What is a router?

A router is a device that connects two or more networks or subnetworks. It mainly manages "routes", traffic between networks, forwarding data packets to their IP addresses, and allows multiple devices to share the same internet connection. The latter is a function greatly utilized in home router setups.

There are several types of routers, but the most common are routers that pass data between LANs (Local Area Networks) and WANs (Wide Area Networks). Largely this means routing traffic from multiple, local, devices, all in the same network, to the internet and back. A router differs from a switch, a device that is often grouped together with, in routing traffic across networks, while switches operate within a single network. We discussed switch monitoring in a separate guide.

Mainly routers are split into wired and wireless. Wired routers connect networks through cables (commonly Ethernet), while wireless ones create WLANs (Wireless Local Area Networks), connecting multiple devices via wireless communication. Furthermore, routers are also split into core and edge ones. A core router is usually put at the center of a local network and is apt to transmit large amounts of data to other LANs. It usually does not connect directly to WANs. In large companies these routers are interconnected through fast optical fiber lines to ensure speed and throughput. Edge routers instead are those which communicate between LANs and WANs, exposed to the internet, and sometimes called border or gateway routers.

Whatever their type is, routers use routing protocols to know about the destination address of packets, referring or building a routing table to quickly identify where on a network an IP is located. That is their primary role in an infrastructure, and the one from which they take their name.

What is router monitoring?

With what is a router hopefully clear, we can move on to what is router monitoring. Monitoring routers is the sum of practices and processes implemented to monitor the health of the device itself, along with the traffic that it routes. This involves using monitoring protocols, like SNMP, that have an agent pre-installed in the router. Not all routers have it, and not all routers have a SNMP-enabled one, but in broad terms either using the agent that came with the router or installing one yourself is how router monitoring is performed.

These agents communicate with a router monitoring software that supports the same monitoring protocol. The king of them is SNMP, and the one that most probably you will use for router monitoring. But recent alternatives like NetFlow, sFlow, IPFIX, WMI and more are increasingly supported. A router monitor should ideally be able to talk with the routers in all these protocols for maximum coverage.

The agents take care to collect the info about the router and the traffic running through it. Then a router monitoring software, like our Checkmk, polls the agents to gather the info, elaborating them, and alerting of malfunctioning or other problems the IT administrators. That is in simple terms what router monitoring is. We will next see what should be monitored, and why.

Why should I monitor my routers?

Applications and devices nowadays use a lot of network traffic to communicate between each other, connect to external resources, and provide end users their services. All this traffic goes through multiple networks and subnetworks, and routers play a vital part in ensuring that the traffic goes to the right IP address and that the networks operate correctly.

A malfunctioning router may cause slower traffic, disruptions, and increased latency. These can have catastrophic consequences on your infrastructure, especially when client-facing services are affected. Not all of them can be caused by a router only: it may be an application that is using a disproportionate amount of bandwidth, making other services using the same router slower. By implementing a router monitoring solution you can identify what application is the culprit and differentiate problems relative to the health of the routers themselves to those caused by connected applications and services.

Router monitoring is key to know how healthy and performing are your routers, which in turn helps to keep your networks working. Paired with the equally important switch monitoring, monitoring routers is an obvious endeavor to take for any serious organization.

How to monitor a router

We have briefly touched how monitoring a router is done in practice, but we will go in more detail now. Most home routers or routers in small networks have an interface that, once logged in, can show you basic info about mainly the traffic going through the router. This is a barebone router traffic monitor setup that anybody can use from time to time, perhaps to troubleshoot sudden drops in internet speed. For any middle-sized and above companies, this is obviously too simple and limiting.

The vast majority of routers nowadays come with support of monitoring protocols through pre-installed monitoring agents. These are unmodifiable, with a few exceptions, but generally enough to get started with monitoring your routers. Router monitoring is therefore primarily done with the help of these agents and monitoring protocols. SNMP most of the time, but NetFlow, sFlow, IPFIX, and WMI, just to name the most widespread ones, provide a great deal of information about routers for their monitoring, and have a more detailed router traffic monitor view.

