Terraform for Serverless: Best Practices, Lessons Learned
summary
Shows concrete Terraform for serverless practices: split a monolithic repo into ~100 components and reduce 34,000 HCL lines to ~22,000 by converting to YAML, use TerraHub CLI for orchestration and cross-component parallelism, avoid local provisioner and external data, pass values via remote state, use workspaces for envs and rely on git diff optimization to run automation only on changed components (example cuts runs from nine to two). Demo deploys 7–9 of 100 components, with init/plan/apply/destroy and pre/post hooks for testing.
AI-written summary · about
chapters (11)
- 0:00Intro
- 0:46Introduction
- 5:15Microservices Architecture
- 6:18Terra for Remote State
- 7:11Avoid Local Provisioning
- 9:10Use Git Diff
- 10:11Separate Environments
- 11:19Variable Precedence
- 12:28Reduce Code
- 13:25Conclusion
- 14:48Demo
related talks
40mTerraform for Serverless. Best Practices. Lessons Learned.
24mTerraforming in a Serverless World
26mDoing Serverless On AWS With Terraform For Real
43mBuilding an AWS Serverless Application with Terraform
09mWorking with Serverless Functions and CI/CD
31mGoing Serverless For Fun and Profit
38mUnlocking AWS Serverless with Terraform
46mServerless Architecture Patterns and Best Practices