Success Story
Waking Up migrates from Rancher to Amazon EKS and gains control, scalability and security over its infrastructure
At a glance
Waking Up is Sam Harris's mobile meditation app, with more than one million users. Its Kubernetes clusters depended on Rancher, a layer the team could not manage and whose nodes died within a short time. Craftech migrated the microservices to Amazon EKS with Terraform, AWS Graviton nodes and security controls. Today its infrastructure responds automatically during high-demand events, with fully automated CI/CD.
Waking Up is a mobile meditation app by Sam Harris, a New York Times best-selling author. With more than one million users, it offers meditation courses for beginners and advanced practitioners. It is based in the United States and keeps growing worldwide.
Solution implemented
- Migration of the microservices to Amazon EKS with Terraform
- Infrastructure split across different organization accounts to isolate development from production
- Nodes on Reserved Instances with AWS Graviton (ARM64) processors
- Amazon CloudFront as a CDN, with origin access identity on S3
- AWS WAF and AWS CloudTrail for security
- Datadog Cloud SIEM to audit and alert on IAM events
- Monitoring, logging and alerting stack with Datadog for infrastructure and business metrics
Business Challenge
Waking Up provisioned and managed its Kubernetes clusters with Rancher on AWS, but the team had no access to or control over that layer. In addition, the nodes Rancher provisioned were not suitable and died within a short time, causing infrastructure shortfalls. It had to migrate, in record time, to a platform with more control and adopt a DevOps culture with a faster deployment cycle. The goals were to gain autonomy in managing the infrastructure, move it to code, get billing visibility by environment and product, achieve scalability and high availability, optimize costs, add controls to detect security events, and have real-time monitoring, alerting and auditing.
Solution
Together with the Waking Up team, Craftech migrated the microservices to Amazon EKS with Terraform and split the infrastructure across different organization accounts to isolate development from production. For the nodes, it chose Reserved Instances with AWS Graviton (ARM64) processors, which considerably reduced instance costs and optimized performance. It added Amazon CloudFront as a CDN, with origin access identity on S3. For security, it implemented AWS WAF and AWS CloudTrail, plus Datadog Cloud SIEM to audit and alert on IAM events. It also set up a monitoring, logging and alerting stack with Datadog for infrastructure and business metrics.
Outcomes
- Better scalability: the infrastructure responded automatically during high-demand events
- Fully automated CI/CD with GitHub Actions
- Automatic database migrations with Helm on every deployment
- Code instrumentation for business logic
- Early alerts on events and performance metrics
- VPN for secure access to private resources
Stack & technologies
