Chapter 118. Google BigQuery Component

The Google Bigquery component provides access to Cloud BigQuery Infrastructure via the Google Client Services API.

The current implementation does not use gRPC.

The current implementation does not support querying BigQuery i.e. is a producer only.

Maven users will need to add the following dependency to their pom.xml for this component:

<dependency>
    <groupId>org.apache.camel</groupId>
    <artifactId>camel-google-bigquery</artifactId>
    <version>x.x.x</version>
    <!-- use the same version as your Camel core version -->
</dependency>

Google BigQuery component authentication is targeted for use with the GCP Service Accounts. For more information please refer to Google Cloud Platform Auth Guide

Google security credentials can be set explicitly via one of the two options:

If both are set, the Service Account Email/Key will take precedence.

Or implicitly, where the connection factory falls back on Application Default Credentials.

OBS! The location of the default credentials file is configurable - via GOOGLE_APPLICATION_CREDENTIALS environment variable.

Service Account Email and Service Account Key can be found in the GCP JSON credentials file as client_email and private_key respectively.

        google-bigquery://project-id:datasetId[:tableId]?[options]

The Google BigQuery component supports 4 options, which are listed below.

The Google BigQuery endpoint is configured using URI syntax:

google-bigquery:projectId:datasetId:tableName

with the following path and query parameters:

The component supports 4 options, which are listed below.

Producer endpoints can accept and deliver to BigQuery individual and grouped exchanges alike. Grouped exchanges have Exchange.GROUPED_EXCHANGE property set.

Goole BigQuery producer will send a grouped exchange in a single api call unless different table suffix or partition decorators are specified in which case it will break it down to ensure data is written with the correct suffix or partition decorator.

Google BigQuery endpoint expects the payload to be either a map or list of maps. A payload containing a map will insert a single row and a payload containing a list of map’s will insert a row for each entry in the list.

Reference: https://cloud.google.com/bigquery/streaming-data-into-bigquery#template-tables

Templated tables can be specified using the GoogleBigQueryConstants.TABLE_SUFFIX header.

I.e. the following route will create tables and insert records sharded on a per day basis:

from("direct:start")
.header(GoogleBigQueryConstants.TABLE_SUFFIX, "_${date:now:yyyyMMdd}")
.to("google-bigquery:sampleDataset:sampleTable")

Note it is recommended to use partitioning for this use case.

Reference: https://cloud.google.com/bigquery/docs/creating-partitioned-tables

Partitioning is specified when creating a table and if set data will be automatically partitioned into separate tables. When inserting data a specific partition can be specified by setting the GoogleBigQueryConstants.PARTITION_DECORATOR header on the exchange.

Reference: https://cloud.google.com/bigquery/streaming-data-into-bigquery#dataconsistency

A insert id can be set on the exchange with the header GoogleBigQueryConstants.INSERT_ID or by specifying query parameter useAsInsertId. As an insert id need to be specified per row inserted the exchange header can’t be used when the payload is a list - if the payload is a list the GoogleBigQueryConstants.INSERT_ID will be ignored. In that case use the query parameter useAsInsertId.