Skip to main content

gcp_bigquery_select

BETA

This component is mostly stable but breaking changes could still be made outside of major version releases if a fundamental problem with the component is found.

Executes a SELECT query against BigQuery and creates a message for each row received.

Introduced in version 3.63.0.

# Config fields, showing default values
input:
  label: ""
  gcp_bigquery_select:
    project: "" # No default (required)
    table: bigquery-public-data.samples.shakespeare # No default (required)
    columns: [] # No default (required)
    where: type = ? and created_at > ? # No default (optional)
    auto_replay_nacks: true
    job_labels: {}
    priority: ""
    args_mapping: root = [ "article", now().ts_format("2006-01-02") ] # No default (optional)
    prefix: "" # No default (optional)
    suffix: "" # No default (optional)

Once the rows from the query are exhausted, this input shuts down, allowing the pipeline to gracefully terminate (or the next input in a sequence to execute).

Examples​

Here we query the public corpus of Shakespeare's works to generate a stream of the top 10 words that are 3 or more characters long:

input:
  gcp_bigquery_select:
    project: sample-project
    table: bigquery-public-data.samples.shakespeare
    columns:
      - word
      - sum(word_count) as total_count
    where: length(word) >= ?
    suffix: |
      GROUP BY word
      ORDER BY total_count DESC
      LIMIT 10
    args_mapping: |
      root = [ 3 ]

Fields​

project​

GCP project where the query job will execute.

Type: string

table​

Fully-qualified BigQuery table name to query.

Type: string

# Examples

table: bigquery-public-data.samples.shakespeare

columns​

A list of columns to query.

Type: array

where​

An optional where clause to add. Placeholder arguments are populated with the args_mapping field. Placeholders should always be question marks (?).

Type: string

# Examples

where: type = ? and created_at > ?

where: user_id = ?

auto_replay_nacks​

Whether messages that are rejected (nacked) at the output level should be automatically replayed indefinitely, eventually resulting in back pressure if the cause of the rejections is persistent. If set to false these messages will instead be deleted. Disabling auto replays can greatly improve memory efficiency of high throughput streams as the original shape of the data can be discarded immediately upon consumption and mutation.

Type: bool
Default: true

job_labels​

A list of labels to add to the query job.

Type: object
Default: {}

priority​

The priority with which to schedule the query.

Type: string
Default: ""

args_mapping​

An optional Bloblang mapping which should evaluate to an array of values matching in size to the number of placeholder arguments in the field where.

Type: string

# Examples

args_mapping: root = [ "article", now().ts_format("2006-01-02") ]

prefix​

An optional prefix to prepend to the select query (before SELECT).

Type: string

suffix​

An optional suffix to append to the select query.

Type: string