Data Catalog regions

Data Catalog resources are hosted in multiple locations worldwide, called regions. A region is a specific geographical location where you can access your resources. For more information, see Geography and Regions.

To decrease network latency, we recommend using the API endpoint for a region that is close to where you plan to use it. For example, if you plan to access Data Catalog from Asia, then you could use the endpoint for Taiwan.

Supported regions

Region Regional endpoint Supported resources
Montréal BigQuery
São Paulo BigQuery
Iowa TagTemplates, user created entries
South Carolina BigQuery
North Virginia BigQuery
Oregon TagTemplates, user created entries
Los Angeles BigQuery
Finland BigQuery
Belgium TagTemplates, user created entries
London BigQuery
Frankfurt BigQuery
Netherlands TagTemplates, user created entries
Zurich BigQuery
Asia Pacific
Taiwan BigQuery, TagTemplates, user created entries
Hong Kong BigQuery
Tokyo BigQuery
Osaka BigQuery
Seoul BigQuery
Mumbai BigQuery
Singapore BigQuery
Sydney BigQuery
Multi-region Multi-regional endpoint Supported resources
Global Pub/Sub, TagTemplates
European Union BigQuery, TagTemplates, user created entries
United States BigQuery, TagTemplates, user created entries

Multi-regions (us, eu, global)

global is a special multi-region, primarily created to serve resources that do not have a single region associated with them--for example metadata synced for Cloud Pub/Sub topics. Internally, Data Catalog can replicate and serve global resources anywhere around the world.

Similarly, multi-regions such as eu and us are special regions, where Data Catalog can replicate and serve resources anywhere within them.

Accessing resources with regional APIs

Resources can be accessed using regional endpoints (listed above). Client libraries allow you to configure endpoints.

For backward compatibility, regional APIs are still supported by the endpoint, with some limitations (see HTTP batch and gRPC). It is strongly recommended that you use regional endpoints for regional APIs.

Non-regional APIs

While data assets are located in particular regions, Data Catalog's search features are globally scoped. Non-regional APIs like SearchCatalog and LookupEntry return assets from all regions, and can be accessed with the endpoint.

HTTP batch support

Data Catalog supports HTTP batch, letting you batch multiple API calls into a single request.

There are some constraints when using HTTP batch for regional APIs:

  • A HTTP batch request must only include API requests for resources belonging to the same region.
  • You must use regional endpoints when batching regional APIs. will not support batched regional APIs. Regional and non-regional requests cannot be batched together.
  • Non-region dependent APIs (SearchCatalog and LookupEntry) can be batched together, using the endpoint.

If you're not sure whether whether you're using HTTP batch:

  • Check whether you are directly making a HTTP request with a path set to /batch, such as
  • If you are using the Google API client library:

    • Java: Using BatchRequest to call APIs.
    • Python: Using BatchHTTP request to call APIs.
    • Similar bindings exist for other libraries (please see the library specific documentation).

gRPC support

Regional endpoints are the recommended way to use gRPC for regional APIs.

Note that some existing Google Cloud client libraries use gRPC internally. For backwards compatibility, the endpoint supports regional APIs when using gRPC from these libraries.