Infrastructure Automation — Future Priority
Amen Ra flagged (2026-03-13) that we should look into Ansible, CloudFormation, or Terraform for managing QCS infrastructure.
Why: Current QCS setup is manual SSH + SCP + cron. As QCS count grows, this doesn’t scale. Need infrastructure-as-code for:
- Provisioning new QCS instances
- Deploying dispatcher/planner scripts
- Managing cron schedules
- SSH key rotation
- State file sync
How to apply: When planning QCS scaling or experiencing manual deploy pain points, recommend IaC tooling. Not urgent — current 3-QCS setup works. Becomes critical at 5+ environments.
Amen Ra’s Strategy (2026-03-13):
- Spin up a fleet of t3.micros (free tier eligible) instead of fewer t3.larges
- Use the free tier minutes, then destroy and spin up replacements — cost optimization via ephemeral infrastructure
- This is a disposable compute pattern: IaC spins up instances, runs workload, tears down when free tier exhausts
- Cuts costs significantly compared to persistent t3.large instances running 24/7
Candidates:
- Ansible — Simplest for config management, no state file, agentless (SSH-based). Good fit for current patterns.
- CloudFormation — AWS-native, manages EC2/IAM/SSM. Already using SSM Parameter Store. Best for auto-scaling groups + launch templates for the ephemeral t3.micro strategy.
- Terraform — Provider-agnostic, state management, good for multi-cloud if QCS expands beyond AWS.