About geographic filters

You don’t need to assign individual answers for every region. If a requester is in a region without a regional answer, they are given the closest existing regional answer. For example, assume that you have two answers: One that assigns the US-EAST metadata value and one that assigns the ASIAPAC metadata value. If a requester is in US-CENTRAL, they receive the US-EAST answer.

You can use this filter with another one like Select First N to send the user to the closest answer. If you have multiple answers in the same location, consider using Shuffle to randomly pick one of the closest answers.

• Source IP of the request if EDNS Client Subnet (ECS) is supported and activated by the recursive resolver, or
• Source IP of the recursive resolver if ECS is not enabled.

Individual answers for every country, subdivision, state, and province aren’t required. If a requester is in a location for which there is no answer with matching metadata, they are given the closest existing answer based on geographical distance. For example, assume you have two answers: One with JP, HK (Japan and Hong Kong) and another with US, CA (United States and Canada). A user in Taiwan is directed to the JP, HK endpoint since Taiwan is closer to Hong Kong.

You can use this filter with another one like Select First N to send the user to the closest answer. If you have multiple answers in the same location, consider using Shuffle to randomly pick one of the closest answers.

This filter examines the latitude and longitude metadata fields to get the location of your answers. Then, NS1 uses the geoIP database to determine the location of the requester to determine which answer is closest.

You can use this filter with another one like Select First N to send the user to the closest answer. If you have multiple answers in the same location, consider using Shuffle to randomly pick one of the closest answers.