Aflevering 68: How an outage sparked a full rethink of Essent’s cloud architecture

In episode 68, Ronald and Jan chat with Nicolas Braga from Essent about how a major service outage, caused by a surge in traffic during the Ukraine war and rising energy prices, led to a full rethinking of Essent’s cloud architecture.
Nicolas explains how they shifted their strategy to increase team ownership and why they chose AWS Lambda over Kubernetes. He highlights Lambda’s advantages, such as auto-scaling, cost-efficiency, and simplified management, which allows Essent to innovate faster without worrying about infrastructure complexity.
The episode explores how these changes align with Essent’s goal to become a leading energy tech company.

[@edikmkoyan]

The concept of information hiding also helps to set proper domain boundaries, thus your services become loosly coupled. There is another solution tough, called reporting database.

[@edikmkoyan]

Why did they choose to serve millions of requests with lambdas?

[@edikmkoyan]

Honestly lambdas make sense only when something needs to be tested quickly in a very early stage of any project.
Lambda spin up an entire environment to serve a tiny number of requests, in most cases it is 1.

[@edikmkoyan]

even when the exact volume of the load is hard to predict but it does not drop to 0, lambda should not be used.