3.3 Tips & Best Practices

Tips & Best Practices

Set Up Budget Alerts Immediately

This is the single most important thing to do after getting access.

  1. Search for Budgets in the console
  2. Click Create budget → Cost budget
  3. Set a monthly amount and add alerts at 50%, 80%, and 100%
  4. Add your email for notifications

Cost Awareness

Cost Factor Description
Compute EC2 instances — charged per second while running
Storage EBS volumes, S3 — charged continuously
Data Transfer Data leaving AWS — charged per GB
Idle resources Stopped instances still incur EBS charges

The golden rules:

  • Stop instances when not in use
  • Delete resources you no longer need
  • Start small, scale up based on actual usage
  • Check costs weekly via Cost Explorer

Resource Cleanup Routine

Daily

  • Stop EC2 instances not actively in use

Weekly

  • Delete unattached EBS volumes (EC2 → Volumes → filter by “available”)
  • Release unused Elastic IPs
  • Check for resources in other regions

When Finishing a Project

  1. Terminate all instances
  2. Delete EBS volumes, snapshots, and S3 data
  3. Release Elastic IPs and delete security groups
  4. Notify the RACE team

Resource Tagging

Tag everything you create so you can track costs and ownership:

Project: MyResearchProject
Owner: your.email@rmit.edu.au
Environment: Development

Security Essentials

Do Don’t
Use federated login only Create IAM users without approval
Restrict security group ports Open 0.0.0.0/0 on all ports
Encrypt EBS volumes and S3 Store credentials in code
Use private subnets where possible Leave unused security groups around

Common Mistakes to Avoid

Mistake Fix
Instances left running overnight Set calendar reminders; use Instance Scheduler
Orphaned EBS volumes after termination Check “Delete on Termination” when launching
Resources in wrong region Always verify region before creating
Oversized instances Start with t3.small, scale up based on CloudWatch metrics
No budget alerts Set up on day one