I got involved in a conversation at the whiteboard the other day where we were talking about the value of managing network devices via their loopback address. It occurs to me that this may be a best practice that many of you haven't implemented so I thought I'd write a little bit about why it's important and how it can help you.
For the sake of keeping things simple, let's talk about managing/monitoring routers. As you know, a router will typically have several IP addresses - at least one per interface and you can usually use any of these addresses to communicate with the router from your NMS. Now, let's assume that you use the IP address of interface serial 1/0/1.1 for monitoring of the device. What happens if that interface is down? Suddenly, your NMS thinks that the entire device is offline and you'll get alerts that the router is down and the NMS will stop collecting the data for the other interfaces as well as the router level statistics like latency, CPU, memory, and buffer usage.
So, how do you get around this issue? You implement a loopback addresss on all of your routers and you be sure that these addresses are reachable throughout the network. A loopback address is basically a virtual interface at the core of the device. As long as you have a path to the device, the loopback address should be reachable.
This best practice can save you a lot of time and effort down the road if you implement the loopback addresses before upgrading or deploying a new NMS.
Follow me on Twitter