Application Performance Monitoring

Alexander Podelko, a performance expert working with Oracle, recently wrote a blog post in which he takes a look at the current state of the APM market and what he thinks needs to change in order for companies to get the most out of their solutions.

In his post, Alex discusses the vagueness around the term APM – is it Application Performance Monitoring or Application Performance Management? The APM market is fairly crowded and not well defined – every solution claims to be the best at monitoring all of your application performance needs; however, that just isn’t the reality. And that name APM is also a bit misleading. Few customers use it. It seems to be more of a creation of analysts and the media in some conspiracy theory to sell more newspapers.  Well, probably not newspapers…but you get the idea. Few enterprises have a division of IT called APM.

The real challenge here is “Who owns application performance monitoring?” Of course, there are network performance specialists, typically developers tune Java and .NET applications and there have been mainframe performance experts almost forever. But for today’s composite applications that span silos, there seems to be no owner for their performance. Instead many different groups work on parts of it and perhaps due to this fragmented approach, not always successfully. This points out an interesting challenge for APM solution vendors – who exactly is the user? Part of the co-evolution of APM in the enterprise and of the solutions from the vendor will be on solidifying who the user is and ensuring that the functionality in these solutions directly addresses their specific needs and is not “one size fits all.”

As Alex points out, the lack of set criteria to compare APM tools is what makes it difficult to define the market. In an attempt to bring some clarity to this space, Gartner published its Magic Quadrant for Application Performance Monitoring, which divided the APM market into five dimensions of functionality. However, Alex explains that there are certain areas still left uncategorized, such as enterprise monitoring and performance analysis, system modeling, and capacity planning.

Application performance management has become the overarching umbrella solution, but companies need to clearly identify their needs before investing in a technology in order to ensure that it best supports their business. In addition, since the APM marketplace has exploded over the last several years, companies need to better educate themselves on the different functionality of APM solutions in order to determine what technology best meets their requirements.

In addition, Alex talks about the ability of first-generation APM tools to monitor applications in production – but on a very limited scale. As we have discussed previously in our article on APM and DevOps, there is an opportunity for companies to use APM tools to connect the development and operations teams. The cooperation between these two teams can greatly improve business outcomes at all levels.

In conclusion, Alex notes that there isn’t a product on the market that will do APM across all platforms and systems, and companies should not believe a software vendor that claims they can do this. We think this is a fair point – however, as the market continues to mature and expand, companies need to recognize the difference between the vision a software vendor has and the capabilities of their solutions. Any given organization can have hundreds of different applications that need monitoring and they must be sure what they are investing in will address these specific needs. Companies need to look at the whole lifecycle of their applications and make sure they are monitoring the different phases with the appropriate solution.  For those of you  who utilize middleware messaging as the nervous system of your application infrastructure and need a solution that can support all of your software and appliance-based middleware technology needs (including WebSphere MQ, TIBCO EMS and RV, Solace, WebSphere DataPower and home-grown) consider Nastel AutoPilot.