
8/17/2021 · Darwin Sanoy
What this post added
This post introduces the "GitLab HA Scaling Runner Vending Machine for AWS," an Infrastructure as Code (CloudFormation) solution for deploying and managing elastic GitLab Runners on AWS. It focuses on cost savings through AWS Autoscaling Groups (ASG), Spot compute instances, scheduled uptime, and ARM architecture (Graviton). The post details how to configure jobs to leverage Spot runners and includes an example of using the `retry:` keyword for resilience. It also provides a cost savings calculation demonstrating up to 97% reduction by combining scheduling, Spot, and ARM.