follows allowing for a more holistic assessment of application behavior. This type of assessment can also be done in application development or pre-production stage, to nip defects in the bud (taking a proactive approach). If it’s done post deployment of the application, it’s more of a reactive approach.

Different types of assessments and outcomes in both approaches are captured in the figures below:

It’s a blessing that almost all of this analysis now can be easily automated using a variety of tools. These next generation tools enable deep diagnostics and aid in initiating permanent remediation of non-performing applications. It’s just that the service provider should be intelligent enough to use the right tools for the right application environments using the right benchmarks.

Organizations intensify testing efforts when critical applications crash or begin to show signs of malfunction. However, it is better to pre-empt such glitches and manage applications proactively arresting possible drain on revenues.

The automation approach described here enables reduction in the effort spent on downstream activities of application development like testing, review and rework activities. This in turn helps reduce overall cost of application maintenance. The framework based on proactive and reactive application assessments can be used for even the most complex of IT application landscapes, found nowadays in large organizations.

The author is Platform Leader, Global Transformation, Wipro Ltd.