Examples and Use Cases

July 28, 2025 · View on GitHub

This document provides real-world examples of using spotinfo for common DevOps scenarios.

DevOps Use Cases

1. Production Workload Deployment

Scenario: Deploy a production web application requiring high availability and cost optimization.

Requirements:

At least 4 vCPU and 16GB RAM
High placement score (8+) for reliability
Compare multiple regions
Budget constraint: $0.30/hour

# Find optimal instances across multiple regions
spotinfo \
  --cpu 4 \
  --memory 16 \
  --with-score \
  --min-score 8 \
  --price 0.30 \
  --region "us-east-1" \
  --region "us-west-2" \
  --region "eu-west-1" \
  --sort score \
  --order desc \
  --output table

Expected Output:

┌───────────────┬──────────┬──────┬────────────┬──────────┬────────────────────────────┐
│ REGION        │ INSTANCE │ VCPU │ MEMORY GIB │ USD/HOUR │ PLACEMENT SCORE (REGIONAL) │
├───────────────┼──────────┼──────┼────────────┼──────────┼────────────────────────────┤
│ us-east-1     │ m5.xlarge│    4 │         16 │   0.0817 │ 🟢 9                       │
│ eu-west-1     │ m5.xlarge│    4 │         16 │   0.0923 │ 🟢 8                       │
└───────────────┴──────────┴──────┴────────────┴──────────┴────────────────────────────┘

2. Development Environment Setup

Scenario: Cost-effective development instances for a team of developers.

Requirements:

Small instances (t3 family)
Acceptable reliability (score 5+)
Multiple options for flexibility
Lowest cost priority

# Find cheapest t3 instances with decent reliability
spotinfo \
  --type "t3.*" \
  --with-score \
  --min-score 5 \
  --sort price \
  --order asc \
  --region "us-east-1" \
  --output table

3. Machine Learning Training Jobs

Scenario: GPU instances for ML training workloads that can handle interruptions.

Requirements:

GPU instances (p3, g4 families)
Cost optimization priority
AZ-level placement for precise targeting

# Compare GPU instances with AZ-level scores
spotinfo \
  --type "(p3|g4).*" \
  --with-score \
  --az \
  --region "us-east-1" \
  --sort price \
  --order asc \
  --output table

4. Batch Processing Workloads

Scenario: Large-scale data processing requiring high memory.

Requirements:

Memory-optimized instances (r5 family)
At least 64GB RAM
Regional analysis for capacity planning

# Find high-memory instances across regions
spotinfo \
  --type "r5.*" \
  --memory 64 \
  --with-score \
  --region "all" \
  --sort score \
  --order desc \
  --output json > high_memory_options.json

Automation Examples

5. Infrastructure as Code Integration

Terraform Variable Generation:

#!/bin/bash
# Generate Terraform variables for spot instances

INSTANCE_DATA=$(spotinfo \
  --type "m5\.(large|xlarge)" \
  --with-score \
  --min-score 7 \
  --region "us-east-1" \
  --output json)

# Extract best instance type
BEST_INSTANCE=$(echo "$INSTANCE_DATA" | jq -r '.[0].instance')

# Generate Terraform variables
cat > terraform.tfvars <<EOF
spot_instance_type = "$BEST_INSTANCE"
availability_zone = "us-east-1a"
max_price = "$(echo "$INSTANCE_DATA" | jq -r '.[0].price * 1.1')"
EOF

echo "Generated terraform.tfvars with optimal spot instance configuration"

6. CI/CD Pipeline Integration

GitLab CI Example:

# .gitlab-ci.yml
variables:
  MAX_SPOT_PRICE: "0.50"

validate_spot_cost:
  stage: validate
  image: ghcr.io/alexei-led/spotinfo:latest
  script:
    - |
      CURRENT_PRICE=$(spotinfo --type "$INSTANCE_TYPE" --region "$AWS_REGION" --output number)
      if (( $(echo "$CURRENT_PRICE > $MAX_SPOT_PRICE" | bc -l) )); then
        echo "Spot price $CURRENT_PRICE exceeds budget $MAX_SPOT_PRICE"
        exit 1
      fi
      echo "Spot price validation passed: $CURRENT_PRICE <= $MAX_SPOT_PRICE"

deploy_infrastructure:
  stage: deploy
  dependencies:
    - validate_spot_cost
  script:
    - terraform apply -auto-approve

7. Cost Monitoring Script

Daily Cost Analysis:

#!/bin/bash
# daily_spot_analysis.sh - Monitor spot instance costs

DATE=$(date +%Y-%m-%d)
REPORT_FILE="spot_report_$DATE.json"

# Generate comprehensive spot analysis
spotinfo \
  --type "m5.*" \
  --with-score \
  --region "us-east-1" \
  --region "us-west-2" \
  --region "eu-west-1" \
  --sort price \
  --order asc \
  --output json > "$REPORT_FILE"

