From the course: DevOps Foundations: Monitoring and Observability

Overview of monitoring and observability

- [Instructor] Observability and monitoring are often used interchangeably. However, while these two concepts are related, they're also very different. Observability is a characteristic of a system that makes it easy to understand its inner workings through the signals it emits. Monitoring, on the other hand, is the use of these signals to visualize or alert on system performance. By practicing observability, we're able to uncover known and unknown problems that lie within our services. Since we emit both good and bad signals. While monitoring could potentially be targeted at known problems, meaning that there could be a plethora of problems we may not be uncovering. In a nutshell, monitoring and observability go hand in hand. With monitoring, we can see or be alerted to what is wrong. While with observability, we can dig deeper into why things went wrong depending on the data we're getting from our services. Observability relies on signals, which are called the pillars of observability. They are metrics, logs, and traces. However, there's another class of signals called events, which as the name implies emits the events happening in your system. Having coverage of these pillars gives high visibility into system performance. Monitoring uses these signals to paint a picture of what is happening within our services, either using dashboards or visualizations or alerting. There are several monitoring techniques depending on which aspects of a service is being monitored. Usually having more than one of these techniques in place gives a full picture of service performance. For example, having application monitoring alone for your backend services doesn't quite provide insight into your frontend performance, which is equally an important part of your user experience. We'll explore these and a few others as we progress in the course.

Contents