Synthetic Monitoring

A new way of monitoring in Enterprise products and services.

Apart from Observability, telemetry tracing, performance monitoring the neglected concept of Synthetic Monitoring has started to thrive in DevOps and SRE community.

Monitoring the services end-to-end by emulating a user like action.

Consider that I have a web application where the user’s are actively using it with 2000+ TPS during peak hours, and the application is deployed globally in 10+ regions. Given the scale of the application, I take into consideration that all the traditional observability metrics have been setup and actively monitored. But the challenge is, monitoring all these regions is quite challenging for the support team.

Well, here Synthetic monitoring comes with added benefits. By emulating a user action, such as placing an order, we can test features like login, search, adding to cart, and placing an order. This ensures that the application is kept from cold starts and that the core functionalities of the product/service are also tested.

Moreover, this methodology enables us to inject and measure metrics from Grafana, Prometheus and trigger alerts if there is an abnormality in the latency of operations. This could potentially reduce system failures. Overall, this is a simple but interesting solution to strengthen application monitoring.

Cheers!