Routing Policies
Route53 Health Checks
- Health Checks are created and configured seperately
- Records an use it for routing
- Health Checkers are located globally
- Interval - every 30s (reduce to 10s for extra cost)
- Supports multiple method of health checks
- Endpoint: Ip, TCP (2s), HTTP/HTTPS (2s + 2s body) string matching
- Cloudwatch alarms
- Other health checks - checks of checks
Simple Routing
- Supports 1 record per NAME
- 1 record can have multiple values
- Single value is returned randomly
- DOES NOT SUPPORT HEALTH CHECKS
Failover Routing
- Active/passive failover
- health check on primary record and failover to secondary record if health check is failing on primary record
Multi-Value Routing
- Multiple records with same name
- each record is independent and can have an associated health check
- NOT THE SUBSTITUTION FOR LOAD BALANCER
- Upto 8 healthy records can be retured to client (> 8 records - randomized 8)
- unhealthy records are not returned
Weighted Routing
- Simple load balancing with weights
- helpful for testing new software versions
- control the distribution for same record
Latency-Based Routing
- Optimze performance and user experience
- AWS maintain database of latency to redirect traffic to record with least latency
- health check can be associated with routes to ensure healhty routes are returned
Geolocation Routing
R53 Geolocation Docs - Relevant (location) records - State - SubdivisionCode | SubdivisionName (only for US) - Country - CountryCode | CountryName - Continent - ContinentCode | ContinentName - Default
- Cane be used of regional restrictions, language specific content or load balancing across regional endpoints
Geoproximity Routing
- Records can be tagged with an AWS Region or lat and long cordinates
- Routing is distance based but can be influence with bias
+or-bias can be added to rules to increase a region size.
Route 53 Interoperability
- Domain registrar or Domain Hosting