These protocols are generally to be used from the command line but UIs exist. Router monitoring software, like Checkmk, include support for more than one monitoring protocol, thus being able then to monitor many types of routers with a single software. Fragmentation is a problem among routers (and switches too), as different vendors support different protocols. A monitoring software that supported only SNMP, for example, would be capable of monitoring many routers but far from all. Thus, most router monitoring software support communication with monitoring agents in multiple protocols, making router monitoring much less complicated than it has to be.

Furthermore, monitoring through agents and relative protocols means not just router traffic monitoring but includes a wealth of info about the actual health of the device like CPU usage, memory utilization, and even temperature sensor reading. IT administrators collect these metrics and compare it with general traffic patterns to identify bottlenecks, bandwidth hogs, and prevent disruptions. A router monitor software basically includes functions of a router traffic monitor, router bandwidth monitor, and general health of the routers, and by extension, of the networks.

Besides the router itself, it is also important to monitor its ports for faulty connections or loose cables. Port monitoring is a whole sub-area of router and switch monitoring, which warranted its own guide.

Advantages of router monitoring

Setting up a router monitor in your company to check how your routers are operating presents several advantages. By comprehensive router monitoring in an infrastructure it is possible to discover lots of problems. These go from an application using a high amount of bandwidth, reducing it to other applications and causing the network to slow down, to identifying security threats by comparing the usual network traffic with, implied, unusual activity. Router monitoring helps IT administrators have better visibility in their infrastructure, being more aware of how the networks are performing, and spotting usage trends that can give important insight into future capacity expansion.

Routers fail too, as any hardware, and their failures can be catastrophic for the service of your company. Preventing them is a big part of router monitoring. Regularly checking the health of the routers is a must for any IT administrator.

Router monitoring helps in combating shadow IT too. Router monitor software such as Checkmk have auto-discovery features that automatically discover nodes on your network with their relative devices and services. By monitoring all your routers, given their central position and role in a network, you can discover more easily resources that may otherwise escape your mind when setting up a monitoring solution. And by monitoring them you are aware of how they are performing, and spot issues.

And any issue can be reported immediately through alerts and notifications. If a device cannot connect to the internet, or a router goes offline for any reason, it is important to be promptly alerted. Checkmk has a vast alerting system, easily customizable to your specific needs, and router monitoring is definitely one of the areas that you want to be alerted on.

Metrics to look for in router monitoring

There are a handful of metrics that are especially key to be aware of in router monitoring. These are those which can give you the best insight in how your routers are performing, and if there are points in your networks that need your intervention.

Firstly, make sure to check general router availability and status. These are definitely available in any router monitoring software. Set up alerts connected to your routers going down, so you can quickly act on network disruptions.

Any metric that tells you about packets is important. These include packet loss, duplication, and reordering. The first are packets that did not arrive at their destination, the second those which arrived twice, and the last are packets that arrived in an order different from expected. These may be caused by a hiccup in the network connection or be due to router misconfiguration (check your QoS rules, if any), or sign of something more serious.

Set up your router monitor to check metrics like network bandwidth, throughput, and speed. These can fluctuate a bit, but when larger drops happen, especially if for increasingly longer bursts of time, there is something that needs to be checked in your network.

Depending on whether you are using them or not, monitoring your router QoS (Quality of Service) rules is important to avoid misconfigurations. Check info about your flows (for monitoring through flow-based protocols like NetFlow or sFlow) to identify traffic patterns and trends, to know how and when to expand your network resources to sustain increased demand.

Also do not forget to check any metrics that tell you about the health of the router themselves. CPU and memory usage, temperatures, fan speeds and so on are vital to prevent hardware failures of such a critical part of your infrastructure.

These, as a minimum, are the metrics to monitor in router monitoring. Depending on the router's vendor and router monitoring software used some may not be possible to monitor. But if so, make sure to set up monitoring them. With its over 2000 out-of-the-box plug-ins, it is highly probable that Checkmk can monitor your specific router and metric combination.

FAQ

What is router management

Router management is the process of configuring and monitoring your routers. It supersedes router monitoring then, including capabilities like configuring routers on remote, backing up the configuration, and automating setting up new routers.