Region Routing

Introduction

With region routing, you can define the deployment configuration of a data connector for different regions.

For data connectors that connect to a data source, e.g. PostgreSQL, it is recommended to deploy the connector in a region closest to the data source to ensure efficient communication between the connector and the data source.

For other data connectors, e.g. Typescript, it is recommended to deploy the connector in a region closest to the consumers of the API to ensure efficient communication between the connector and the Hasura engine.

If you have a distributed data source, with multi-region routing, you can ensure that data is fetched from the data source closest to the user, thus minimizing latency for the request, improving the performance of your application, and providing a better user experience.

See the list of supported regions below.

Single-Region Routing

You can modify the Connector object as per the highlighted values in the example below to force the deployment of your data connector to a specific region. If the region is not specified, the connector will be deployed randomly to one of the supported regions.

For example, in my_subgraph/connector/my_connector/connector.yaml:

kind: Connector
version: v2
definition:
  name: my_connector
  subgraph: my_subgraph
  source: hasura/connector_name:<version>
  context: .
  regionConfiguration:
    - region: <region from the list below>
      mode: ReadWrite
      envMapping:
        <CONNECTOR_ENV_VAR>: # e.g. CONNECTION_URI
          fromEnv: <CONNECTOR_ENV_VAR> # e.g. Env Var set as DB read write URL

Multi-Region Routing

You can modify the Connector object as per the highlighted values in the example below to define the deployment configuration of your connector across multiple regions.

Currently only supported for PostgreSQL

Multi-region routing is currently supported only for the PostgreSQL connector.

Support for other data connectors will be added soon.

For example, in my_subgraph/connector/my_connector/connector.yaml:

kind: Connector
version: v2
definition:
  name: my_connector
  subgraph: my_subgraph
  source: hasura/connector_name:<version>
  context: .
  regionConfiguration:
    - region: <region1: region from the list below>
      mode: ReadWrite
      envMapping:
          <CONNECTOR_ENV_VAR>: # e.g. CONNECTION_URI
              fromEnv: <CONNECTOR_ENV_VAR_REGION_1> # e.g. Env Var set as DB read write replica URL in region1
    - region: <region2: region from the list below>
      mode: ReadOnly
      envMapping:
          <CONNECTOR_ENV_VAR>: # e.g. CONNECTION_URI
              fromEnv: <CONNECTOR_ENV_VAR_REGION_2> # e.g. Env Var set as DB read only replica URL in region2
    - region: <region3: region from the list below>
      mode: ReadOnly
      envMapping:
          <CONNECTOR_ENV_VAR>: # e.g. CONNECTION_URI
              fromEnv: <CONNECTOR_ENV_VAR_REGION_3> # e.g. Env Var set as DB read only replica URL in region3

Supported regions

Currently, Hasura DDN supports the following regions in GCP for multi-region routing:

• gcp-asia-south1
• gcp-asia-southeast1
• gcp-australia-southeast1
• gcp-europe-west1
• gcp-southamerica-east1
• gcp-us-east4
• gcp-us-west2