Cloud DNS Routing Policies

 🔸 Routing is the process of selecting best path (route) to reach destination network from your source network.

🔸 𝗦𝗲𝗿𝘃𝗲𝗿 𝗽𝗼𝗹𝗶𝗰𝗶𝗲𝘀 apply private DNS configuration to a Virtual Private Cloud (VPC) network (DNS forwarding, logging).

🔸 𝗥𝗲𝘀𝗽𝗼𝗻𝘀𝗲 𝗽𝗼𝗹𝗶𝗰𝗶𝗲𝘀 override private DNS responses based on the query name.

🔸 𝗥𝗼𝘂𝘁𝗶𝗻𝗴 𝗽𝗼𝗹𝗶𝗰𝗶𝗲𝘀 steer traffic based on query (for example, round robin, geolocation).

Cloud DNS routing policies

Each policy enables distinct traffic routing methods

𝟭. 𝗦𝗶𝗺𝗽𝗹𝗲:
Route 100% of the traffic to a single resource.

𝟮. 𝗪𝗲𝗶𝗴𝗵𝘁𝗲𝗱:
Specify the % of requests for a specific resource.

𝘍𝘶𝘯 𝘧𝘢𝘤𝘵: If all resources have weight=0, traffic is evenly distributed.

𝟯. 𝗙𝗮𝗶𝗹𝗼𝘃𝗲𝗿:
▪ Used for DR purposes, active-passive setup with resource health checks.
▪ On primary health check failure, requests auto-routed to secondary.

𝟰. 𝗟𝗮𝘁𝗲𝗻𝗰𝘆 𝗕𝗮𝘀𝗲𝗱:
▪ Redirect users to low-latency resource regions.

𝘍𝘶𝘯 𝘧𝘢𝘤𝘵: Not always closest region to the user.

𝟱. 𝗚𝗲𝗼𝗹𝗼𝗰𝗮𝘁𝗶𝗼𝗻:
▪ Route traffic by user location.
▪ Matched users routed to pre-configured locations.
▪ No-match users redirected to the default location.

𝘜𝘴𝘦 𝘤𝘢𝘴𝘦: OTT platforms use this to restrict programs based on the user location.

𝟲. 𝗚𝗲𝗼𝗽𝗿𝗼𝘅𝗶𝗺𝗶𝘁𝘆:
▪ Shifts traffic based on a bias.
▪ To allow more traffic (bias range: 1 to 99)
▪ To shrink traffic (bias range: -1 to -99)

𝟳. 𝗠𝘂𝗹𝘁𝗶-𝗩𝗮𝗹𝘂𝗲 𝗔𝗻𝘀𝘄𝗲𝗿:
▪ When you want DNS to respond to DNS queries with up to 8 healthy records selected at random.

𝗡𝗼𝘁𝗲: Multi-value isn't a load balancer replacement