There are many ways to approach a problem.
There’s the direct way, which involves seeing the immediate face of the situation and tackling the issue head-on. There’s merit to this, since often times a problem is simple enough that fixing the immediate symptom is more than enough to fix the entire issue.
But not always. Sometimes problems are too complex to be solved directly. In those cases, it’s better to back up and take a step or two around the situation to see what other complexities can be found.
That is, without doubt, the best approach to application performance management and how its perceived in IT. While APM is not a problem in itself, its perception among IT managers could stand a lot of improvment. So, to get an idea of what we’re dealing with, let’s try to examine the problem from all sides.
First, it’s important to understand what Application Performance Management is and what it isn’t. Application Performance Management lives, first and foremost, in application space. This isn’t the monitoring you get with network monitors like Zenoss or Nagios. Nor is it the web-site monitoring people try to sell you all the time on the Internet.
Application Performance Management includes end-user experience monitoring, which uses real-time end-to-end monitoring of analytics, usage patterns, and other metrics to determine exactly where a user’s interaction with the app can be improved. (Some wags will tell you it’s the user who needs improved, but I’d like to think there’s always room for improvement in the app.)
This is the nickel definition of Application Performance Management, and it sounds pretty innocuous, right? Yet, Application Performance Management still brings up a lot of problems for IT managers, who are not always convinced in the value of Application Performance Management.
The simple solution would be to just encourage those managers to use Application Performance Management tools better. Or outright use them, if they’ve never applied Application Performance Management before. That’s a bit of a conceit, since it ignores the core reason why IT managers are likely to distrust Application Performance Management in the first place.
Circle around the problem some more, and you will see why: IT managers, as a general rule, tend to be systems people. That means, if they see a bottleneck somewhere, they will through more systems at it. Slow app? Needs more processor speed. Slow network app? Build more pipe. Slow datacenter app? Build a bigger datacenter, or ramp out to the cloud.
IT staffers with a background in development are more likely to be the ones who put more stock in Application Performance Management, because they know that apps can hold all sorts of gotchas that can slow things down. Systems folk know this, too, but they their general approach to fixing an app is to go back to the developers on their staff with a generally unhelpful admonition of “fix it.”
And this is where the case for Application Performance Management can really be made. Even systems-biased managers should recognize that fixes will go a lot faster if developers have the proper data from which they can find problems. Instead of blind, one-size-fits-all solutions that may speed up an application, targeted, specific solutions that will speed up application performance can become the norm.
Application Performance Management doesn’t have to be the cause of conflict. Implemented correctly by developers and systems teams, Application Performance Management can save companies a lot of time and money.