DeveloperWeek New York 2019

Like many large enterprises, Intuit faced significant limitations to test and assure resiliency across complex legacy systems. With the recent migration to public cloud such as AWS, our teams began to proactively test and assure resiliency as part of software and infrastructure pre-release pipelines. By automating resiliency earlier into the product development cycle, we significantly reduced testing cycles from many days of manual efforts to just a few hours via an automated pipeline. This new automation proved invaluable during our tax peak, allowing the team to recover from an incident in just a few minutes. Today, the team is able to programmatically automate test failures in various AWS resources (e.g. Load-balancers, EC2, RDS, DynamoDB, etc.) while also validating system resiliency, alerts, and monitors. As a cloud agnostic enterprise, Intuit utilizes container orchestration system such as Kubernetes and has successfully extended this automation in the cloud. In this talk, we will share how Intuit automated resiliency through Cloud-raider, an in-house tool offering natural language construct for engineers and architects, enables behavior-driven development for resilient systems.