bigquery-geotools

Install

Build or download the JARs and copy them into geoserver's WEB-INF/lib folder in the same way you install other geoserver plugins.

A .zip containing the bigquery-geotools JAR and required dependency JARs is built against JDK 8 is available in under releases.

gcloud builds submit .

mvn package -DskipTests

The service account used to authenticate with BigQuery needs the following permissions:

You can use one of three authentication methods to connect to BigQuery, depending on your requirements:

If your geoserver instance is running inside GCP, on a VM or elsewhere, you can make use of Default service account credentials. (Recommended)
Set the GOOGLE_APPLICATION_CREDENTIALS environment variable
Export a service account key JSON file, add it to disk, and manually select the file when configuring the BigQuery data source. Note: Do not store the service account key file in the geoserver data directory, it may not be secure there.

Create a new data soruce he same way as other databases. Under Stores, select Add New Store, then select BigQuery Table.

Now, you will configure the connection to a BigQuery dataset. Datasets can contain many tables, and are roughly analogous to PostGIS schemas.

The Pregenerate Materialized Views setting will do one of the following:

MV_NONE: do not generate or utilize any pregenerated materialized views.
MV_USE_EXISTING: utilize any existing materialized views, but do not generate any.
MV_PREGEN_ALL: pregenerate 4 tables of simplified geometries at tolerances 1m, 10m, 100m, and 1000m.

To generate materialized views manually, see this sample SQL.

Select Publish to configure and publish a vectory layer from BigQuery.

There are several configuration options that can be set when adding a BigQuery data source.

Parameter	Default	Description
Access Method	`QUERY_API`	Select whether to query using the BigQuery Storage API or standard Query API.
Simplify Geometries	`true`	Attempt to simplify geometries at wider zoom levels without impacting accuracy
Use Query Cache	`true`	Use the BigQuery query cache when possible
Query Recent Partition	`true`	When querying a partitioned table, automatically detect the most recent partition and use it for queries