IT operations uses a multitude of specialized tools to keep the modern data center and IaaS deployments functional. Displays relay server, network, application and other performance statistics and health reports that enable IT admins to keep operations running smoothly. Yet, with all of this technology, serious issues still arise.
The origins of failures in IT vary widely, but the real issue is that operations didn't anticipate them. While human factors contribute to dysfunction, IT monitoring software shares that responsibility.
A successful IT and application monitoring strategy demands the right set of tools, as well as a focus on the user experience.
Know the strengths and weaknesses
Every IT environment is unique, which means that there isn't one tool to rule them all, and that's the root of the overall problem with insufficient or ineffective monitoring. While many IT monitoring tools operate similarly, each product often possesses a specialty. For example, Microsoft tools tend to monitor Windows installations better than Linux installs.
IT monitoring software vendors focus on key tasks at which the tool excels and also build secondary functions -- tasks the tool is designed to perform but are not its primary intention. For example, networking monitoring tools can oversee OS events but are not the ideal choice for organizations that require granular OS monitoring. This complication leaves operations with a critical decision between an overall tool package that monitors everything at some level or a collection of diverse tools from different vendors, where each keeps track of a specific area at an ideal level. Most likely, IT operations has already made a decision and now must manage the effects.
Approach tool replacements carefully
Just because a tool isn't optimal for a given task doesn't mean it must be replaced. Understand the intended use of a given tool, and balance that knowledge with the pros and cons that affect other systems and the overall IT environment. For example, a tool that isn't designed for Windows can require additional configuration to monitor the desired data, but that effort can still be more cost-effective than replacing it -- and switching tools might incur data losses or performance degradation.
Regularly review the tools and procedures in place across the IT organization. One additional step to make a tool work as desired easily turns into several additional steps over the course of time, as well as an increasing amount of work. When the effort eclipses the yield, consider replacements.
Consider the user's perspective
In addition to the ideal toolset, an application monitoring strategy requires admins to examine issues from the user's or customer's perspective, rather than the server or hardware performance angle. For example, horizontally scaled-out applications mean that a single node with performance issues is less important for overall application health than it was in a vertically hosted application setup, but those single nodes can make up a significant portion of the puzzle that broad performance monitoring tools miss.
The same logic applies to IT performance reporting. Daily and weekly reports quickly lose relevance as the email alerts flood in from the data center. Do you spend more time deleting the reports and emails daily than reading them? Once you reach that tipping point, take action to reduce the volume of alerts, or obsolete data could bury you.
Should these warning signs incite a switch or new buy, realize first that a tool is more than just a tool. Look for an overall system to monitor applications and their environments that also suits the business. Listen to customer concerns, and relate them to where the IT infrastructure might have failures or bottlenecks. Choose monitoring software that can capture trends in performance and how issues in the infrastructure affect the customer.
Build upon existing monitoring and support tools to prevent too much overlap and excess effort in IT admins' days. An application monitoring strategy relies on tool choices that make sense from both a technical and business sense, and new additions and optimizations must be worthy of operations' staff efforts.