Cost Management Guide

DevOps Studio › Docs › Cost Management

Strategies for managing and optimizing costs in DevOps Studio labs.

Understanding Costs

💰 Important Cost Note: The costs below are monthly estimates if you keep infrastructure running continuously. Actual costs to complete a lab and delete it are significantly less - typically 5-10% of monthly costs depending on how long you keep resources running. For example, completing Lab 01 in 2-4 hours and then destroying it would cost approximately $2-5 instead of $50-80/month.

Estimated Monthly Costs vs Completion Costs

Monthly costs assume resources run continuously for a full month. Completion costs assume you run the lab for the typical completion time and then destroy all resources.

Lab	Resources	Monthly Est. Cost (USD)	Expected Cost to Complete*	Typical Time	Free Tier Eligible
Lab 01	VPC, ASG, RDS, ALB	$50-80	$2-5	2-4 hours	Partially
Lab 02	+ EKS cluster	$120-180	$5-10	3-5 hours	No
Lab 03	+ CI/CD resources	$20-40	$1-2	1-2 hours	Partially
Lab 04	+ Monitoring stack	$40-60	$5-8	4-6 hours	No
Lab 05	+ Security tools	$30-50	$2-3	2-3 hours	Partially
Lab 06	+ GitOps tools	$20-30	$1-2	1-2 hours	No
Lab 07	+ Serverless	$25-45	$1-2	1-2 hours	Partially
Lab 08	+ Platform tools	$80-120	$3-5	3-4 hours	No
All Labs	Complete platform	$200-300	$20-40	15-25 hours	No

* Expected cost to complete assumes you run the lab for the typical completion time (shown in "Typical Time" column) and then destroy all resources. Actual costs may vary based on your completion time, AWS pricing, and region.

Costs are estimates for us-west-2 region. Actual costs may vary.

Cost Breakdown Example (Lab 01)

Monthly Cost (if running continuously for 30 days): ~$50-80

Resource	Configuration	Monthly Cost	Hourly Cost*
EC2 Instances	2x t3.micro	$16.00	$0.022
Application Load Balancer	Standard ALB	$22.00	$0.030
RDS MySQL	db.t3.micro, Multi-AZ	$25.00	$0.034
NAT Gateways	2x Standard	$45.00	$0.062
Data Transfer	Typical usage	$5.00	Variable
CloudWatch	Logs and metrics	$3.00	$0.004
S3 Storage	Terraform state	$1.00	Negligible
Total		~$117/month	~$0.15/hour

* Hourly costs are approximate and based on pro-rated monthly costs. Some services charge per-hour, others per-GB or per-request.

Cost to Complete (run for 3 hours then destroy): ~$2-4

Component	3-Hour Cost	Notes
EC2 Instances	$0.07	2 instances × 3 hours × $0.0116/hour
Application Load Balancer	$0.09	ALB charges per hour + LCU
RDS Database	$0.10	Multi-AZ charges per hour
NAT Gateways	$0.37	2 NAT Gateways × 3 hours × $0.062/hour
Data Transfer	$0.10	Minimal during lab completion
CloudWatch	$0.01	Minimal logs during short run
Total	~$0.74	Plus any data transfer charges

💡 Key Insight: Completing Lab 01 in 3 hours costs approximately $0.74-2.00 (depending on data transfer), which is about 1-4% of the monthly cost. The biggest savings come from destroying resources immediately after completion!

Cost Optimization Strategies

For Learning/Development (Minimize Costs)

🎯 Best Practice: Complete labs in focused sessions and destroy resources immediately after. This is the single biggest cost saver!

1. Destroy Immediately After Completion

bash

# Complete the lab (typically 1-4 hours)
make apply
# ... work through the lab exercises ...

# Destroy resources immediately when done
make destroy  # Per lab
# or
./tools/cleanup.sh  # All labs

Savings: This is the biggest cost saver! Completing and destroying a lab costs 5-10% of keeping it running for a month.