# Extract insights
CHEAPEST=$(jq -r '.[0] | "\(.instance) in \(.region): $\(.price)/hour"' "$REPORT_FILE")
HIGHEST_SCORE=$(jq -r 'sort_by(.region_score) | reverse | .[0] | "\(.instance) in \(.region): score \(.region_score)"' "$REPORT_FILE")

# Send to monitoring system
cat <<EOF | curl -X POST -H 'Content-Type: application/json' -d @- "$WEBHOOK_URL"
{
  "text": "Daily Spot Analysis",
  "blocks": [
    {
      "type": "section",
      "text": {
        "type": "mrkdwn",
        "text": "*Cheapest*: $CHEAPEST\n*Highest Score*: $HIGHEST_SCORE"
      }
    }
  ]
}
EOF

Advanced Queries

8. Multi-Criteria Optimization

Find the sweet spot between cost and reliability:

# Score instances by combined cost-reliability metric
spotinfo \
  --type "c5.*" \
  --with-score \
  --region "us-east-1" \
  --output json | \
jq 'map(. + {
  "cost_reliability_ratio": (.price / (.region_score // 1))
}) | sort_by(.cost_reliability_ratio) | .[0:5]'

9. Capacity Planning Analysis

Compare placement scores for different instance families:

# Analyze placement scores across instance families
for family in "m5" "c5" "r5"; do
  echo "=== $family family ==="
  spotinfo \
    --type "${family}.*" \
    --with-score \
    --region "us-east-1" \
    --output json | \
  jq -r '.[] | "\(.instance): \(.region_score // "N/A")"' | \
  head -5
  echo
done

10. AZ-Level Deployment Strategy

Compare regional vs AZ-level scores for deployment decisions:

#!/bin/bash
INSTANCE_TYPE="m5.large"
REGION="us-east-1"

echo "Regional Score:"
spotinfo --type "$INSTANCE_TYPE" --with-score --region "$REGION" --output text

echo -e "\nAZ-Level Scores:"
spotinfo --type "$INSTANCE_TYPE" --with-score --az --region "$REGION" --output text

echo -e "\nRecommendation:"
REGIONAL_SCORE=$(spotinfo --type "$INSTANCE_TYPE" --with-score --region "$REGION" --output json | jq -r '.[0].region_score // 0')
AZ_SCORE=$(spotinfo --type "$INSTANCE_TYPE" --with-score --az --region "$REGION" --output json | jq -r '.[0].zone_scores | to_entries[0].value // 0')

if [ "$AZ_SCORE" -gt "$REGIONAL_SCORE" ]; then
  echo "Deploy to specific AZ for better placement score"
else
  echo "Regional deployment recommended"
fi

Output Format Examples

11. CSV Export for Spreadsheet Analysis

# Generate CSV report for stakeholders
spotinfo \
  --type "m5.*" \
  --with-score \
  --region "all" \
  --output csv > "spot_instances_$(date +%Y%m%d).csv"

# Import into Google Sheets or Excel for further analysis

12. JSON Processing with jq

# Extract specific fields for monitoring
spotinfo \
  --type "c5.*" \
  --with-score \
  --region "us-east-1" \
  --output json | \
jq -r '.[] | select(.region_score >= 8) | 
  "Instance: \(.instance), Score: \(.region_score), Price: $\(.price)/hour"'

13. Text Format for Logging

# Log format for system monitoring
spotinfo \
  --type "t3.medium" \
  --with-score \
  --region "us-east-1" \
  --output text | \
  logger -t "spotinfo" -p user.info

Integration Patterns

14. Kubernetes Cluster Autoscaler

Node group optimization:

# Find optimal instance types for Kubernetes node groups
spotinfo \
  --cpu 2 \
  --memory 8 \
  --with-score \
  --min-score 7 \
  --region "us-east-1" \
  --sort price \
  --order asc \
  --output json | \
jq -r '.[0:3][] | .instance' | \
tr '\n' ',' | \
sed 's/,$//'

15. AWS Auto Scaling Groups

Mixed instance policy configuration:

# Generate instance types for ASG mixed instance policy
INSTANCE_TYPES=$(spotinfo \
  --type "m5\.(large|xlarge|2xlarge)" \
  --with-score \
  --min-score 6 \
  --region "us-east-1" \
  --output json | \
jq -r '[.[].instance] | unique | join(",")')

echo "InstanceTypes: $INSTANCE_TYPES"

Troubleshooting Examples

16. Permission Debugging

# Test placement score permissions
if spotinfo --type "t3.micro" --with-score --region "us-east-1" &>/dev/null; then
  echo "✅ Placement score permissions working"
else
  echo "❌ Placement score permissions failed"
  echo "Check IAM policy for ec2:GetSpotPlacementScores"
fi

17. Performance Testing

# Measure query performance
time spotinfo --type "m5.*" --region "all" --output json >/dev/null

# Test with placement scores
time spotinfo --type "m5.*" --with-score --region "us-east-1" --output json >/dev/null

These examples demonstrate the flexibility and power of spotinfo for various DevOps scenarios. Adapt them to your specific requirements and integrate them into your infrastructure automation workflows.