Application Monitoring Overview

In my previous article on network monitoring basics, we did an overview of some of the network areas that you might want to monitor, and some of the tools available to do the job.  In today’s article, we’ll discuss monitoring applications, which is complementary to network monitoring

Network & Application Monitoring Are Not the Same!

If your organization is like most, putting in the request for application monitoring after the request for network monitoring will probably get you a statement something like: “Didn’t we just buy a monitoring solution?”

How can you explain the difference between application monitoring and generalized networkmonitoring?

One easy way to explain it is to point out that network monitoring tells you when people can’t get toan application, application monitoring tells us when an application is not working properly, even though people can get to it.

Application monitoring will let you know when your main line of business apps, or their related databases, email system, etc… are not performing properly.  Proper app monitoring software will give you a visual dashboard to trend usage, performance, and growth.  All these are extremely important for capacity planning, meeting SLA’s, and finding problems before they cause outages.

App Monitoring Options

The Application Monitoring field is large, and there are even a couple of mature, well used open-source options, such as Nagios and Hyperic which provide powerful monitoring solutions for all sizes of business.  In addition to these open source options, there are several commercial options available as well.  Solarwinds has their powerful APM module for their Orion integrated management system, which can monitor your applications without an agent installation.  There’s also options from ManageEngine, which work in a similar fashion to the Solarwinds product line.

How it Works

All Application monitoring packages work in one of two ways, using either an agent-push, or apolling method.  In some cases, they use a combination of both methods.  Agent-push uses a locally-installed monitoring agent on the server to push monitored data, while the polling method uses WMI, SNMP, or other similar methods of gathering application information.

Historically, agent push has had more powerful monitoring options, but it introduces a management headache, as you have to install and update agents to the monitored servers.  Polling is becoming the preferred method, due to the ease of deployment of the monitoring solution.

Summary

App monitoring is another essential piece of the puzzle to ensuring your infrastructure stays operational, you know how your applications are running, how they’re being utilized, and how they’re growing.  With this data, you can more easily plan equipment acquisition, find problems with your applications easier, and know that your clients are able to complete their transactions smoothly.

Don’t forget or neglect this crucial piece of IT infrastructure.  When you’re in a bind, your application monitoring package will probably save your bacon.