Sink data from RisingWave to Google BigQuery

This guide describes how to sink data from RisingWave to Google BigQuery.

BigQuery is Google’s fully managed data warehouse and data analytics platform, capable of handling and analyzing large volumes of data as it is highly scalable.

You can test out this process on your own device by using the big-query-sink demo in the integration_test directory of the RisingWave repository.

PREMIUM FEATURE

This is a premium feature. For a comprehensive overview of all premium features and their usage, please see RisingWave premium features.

Prerequisites

Before sinking data from RisingWave to BigQuery, please ensure the following:

• The BigQuery table you want to sink to is accessible from RisingWave.
• Ensure you have an upstream materialized view or table in RisingWave that you can sink data from.

Syntax

CREATE SINK [ IF NOT EXISTS ] sink_name
[FROM sink_from | AS select_query]
WITH (
   connector='bigquery',
   connector_parameter = 'value', ...
);

Parameters

Parameter Names	Description
sink_name	Name of the sink to be created.
sink_from	A clause that specifies the direct source from which data will be output. sink_from can be a materialized view or a table. Either this clause or select_query query must be specified.
AS select_query	A SELECT query that specifies the data to be output to the sink. Either this query or a sink_from clause must be specified. See SELECT for the syntax and examples of the SELECT command.
type	Required. Data format. Allowed formats: append-only: Output data with insert operations. upsert: For this type, you need to set corresponding permissions and primary keys based on the Document of BigQuery.
force_append_only	Optional. If true, forces the sink to be append-only, even if it cannot be.
bigquery.local.path	Optional. The file path leading to the JSON key file located on your local machine. Details can be found in Service Accounts under your Google Cloud account. At least one of bigquery.credentials, bigquery.local.path, or bigquery.s3.path must be specified.
bigquery.s3.path	Optional. The file path leading to the JSON key file located in S3. Details can be found in Service Accounts under your Google Cloud account. At least one of bigquery.credentials, bigquery.local.path, or bigquery.s3.path must be specified. If this parameter is used, aws.credentials.access_key_id, aws.credentials.secret_access_key, and region are also required.
bigquery.project	Required. The BigQuery project ID.
bigquery.dataset	Required. The BigQuery dataset ID.
bigquery.table	Required. The BigQuery table you want to sink to.
bigquery.credentials	Optional. Enter the JSON key credentials as a string. Both plain JSON and Base64-encoded JSON are accepted; Base64 is recommended for inline use to avoid escaping issues. At least one of bigquery.credentials, bigquery.local.path, or bigquery.s3.path must be specified.
bigquery.retry_times	Optional. The number of times the system should retry a BigQuery insert operation before ultimately returning an error. Defaults to 5.
auto_create	Optional. Defaults to false. If true, a new table will be automatically created in BigQuery when the specified table is not found.
aws.credentials.access_key_id	Optional. The access key for accessing the S3 file storing the JSON key. Required when bigquery.s3.path is specified.
aws.credentials.secret_access_key	Optional. The secret access key for accessing the S3 file storing the JSON key. Required when bigquery.s3.path is specified.
region	Optional. The AWS region of the S3 file storing the JSON key. Required when bigquery.s3.path is specified.

Examples

RisingWave supports three ways to provide the Google Cloud service account JSON key for authenticating with BigQuery.

Option 1: Inline credentials string

Example using inline credentials

CREATE SINK big_query_sink_credentials
FROM mv1
WITH (
    connector = 'bigquery',
    type = 'append-only',
    bigquery.credentials= '${base64_encoded_service_account_json}',
    bigquery.project= '${project_id}',
    bigquery.dataset= '${dataset_id}',
    bigquery.table= '${table_id}',
    force_append_only='true'
);

Option 2: Local JSON key file

Example using a local JSON key file

CREATE SINK big_query_sink_local
FROM mv1
WITH (
    connector = 'bigquery',
    type = 'append-only',
    bigquery.local.path= '${bigquery_service_account_json_path}',
    bigquery.project= '${project_id}',
    bigquery.dataset= '${dataset_id}',
    bigquery.table= '${table_id}',
    force_append_only='true'
);

Option 3: JSON key file stored in S3

When the JSON key file is stored in S3, you must also provide the AWS credentials and region to access it.

Example using a JSON key file stored in S3

CREATE SINK big_query_sink_s3
FROM mv1
WITH (
    connector = 'bigquery',
    type = 'append-only',
    bigquery.s3.path= '${s3_service_account_json_path}',
    bigquery.project= '${project_id}',
    bigquery.dataset= '${dataset_id}',
    bigquery.table= '${table_id}',
    aws.credentials.access_key_id = '${aws_access_key}',
    aws.credentials.secret_access_key = '${aws_secret_access}',
    region = '${aws_region}',
    force_append_only='true'
);

Data type mapping

RisingWave Data Type	BigQuery Data Type
boolean	bool
smallint	int64
integer	int64
bigint	int64
real	unsupported
double precision	float64
numeric	numeric
date	date
character varying (varchar)	string
time without time zone	time
timestamp without time zone	datetime
timestamp with time zone	timestamp
interval	interval
struct	struct
array	array
bytea	bytes
JSONB	JSON
serial	int64