Getting ready for the next generation APM

by Sohini Bagchi    Jul 03, 2013

Scott ca

As companies rely more on virtualization and cloud environments, application performance management (APM) systems help them run their enterprise applications more efficiently and at a much greater speed, providing greater value to business. In an exclusive interaction with CXOtoday, Scott Williamson, Senior Director, Product Management at CA Technologies explains the latest trends in the application performance management space and how APM can help companies simplify their IT complexities and make strategic decisions.

What is it that companies still do not understand about application performance management?

Companies still seem to be taking an almost siloed approach to APM where they are focusing on a single aspect – say the ability to deep dive into transactions and the underlying application delivery environment to troubleshoot problem – but not on a holistic strategy for monitoring and managing the performance of critical applications. If a business depends on applications for revenue and productivity, it needs to have that deep dive capability as well as the ability to measure end-user experience. IT organizations must have that complete picture of application health and end-user experience to ensure that the business is running efficiently and customers are happy.

What are the most common mistakes organizations make when selecting and deploying an APM solution?

I see three areas where IT organizations tend to make mistakes when it comes to selecting and deploying an APM tool for assuring the reliability of complex, business-critical applications and services.  First of all, organizations overlook the production scale. Maybe the development team picked an APM tool for pre-production testing or in a proof of concept test, and everything works fine in those limited environments. But a production environment is a completely different animal and enterprise-grade APM tools need to be able to scale accordingly. Organizations that had no problem with its APM choice in those earlier tests may suddenly find themselves with a tool that either can’t keep up with the number of transactions and metric data being processed or require such hefty storage and hardware requirements that ROI is hard to achieve.

The second problem is companies often do not understand that customers require a holistic view of application performance. This means they need the ability to delve into the problems in a transaction or service and have the ability to measure end-user experience. The latter is important because your server and database lights may be green, but your end-users may still be having a terrible time using your application.  This combination of views helps IT better assure business service reliability and limit the impact of slowdowns and outages by finding and fixing problems more quickly and efficiently.

Finally, IT organizations aren’t getting enough value from their APM data. APM systems themselves are creators of Big Data based on the sheer number of metrics being collected. For instance, we have a customer that collects over a billion metric instances per day. While that data is very useful for alerting to extreme problems based on thresholds and baselines, it can be further leveraged using advanced analytics to find more subtle problems that may be occurring, helping IT get ahead of problems before end-users are impacted.

What APM capabilities are required by a company to deploy business critical applications in the virtual environment?

Most large enterprises are deploying virtualization and cloud technologies to gain better efficiency in delivering applications and services. While this adds a layer of complexity when it comes to monitoring the performance of an application as its underlying hardware may be changing on a near constant basis, IT organizations still require the ability to triage issues into virtual layer. To do this, they need the ability for their APM solution to be able to move with the application in the virtualized or cloud environment. Our APM handles this by monitoring user end-users transactions from end to end, regardless of which physical server(s) process the transaction. This gives our customers greater ability to pinpoint problem spots in highly virtualized environments.

There’s a lot of buzz around APM 2.0? Can you kindly explain?

APM 2.0 is the next generation of application management. Some might say the APM market has been a bit commoditized as the basic technology has been around for a while and there are a lot of players in the space. But, I think there’s a lot of innovation and change continuing to happen which businesses can look forward to in the coming days. For example, Cloud APM or SaaS-based APM are already being deployed by organizations that are looking for greater flexibility or agility. Likewise companies will also run APM in Infrastructure-as-a-Service (IaaS) and Platform as-a-Service (PaaS) environments that can generate an excellent user experience. One way to do that is to run an APM solution and give customers access to the performance dashboards so they can see for themselves that their applications are running as they expect.

The other innovation is Java Script Injection-powered end-user experience monitoring. Most APM tools do a good job of telling you there’s a problem somewhere in the data center. But what if the cause of the performance slowdown is somewhere between the edge of your data center and the customer’s browser? Java Script Injection-powered end-user experience monitoring helps provide that last mile visibility to give IT a complete picture of performance and help troubleshoot problems that might not reside inside IT domain.

Then there is advanced analytics for root cause diagnosis. When it comes to fixing problems with critical systems, cutting down mean time to repair can positively impact the bottom line. With our most recent release, we’ve added Application Behavior Analytics, an automated engine that can help find performance anomalies in complex applications. In this, the engine doesn’t need to be “taught” the application and its components, it automatically figures out the “normal” behavior and alert to issues that deviates from what it considers normal. In other words, it’s another set of “eyes” looking for potential problems – and working at much faster machine speed, not human speed, which helps IT organizations be more efficient. There are many more such new generation technologies that are coming into APM systems that are aiming for greater efficiency and improving bottomline.

How can APM help companies simplify their IT complexities and make strategic decisions.

A standard APM solution provides deep visibility into all layers of a complex application can help simplify things by making troubleshooting more efficient and helping to assure the great experience users have come to expect from modern applications. But APM can go far beyond that, providing greater value to customers. For example, it can help reduce hardware cost through predictive reliability. In the case of CA APM, we’ve integrated it with our CA Capacity Management suite so that rich performance data from production environments can be usedto build more accurate capacity planning models, allowing customers to right-size their future application environment while reducing risk of failure and optimizing capital expenditures.

Besides, APM can help increase production readiness by leveraging the technology in pre-production testing environments to help find performance bottlenecks earlier in the development cycle. Issuesfound earlier in development are cheaper to fix than after an application has been deployed in production. You also avoid production downtime and customer satisfaction issues. It can also help organizations understand the performance of its applications and target those underperforming applications for upgrades, rewrites or elimination. Production performance data is great scorecard for figuring out which applications are running effectively and properly supporting the bottom line.

On the whole, strategic use of APM can help customers understand how end users are being served and the relationship between application performance to customer satisfaction, revenue, and retention. This is of particular importance to businesses where customers work on a virtualized setup.