Example:

Lab 01 running for 1 month: ~$50-80
Lab 01 completed in 3 hours and destroyed: ~$2-5
Savings: 95-96%

2. Use Development Environment

bash

# Deploy with dev configuration (smaller resources)
make apply ENV=dev

Dev Environment Savings:

Smaller instance types (t3.micro vs t3.medium)
Fewer instances (1-2 vs 3+)
Single NAT Gateway option
Minimal storage

2. Scale Down When Not in Use

bash

# Scale to zero instances
terraform apply -var="desired_capacity=0" -var="min_size=0"

# Scale back up when needed
terraform apply -var="desired_capacity=2" -var="min_size=1"

Savings: ~$16/month per instance stopped

3. Use Single NAT Gateway

hcl

# In terraform.tfvars or dev.tfvars
single_nat_gateway = true

Savings: ~$22.50/month (one NAT Gateway instead of two)

4. Disable Deletion Protection

hcl

# For development only!
enable_deletion_protection = false

Allows easier cleanup and prevents accidental costs.

5. Use Spot Instances (Advanced)

hcl

# Modify launch template to use spot instances
# Can save 50-90% on compute costs
# Note: Instances can be terminated with 2-minute notice

6. Reduce RDS Costs

hcl

# Use single-AZ (dev only)
multi_az = false

# Smaller instance
db_instance_class = "db.t3.micro"

# Minimal storage
db_allocated_storage = 20

Savings: ~$12.50/month (single-AZ vs Multi-AZ)

7. Clean Up Regularly

bash

# Destroy infrastructure immediately after completing lab
make destroy

# Or emergency cleanup for all labs
./tools/cleanup.sh

💡 Pro Tip: Set a reminder or use a timer to ensure you destroy resources after completing each lab. Even leaving resources running overnight can add significant costs!

For Production Reference

1. Right-Sizing

Monitor Usage: Use CloudWatch metrics
Analyze Performance: Identify over-provisioned resources
Adjust Gradually: Start conservative, scale up as needed

2. Reserved Instances

For predictable workloads:

Standard RIs: Up to 72% savings (1-3 year term)
Convertible RIs: Up to 54% savings (flexible)
Savings Plans: Up to 72% savings (flexible compute)

3. Auto Scaling

hcl

# Configure proper scaling policies
min_size = 2
max_size = 10
desired_capacity = 3

# Scale based on metrics
scale_up_threshold = 70
scale_down_threshold = 30

4. Cost Allocation Tags

All resources are automatically tagged:

Project: DevOps Studio
Environment: dev/staging/prod
ManagedBy: Terraform
CostCenter: Engineering

Use tags for cost allocation and tracking.

5. S3 Lifecycle Policies

hcl

# Automatically move old logs to cheaper storage
# Or delete after retention period

6. CloudWatch Log Retention

hcl

# Reduce log retention for non-production
log_retention_days = 7  # Instead of 30

Savings: ~$1-2/month per log group

Cost Monitoring

AWS Cost Explorer

Access: AWS Console → Cost Management → Cost Explorer
View Costs: Daily, monthly, by service
Filter by Tags: Use Project=DevOps Studio tag
Forecast: Predict future costs

Cost Estimation Tools

Using infracost

bash

# Install infracost
brew install infracost  # macOS
# or see: https://www.infracost.io/docs/

# Estimate costs
cd labs/01-terraform-foundations
infracost breakdown --path . --terraform-var-file environments/dev.tfvars

Using Cost Estimation Script

bash

# Run cost estimation
./tools/cost-estimate.sh

AWS Budgets

Set up budget alerts:

bash

# Create budget (one-time setup)
aws budgets create-budget \
  --account-id $(aws sts get-caller-identity --query Account --output text) \
  --budget file://budget-config.json

Budget Configuration (budget-config.json):

json

{
  "BudgetName": "DevOps-Studio-Monthly",
  "BudgetLimit": {
    "Amount": "100",
    "Unit": "USD"
  },
  "TimeUnit": "MONTHLY",
  "BudgetType": "COST",
  "CostFilters": {
    "TagKeyValue": [
      "user:Project$DevOps Studio"
    ]
  }
}

