Route 53 Routing Policies

How It Works:

• Assign a weight (0-255) to each record
• Traffic percentage = (record weight) / (sum of all weights)
• Weight of 0 = no traffic to that record

Example Distribution:

Record A: Weight 70 → 70% traffic
Record B: Weight 20 → 20% traffic
Record C: Weight 10 → 10% traffic

Health Checks: Supported - unhealthy records excluded from responses

How It Works:

1. Create latency records for resources in multiple AWS regions
2. Route 53 measures latency between users and AWS regions
3. Returns record for region with lowest latency to user

Key Points:

• Latency is measured from user to AWS region, not to your resource
• Latency data is collected over time by AWS
• Not necessarily the geographically closest region
• Requires resources in multiple AWS regions

Record Types:

• Primary: Active resource (must have health check)
• Secondary: Standby resource (health check optional)

Behavior:

1. Route 53 monitors primary via health check
2. When primary is healthy → returns primary record
3. When primary fails → returns secondary record
4. When primary recovers → returns to primary (failback)

Secondary Options:

• Another active resource in different region
• Static S3 website (maintenance page)
• CloudFront distribution

Location Hierarchy:

1. Continent: Africa, Antarctica, Asia, Europe, North America, Oceania, South America
2. Country: Specific country codes (US, UK, JP, etc.)
3. State/Province: US states, Canadian provinces (subdivisions)

Matching Order:

1. Most specific match (state/province)
2. Country match
3. Continent match
4. Default record (if configured)

Important: Always configure a default record for unmatched locations!

How It Works:

• Routes to geographically closest resource
• Bias expands or shrinks the geographic region

Bias Values:

• Positive bias (1 to 99): Expands region, attracts more traffic
• Negative bias (-1 to -99): Shrinks region, repels traffic
• Bias of 0: No adjustment (default)

Example:

• Resource A at 150km with +50 bias → treated as 75km
• Resource B at 100km with no bias → remains 100km
• Result: Traffic routes to A (appears closer)

Requirement: Must use Route 53 Traffic Flow

How It Works:

• Returns up to 8 healthy records per query
• Each record must have a health check
• Unhealthy records excluded from responses
• Client typically uses first IP returned

Key Characteristics:

• Simple client-side load balancing
• NOT a replacement for load balancer
• Health check required for each record
• Random selection of healthy records

How It Works:

• Define CIDR blocks and map them to specific endpoints
• Route 53 matches client IP against CIDR collections
• Returns record for matching CIDR

Use Cases:

• Route ISP customers to specific endpoints
• Enterprise networks to dedicated resources
• Optimize for known IP ranges (CDN, partners)

Configuration:

1. Create CIDR collection with IP blocks
2. Create CIDR location for each block
3. Associate records with CIDR locations

Active-Passive Failover:

• Uses Failover routing policy
• Primary resource handles all traffic when healthy
• Secondary (standby) only receives traffic when primary fails
• Best for: DR scenarios, maintenance pages

Active-Active Failover:

• Uses any other routing policy (weighted, latency, geolocation, etc.)
• All resources actively serve traffic
• Unhealthy resources automatically excluded
• Best for: High availability, load distribution

Example:

• Active-Active with Weighted: Both regions get traffic, unhealthy region excluded
• Active-Passive with Failover: Only primary gets traffic until it fails

Scenario: Global application with regional failover

Architecture:

                    ┌─────────────────────┐
                    │  Latency Records    │
                    │  (top level)        │
                    └──────────┬──────────┘
                               │
           ┌───────────────────┼───────────────────┐
           │                   │                   │
    ┌──────▼──────┐     ┌──────▼──────┐     ┌──────▼──────┐
    │  Failover   │     │  Failover   │     │  Failover   │
    │  us-east-1  │     │  eu-west-1  │     │  ap-south-1 │
    └──────┬──────┘     └──────┬──────┘     └──────┬──────┘
           │                   │                   │
     ┌─────┴─────┐       ┌─────┴─────┐       ┌─────┴─────┐
     │           │       │           │       │           │
   Primary   Secondary Primary   Secondary Primary   Secondary

Behavior:

1. Latency routing selects closest region
2. Within region, failover handles availability
3. If regional primary fails, traffic goes to regional secondary