Bulk delete Cloud Storage objects using BigQuery remote functions¶

A common way to remove objects from Cloud Storage is to use object lifecycle management, which allows Time to live (TTL) driven automatic changes, like moving objects storage class and delete objects.

When your Cloud Storage bucket does not have life-cycle management enabled, you can use the gsutil command.

A limitation of this approach is the complexity that arises when you need to delete a very large number of files:

You might need to spread the operation across multiple machines
You might need to retry on failures

In cases when there are hundreds of thousands or more objects to delete, if you have the list of URIs (or some logic) to remove in a database, BigQuery remote functions provide an easy mechanism to scale out the operation through micro-batching, with resilience and retries.

This article shows you how to deploy a BigQuery remote function that bulk-deletes objects and gives the option to stores the results in a table.

Google Cloud services¶

Brief introduction to a few of the Google Cloud services that are referenced here:

Cloud Storage is a managed service for Object storage that can be used for storing unstructured data.
BigQuery is a fully managed, data analytics platform that enables multi-engine, multi-format SQL driven data processing. BigQuery supports remote functions that allow writing a user-defined processing logic as Cloud Run or Cloud Function services.
Cloud Run

Architecture¶

A Cloud Run service runs a Spring boot application that receives a list of object URIs to be deleted as an HTTP request body. The application splits the list into supported batch sized sub-lists and makes batch calls to Cloud Storage API to delete the objects. The status of the deletion request for each URI from the Storage API is then relayed back as the HTTP response.

BigQuery invokes The Cloud Run application as a remote function with a list of URIs by reading an existing table or through other SQL-based logic. The responses are shown on screen or can be stored to a result table if required. BigQuery handles the job of splitting the original list into smaller batches before calling the remote functions and stitches the responses back.

architecture

Objectives¶

Deploy the Bulk delete remote function to Cloud Run
Run a Demo SQL query in BigQuery to delete a small list of object URIs from Cloud Storage
Verify that the files are deleted from the Cloud Storage bucket.

Costs¶

This tutorial uses billable components of Google Cloud, including the following:

Use the pricing calculator to generate a cost estimate based on your projected usage.

Before you begin¶

For this tutorial, you need a Google Cloud project. To make cleanup easiest at the end of the tutorial, we recommend that you create a new project for this tutorial.

Create a Google Cloud project
Make sure that billing is enabled for your Google Cloud project
Open Cloud Shell At the bottom of the Cloud Console, a Cloud Shell session opens and displays a command-line prompt. Cloud Shell is a shell environment with the Cloud SDK already installed, including the gcloud command-line tool, and with values already set for your current project. It can take a few seconds for the session to initialize.

In Cloud Shell, clone the source repository and go to the directory for this tutorial:

git clone https://github.com/GoogleCloudPlatform/cloud-solutions.git
cd cloud-solutions/projects/bulk-delete-gcs-files-fn

Enable the required APIs for the services required for this tutorial:

gcloud services enable \
  artifactregistry.googleapis.com \
  bigquery.googleapis.com \
  bigqueryconnection.googleapis.com \
  cloudbuild.googleapis.com \
  run.googleapis.com

Deploy the remote function¶

Authenticate using User Application Default Credentials (ADC).
```
gcloud auth application-default login
```
Edit the project_id and other fields in example.tfvars file using a text editor of your choice.
Run the Terraform script to create all resources:

The Terraform script creates:
- A BigQuery remote function that accepts a list of Cloud Storage Object URIs to delete. You can call the remote function as follows:
- Cloud Run service that hosts the remote function
- Cloud Storage bucket with a single file named file1
```
terraform init && terraform apply -var-file="example.tfvars"
```

Use the remote function from BigQuery¶

Open BigQuery console and issue following SQL command:

WITH
  filenames AS (
    SELECT filename
    FROM
    (SELECT 'file1' AS filename)
    UNION ALL (SELECT 'file2' AS filename)
  )
  SELECT
    filename,
    `fns_del`.gcs_del('gs://<project-id>-bulk-delete-test/' + filename) AS delete_status
  FROM
    filenames

The command should produce following output:

filename	delete_status
file1	Completed
file2	Not Found

Troubleshooting¶

If you are seeing many errors with message like: Unauthorized, ensure that the Service Account used by the delete function's Cloud Run service has appropriate IAM permissions to the GCS buckets to be used through the BigQuery function.

Cleaning up¶

Caution: Deleting a project has the following effects:

Everything in the project is deleted. If you used an existing project for this tutorial, when you delete it, you also delete any other work you've done in the project.
Custom project IDs are lost. When you created this project, you might have created a custom project ID that you want to use in the future. To preserve the URLs that use the project ID, such as an appspot.com URL, delete selected resources inside the project instead of deleting the whole project.

If you plan to explore multiple tutorials and quickstarts, reusing projects can help you avoid exceeding project quota limits.

To avoid incurring charges to your Google Cloud account for the resources used in this tutorial, you can delete the project:

In the Cloud Console, go to the Manage resources page.
In the project list, select the project that you want to delete and then click Delete
In the dialog, type the project ID and then click Shut down to delete the project.

Disclaimer¶

This is not an official Google product.