Bigquery dataset using pub/sub subscription

Question

Trying to create a dataset using pub/sub subscription message. https://cloud.google.com/pubsub/docs/bigquery. This one provides how to write message to the table, but how can I create a dataset using the message. For example – if I have this is test1 in my subscription, then I should be able to take the message(test1) and create a dataset based on that message. Just an empty dataset, no table needed. I have done all the research and didn’t see any solution from google.

Asked By: user3713336

||

Source

Answer 1

You can trigger a Cloud Function when the message is published on the topic with Pub Sub triggers :

https://cloud.google.com/functions/docs/tutorials/pubsub

gcloud functions deploy python-pubsub-function 
--gen2 
--runtime=python310 
--region=REGION 
--source=. 
--entry-point=subscribe 
--trigger-topic=YOUR_TOPIC_NAME

You can also check this link :

https://cloud.google.com/functions/docs/calling/pubsub

The Cloud Function can use Python client to create your dataset, here an example :

from google.cloud import bigquery

import base64

import functions_framework

@functions_framework.cloud_event
def subscribe(cloud_event):
    dataset_id_from_topic =  base64.b64decode(cloud_event.data["message"]["data"]).decode()

    # Construct a BigQuery client object.
    client = bigquery.Client()

    # Construct a full Dataset object to     send to the API.
    dataset = bigquery.Dataset(dataset_id_from_topic)

    # TODO(developer): Specify the geographic location where the dataset should reside.
    dataset.location = "US"

    # Send the dataset to the API for creation, with an explicit timeout.
    # Raises google.api_core.exceptions.Conflict if the Dataset already
    # exists within the project.
    dataset = client.create_dataset(dataset, timeout=30)  

    # Make an API request.
    print("Created dataset {}.   {}".format(client.project, dataset.dataset_id))

Answered By: Mazlum Tosun

Bigquery dataset using pub/sub subscription

Question:

Answers: