Welcome to Virtualization and Beyond
More data, more problems? Not with the right context.
Written by Chris Paap, Technical Product Manager, SolarWinds
When it comes to monitoring your IT infrastructure, especially virtual environments, conventional wisdom dictates that the more data you can gather, the better. In fact, you'd be hard pressed to find a systems engineer who would argue against having more monitoring in their environment. You likely fall into this camp.
Following that logic, the more monitoring the better prepared we should be to address performance problems and general outages. So, why is it still so difficult to quickly identify not only where problems reside in our infrastructures, but what exactly those problems are, which would lead to solving them once and for all, regardless of the amount of monitoring we have in place?
The answer is that the monitoring needs of today are evolving beyond passive monitoring. Monitoring that simply collects data and provides counters is no longer adequate. As with any modern data collection endeavor, the critical task is to determine what data is truly important and then taking action as a result of the most relevant data.
Therefore, having the ability to receive important context along with counters and alerts helps expedite the process of identifying root causes. After all, receiving an alert or warning is meaningless if you don't know what the alert means. Equally important is the need to identify and correlate how performance problems affect the rest of your environment. In essence, any monitoring you put in place should not only provide the "what," but ideally also the remediation path and steps to prevent future occurrences of the same problem.
This migration from a reactional approach to IT to a preventative/proactive one requires monitoring that provides intelligent actions. These actions should be prescriptive to remediate current issues (a.k.a. tactical remediation), or after-the-fact recommendations. Such recommendations assist in eliminating the firefighting that consumes many an administrator's day, eating up valuable hours that could otherwise be spent on bigger projects. Such recommendations should also be predictive in nature, so that they can identify workload utilization patterns and discern anomalies versus true workloads. This will enable infrastructure optimization that provides the highest operational efficiency, while also helping to prevent potential performance issues from occurring.
Monitoring that provides context, relationship and performance correlation capabilities combined with troubleshooting and proactive recommendations becomes a hybrid operational tool that is much more sophisticated than yesterday's monitoring software. However, this added level of sophistication may also invite complexities that need to be balanced with functionality. There is little value in monitoring that has the capability to solve every problem if the monitoring is so complex that it becomes unusable.
In closing, any operational monitoring you implement should provide context and performance correlation along with core monitoring results so that it becomes easier for you to identify the what, why, and how (as in, how to fix the problem). These are the new table stakes.
##
Read more articles from the Virtualization and Beyond Series.
About the Author
With 14 years of IT systems engineering experience across multiple corporate environments, Chris Paap currently serves as a technical product manager for hybrid IT performance management software provider SolarWinds, where he focuses specifically on the award-winning SolarWinds Virtualization Manager. In this role, he is responsible for defining the product roadmap and identifying new, key features to solve IT problems.