DevOps culture and Kubernetes migration for Ripio: infrastructure as code, CI/CD and centralized observability

Challenge

Ripio needed to bring the DevOps culture of its platform and infrastructure to a level that would let them deploy more safely, frequently and in an orderly way. The starting point had clear friction: there was no working local environment for team onboarding, the Kubernetes clusters were self-managed and unmaintained, there was no infrastructure as code and no platform documentation, CI/CD pipelines were outdated, many Docker images had no versioned Dockerfile, and there was no monitoring tool to provide visibility when problems arose.

Solution

Craftech built Terraform modules that allowed the team to update and replicate portions of the infrastructure in a controlled way, migrating the solution to Amazon EKS with infrastructure written as code. It standardized and updated all container images to versioned builds from Dockerfile, with dependencies aligned to current security requirements, so engineers could deploy and update more safely. It created a dockerized, extensible and fully documented local environment to speed up onboarding of new engineers, and rolled out Datadog across all services and servers to deliver centralized, real-time visibility of logs, metrics and traces on platform health.

Results

Ripio ended up with its infrastructure 100% as code and with the lowest possible management overhead, and with fully automated application deployments via CI/CD pipelines. On observability, it gained a reliable system that monitors its systems and alerts the team to any potential issue. On scalability, since the EKS implementation the platform was able to grow quickly, efficiently and cost-effectively alongside the expansion of the business.