AWS Free Tier

Eligible Services (First 12 Months)

EC2: 750 hours/month of t2.micro or t3.micro
RDS: 750 hours/month of db.t2.micro or db.t3.micro
S3: 5GB storage, 20,000 GET requests
Lambda: 1M free requests, 400,000 GB-seconds
CloudWatch: 10 custom metrics, 5GB log ingestion

Free Tier Optimization

hcl

# Use free tier eligible resources
instance_type = "t3.micro"        # Free tier eligible
db_instance_class = "db.t3.micro" # Free tier eligible

Note: Load balancers and NAT Gateways are NOT free tier eligible.

Cost Optimization Checklist

Before Deployment

[ ] Choose appropriate instance sizes
[ ] Use dev environment for learning
[ ] Set up cost alerts
[ ] Review cost estimates
[ ] Enable cost allocation tags

During Deployment

[ ] Monitor costs daily (first week)
[ ] Review CloudWatch metrics
[ ] Adjust resources based on usage
[ ] Clean up unused resources

After Deployment

[ ] Review monthly costs
[ ] Identify optimization opportunities
[ ] Right-size resources
[ ] Set up automated cleanup (if applicable)

Emergency Cost Control

Immediate Actions

bash

# 1. Stop all instances
terraform apply -var="desired_capacity=0" -var="min_size=0"

# 2. Emergency cleanup
./tools/cleanup.sh

# 3. Check for orphaned resources
aws ec2 describe-instances --query 'Reservations[*].Instances[?State.Name==`running`]'
aws rds describe-db-instances --query 'DBInstances[?DBInstanceStatus==`available`]'

Cost Anomaly Detection

Set up CloudWatch anomaly detection:

bash

# Monitor for unusual cost spikes
aws cloudwatch put-metric-alarm \
  --alarm-name "HighCostAlert" \
  --alarm-description "Alert when costs exceed threshold" \
  --metric-name EstimatedCharges \
  --namespace AWS/Billing \
  --statistic Maximum \
  --period 86400 \
  --evaluation-periods 1 \
  --threshold 100 \
  --comparison-operator GreaterThanThreshold

Cost Optimization Examples

Example 1: Development Environment

Configuration:

hcl

instance_type = "t3.micro"
min_size = 1
max_size = 2
desired_capacity = 1
single_nat_gateway = true
multi_az = false

Monthly Cost: ~$60-70 (vs ~$117 for default)

Savings: ~$50/month

Example 2: Production Environment

Configuration:

hcl

instance_type = "t3.medium"
min_size = 3
max_size = 10
desired_capacity = 3
single_nat_gateway = false  # HA required
multi_az = true

Monthly Cost: ~$200-250

Includes: High availability, auto-scaling, monitoring

Example 3: Cost-Optimized Learning (Recommended)

Configuration:

hcl

instance_type = "t3.micro"
min_size = 1
max_size = 2
desired_capacity = 1
single_nat_gateway = true
multi_az = false
db_instance_class = "db.t3.micro"

Monthly Cost: ~$30-40 (if kept running)

Cost to Complete (3 hours): ~$1-2

Usage:

Deploy lab
Complete exercises (2-4 hours)
Destroy immediately when done
Repeat for next lab

Total Cost for All 8 Labs: ~$20-40 (if completed and destroyed immediately)

vs Keeping All Labs Running: ~$200-300/month

Best Practices

🎯 Destroy Immediately: Complete labs in focused sessions and destroy resources right after - this is the #1 cost saver
Start Small: Begin with minimal resources, scale up as needed
Monitor Regularly: Check costs daily during active learning, weekly otherwise
Use Tags: Tag all resources for cost allocation
Set Alerts: Configure budget alerts at 50%, 80%, 100%
Time Your Learning: Complete labs during focused sessions rather than leaving resources running
Use Free Tier: Maximize free tier usage for learning (first 12 months)
Right-Size: Match resources to actual usage
Batch Learning: Complete multiple labs in sequence, then destroy all at once
Set Reminders: Use timers or reminders to ensure cleanup after lab completion

