Monitoring Network Switches with Grafana

MetricFire
6 min readJul 19, 2023

--

Introduction

In monitoring, a target system or device is a deciding factor in designing your monitoring stack. You will have to consider various aspects starting from how you want to collect data in what frequency to how you want to surface metrics to end users. You will have to take this strategic approach when you want to monitor your network infrastructure. In this article, we will discuss how Grafana, an open-source visualization tool, can help you to monitor network switches.

Before we start, check out MetricFire. We provide hosted services not only for the visualization tool but also for other comprehensive monitoring platforms such as Graphite. Using our services, you can view insightful metrics for network switches with minimal configuration.

If you would like to learn more, please book a demo with us, or sign on to the free trial today.

What are network switches?

Network switches are physical devices that connect devices together such as computers, printers, and servers. The switches allow those devices to share information and communicate with each other. With switches, you can build a network of devices and you can have more flexibility in where you want to put devices.

Due to the very characteristics of connecting devices, sometimes, people confuse switches with routers. To clarify what makes them different, a switch connects devices to build a network, whereas a router connects multiple switches, which allows to create an even larger network. The networks built by routers can be in a single location or multiple locations. When you build a small business network, you need one or more routers. Routers, in addition, can enable networked devices and users to have an Internet connection.

In network switches, there are two main types.

  • Unmanaged network switch: this type does not require configuration. You can plug devices in and start using the switch and the devices. The unmanaged switch is usually for a basic network such as the one for your home or areas where only a few ports are needed.
  • Managed network switch: the managed type allows you to have greater security and utilize additional features and customization. This type is more flexible than unmanaged ones since they can be changed to meet the needs of your network. You can have more control and use them for larger networks.

Why do we want to monitor network switches?

Undoubtedly, network switches perform important tasks in your network infrastructure. Broadly, we now know that we want to monitor network switches since they are important. Let’s start with what network switch monitoring means. Then, we can discuss what we can achieve by monitoring switches and what will happen when we fail to do so.

Depending on your profession, the interpretation of network switch monitoring can differ. However, there are two common monitoring aspects. First, you will want to ensure that the devices that are being monitored behave normally. Second, you will want to track how traffic is handled by the devices. Combined with these two, you can monitor the performance and efficiency of network switches.

Important metrics for network switches

There are a variety of metrics you can focus on for effective network switch monitoring. Among many, we will introduce some of the major metrics you may want to consider.

  • CPU usage: like computers, network switches, especially managed ones, rely on the CPU to compute to decide traffic flow and ensure smooth interactiveness among the connected devices. When CPU usage spikes abnormally, you will want to look at the symptom from two perspectives — performance and security. When CPU usage increases, the overall performance will slow down. Also, CPU spikes can be caused by too much traffic, which could be caused by a hacking attack.
  • Memory usage: like CPU usage, having not enough memory can lead to deteriorated performance and exceptions. By exposing memory usage on dashboards, you can maintain optimal network performance.
  • Fan status: when you use highly performant switches, they can have fans to cool down the temperature. Just like your laptop gets slow when its temperature increases, switches are affected by their temperature. You will want to monitor the temperature and also the status of the fans.
  • Incoming traffic: as switches receive requests from external sources, measuring incoming traffic can help you to understand the workload that the switches need to handle. If there is too much traffic that a switch cannot handle and it occurs regularly, you will want to take measures such as adjusting routers to distribute the traffic or increasing the capacity of the switch. In addition to the amount of traffic, you may also want to track incoming error codes that your switch receives.
  • Outgoing traffic: looking at outgoing traffic can help you to understand how well your resources perform. When you see error codes, for example, 4xx or 5xx, you will want to zero in which server caused that error. Knowing the source of an error and the cause lets you take swift corrective actions, which increases service quality.

What is Grafana?

We found out about network switches and major metrics you may want to focus on. The next step is how we want to expose that critical information to end users effectively. Grafana is an open-source analytics and interactive visualization web application. You can customize charts, graphs, and alerts. On top of its visualization features, you can design its back end. With its extensive features, you can create customized dashboards and run queries to generate the metrics you want. The major features of Grafana include the following.

  • Dashboard templates: this means that you can write your dashboard design in templates as if you write code. This can be more effective than creating a dashboard by clicking and dragging and dropping since templates allow you to reproduce the same design across multiple environments such as dev, uat, stage, and production. With templates, you can deploy your template files systematically.
  • Annotations: when you use to discover some patterns in your graph, you will sometimes flag them and leave comments to add explanations. In Grafana, you can not only create a manual annotation but also set up an automatic annotation. This means when a certain event happens, a configured annotation will be added to mark it.
  • Custom plugins: when you want to extend more features, you can rely on external plugins. These plugins enable you to, for example, add more graph types and manage dashboard users on a deeper level.
  • Alerting: notifying users can be considered a simple feature but without alerting, you can lose your sight on important incidents. You can customize alerting criteria and decide to whom you want to send it.
  • SQL support: with SQL support, you can be flexible to retrieve metrics data from various database sources.

What is a Grafana as a Service?

Open-source software is often chosen by many people since they don’t have to pay any cost. However, it comes with a price actually. When you use open-source software, you may have to spend a long time learning about the tool and actually starting to use it. Even when you finish setting up, maintaining it can be your next concern.

Doing your everyday task while maintaining an open-source tool can be a challenge. This applies to Grafana too. Since Grafana is an open-source tool, it may be not easy to complete the initial setup and keep maintaining it.

MetricFire, to reduce your workload, offers Grafana as a Service. Using this, you don’t have to worry about maintenance and focus on what matters the most to your team. You can enjoy all the good features of Grafana and, with our managed service, you can:

  • own and control your data
  • use MetricFire’s flexible pricing plans
  • receive expert engineering support

With these additional features, you can start creating metrics to monitor important metrics of network switches including CPU, memory, fan status, and incoming and outgoing traffic.

Wrapping up

In this article, we learned about network switches and the types of metrics that are important to us. Also, we discussed the conveniences of Grafana as a Service.

Using Grafana as a Service, you can focus on monitoring network switches more efficiently without having to worry about maintaining a monitoring infrastructure. Our hosted Grafana service gives you more control over your data and lets you have an optimal monitoring stack.

Our expert support will lend helping hands whenever you need one.

To access all these services, visit MetricFire today and check our Graphite as a Service.

Try MetricFire now!

Get MetricFire free for 14 days. No credit card is required.

Get Started

--

--

MetricFire

Time series monitoring as a service using Graphite and visualized on Grafana.