Practical Monitoring

Chapter 9. Network Monitoring

Network monitoring has a special place in my heart. Starting with my first job in tech, I’ve been fascinated with how networks worked. I quickly discovered the importance of monitoring: one day while removing some old gear from a closet, I accidentally knocked out the power plug to a switch that was perched precariously on a desk. Given that everyone else had gone home already, this went unnoticed until the next morning, when dozens of people were unable to check their email. I quickly fixed the issue and then searched online for something like “monitor network switch” and then set up Nagios. I’ve been hooked ever since.

Over my next several jobs as a systems administrator, I gravitated back to network engineering and network monitoring. One thing I’ve learned over the years is that the behavior and performance of the network is fundamental to the behavior and performance of everything that relies on it—which, these days, is everything. If your network is only capable of maintaining three nines of availability (99.9%), then your applications can’t possibly maintain four nines (99.99%). Increasing the availability of the network is a nice lever you pull to allow everything that relies on it to improve.

Networking is one of the few “dark arts” left in the tech world. So many people don’t understand it, yet it’s a crucial component of everything we do. To the network engineers reading this chapter, this will likely be a recap of ...