Cost Resources

AWS Pricing Calculator: https://calculator.aws/
AWS Cost Explorer: AWS Console → Cost Management
infracost: https://www.infracost.io/
AWS Well-Architected Cost Optimization: https://aws.amazon.com/architecture/well-architected/

Cost Summary

Key Takeaways

Monthly costs shown are for continuous operation (30 days)
Completion costs are 5-10% of monthly costs when you destroy resources immediately
Biggest savings: Destroy resources immediately after completing each lab
All 8 labs completed: ~$20-40 total (vs $200-300/month if kept running)
Best practice: Complete labs in focused 2-4 hour sessions, then destroy

Cost Comparison

Scenario	Lab 01 Cost	All 8 Labs Cost
Run for 1 month	$50-80	$200-300
Complete & destroy	$2-5	$20-40
Savings	95-96%	90-95%

Remember: Costs are estimates. Monitor your actual usage and adjust accordingly. The most important cost optimization is destroying resources immediately after completing labs!

Cost Management Guide ​

Understanding Costs ​

Estimated Monthly Costs vs Completion Costs ​

Cost Breakdown Example (Lab 01) ​

Cost Optimization Strategies ​

For Learning/Development (Minimize Costs) ​

1. Destroy Immediately After Completion ​

2. Use Development Environment ​

2. Scale Down When Not in Use ​

3. Use Single NAT Gateway ​

4. Disable Deletion Protection ​

5. Use Spot Instances (Advanced) ​

6. Reduce RDS Costs ​

7. Clean Up Regularly ​

For Production Reference ​

1. Right-Sizing ​

2. Reserved Instances ​

3. Auto Scaling ​

4. Cost Allocation Tags ​

5. S3 Lifecycle Policies ​

6. CloudWatch Log Retention ​

Cost Monitoring ​

AWS Cost Explorer ​

Cost Estimation Tools ​

Using infracost ​

Using Cost Estimation Script ​

AWS Budgets ​

AWS Free Tier ​

Eligible Services (First 12 Months) ​

Free Tier Optimization ​

Cost Optimization Checklist ​

Before Deployment ​

During Deployment ​

After Deployment ​

Emergency Cost Control ​

Immediate Actions ​

Cost Anomaly Detection ​

Cost Optimization Examples ​

Example 1: Development Environment ​

Example 2: Production Environment ​

Example 3: Cost-Optimized Learning (Recommended) ​

Best Practices ​

Cost Resources ​

Cost Summary ​

Key Takeaways ​

Cost Comparison ​

Cost Management Guide

Understanding Costs

Estimated Monthly Costs vs Completion Costs

Cost Breakdown Example (Lab 01)

Cost Optimization Strategies

For Learning/Development (Minimize Costs)

1. Destroy Immediately After Completion

2. Use Development Environment

2. Scale Down When Not in Use

3. Use Single NAT Gateway

4. Disable Deletion Protection

5. Use Spot Instances (Advanced)

6. Reduce RDS Costs

7. Clean Up Regularly

For Production Reference

1. Right-Sizing

2. Reserved Instances

3. Auto Scaling

4. Cost Allocation Tags

5. S3 Lifecycle Policies

6. CloudWatch Log Retention

Cost Monitoring

AWS Cost Explorer

Cost Estimation Tools

Using infracost

Using Cost Estimation Script

AWS Budgets

AWS Free Tier

Eligible Services (First 12 Months)

Free Tier Optimization

Cost Optimization Checklist

Before Deployment

During Deployment

After Deployment

Emergency Cost Control

Immediate Actions

Cost Anomaly Detection

Cost Optimization Examples

Example 1: Development Environment

Example 2: Production Environment

Example 3: Cost-Optimized Learning (Recommended)

Best Practices

Cost Resources

Cost Summary

Key Takeaways

Cost Comparison