Observability in 3 minutes

Over the last few years, computer science has made advances in moving from monolithic systems to cloud-native or hybrid applications using virtual machines, containers, and serverless technologies. Typically, modern applications have hundreds to thousands of distributed micro-services in different languages.

Modern applications are becoming more and more complex. Source: https://devops.com/author/breselman/

For simple systems with a single application — if a user knows what to monitor and the possible root cause of a problem is predictable, a simple monitoring tool can very well work. But as the system becomes complex (which is typical for modern applications), there are so many touch-points within the application that it becomes hard to predict what can go wrong. And if troubleshooting takes time — that would ultimately impact the customer experience and could result in loss of business. Proactively measuring system availability, security, performance, and cost where one cannot predict the problems in advance is the power of Observability.

In general, Observability has three pillars:

1. Metrics: These quantify the system behavior over a period of time. Eg: DNS response time, load time, requests per seconds for a website. These metrics can be represented in different types of dashboards for easy viewing and analysis of the data. Let’s take an example of a shopping website (a micro-services based application) and we see a spike in the requests per…