Bulk delete data

You can use the Firestore managed bulk delete service to delete data from your database. This feature supports deletion against one or more collection groups.

This page describes how to delete Firestore documents in bulk using the managed bulk delete service. The Firestore managed bulk delete service is available through the gcloud command-line tool and the Firestore API (REST, RPC).

Before you begin

Before you can use the managed bulk delete service, you must complete the following tasks:

  1. Enable billing for your Google Cloud project. Only Google Cloud projects with billing enabled can use the bulk delete functionality.
  2. Make sure your account has the necessary permissions for Firestore. If you are the project owner, your account has the required permissions. Otherwise, the following roles grant the necessary permissions for bulk delete operations:

    • Firestore roles: Owner, Cloud Datastore Owner, or Cloud Datastore Bulk Admin

Set up gcloud for your project

You can initiate bulk delete operations through the Google Cloud console or the gcloud command-line tool. To use gcloud, set up the command-line tool and connect to your project in one of the following ways:

Bulk delete data

A bulk delete operation first finds all applicable documents in your database and deletes them in batches. You may still query or read these documents while the results may vary based on the progress made. Bulk delete will not delete any documents added or modified after the operation starts.

Bulk delete specific collection groups

gcloud

To bulk delete specific collection groups, use the --collection-ids flag. The operation deletes only the collection groups with the given IDs. A collection group includes all documents and nested documents (at any path) with the specified collection groups.

gcloud beta firestore bulk-delete \
--collection-ids=[COLLECTION_GROUP_ID_1_OR_KIND_1],[COLLECTION_GROUP_ID_2_OR_KIND_2],[SUBCOLLECTION_GROUP_ID_1_OR_KIND_3] \
--database=[DATABASE]
For example, imagine multiple documents representing restaurants in the top level restaurants collection of the cymbal database. Under each restaurant document, there are multiple nested sub-collections such as ratings, reviews, and outlets. To bulk delete the restaurants and reviews collection groups, your command looks as follows:

gcloud beta firestore bulk-delete \
--collection-ids=restaurants,reviews \
--database='cymbal'

Manage bulk delete operations

After you start a bulk delete operation, Firestore assigns the operation a unique name. You can use the operation name to delete, cancel, or check the status of the operation.

Operation names are prefixed with projects/[PROJECT_ID]/databases/[DATABASE_ID]/operations/, for example:

projects/my-project/databases/(default)/operations/ASA1MTAwNDQxNAgadGx1YWZlZAcSeWx0aGdpbi1zYm9qLW5pbWRhEgopEg

However, you can leave out the prefix when specifying an operation name for the describe, cancel, and deletecommands.

List all bulk delete operations

gcloud

Use the operations list command to see all running and recently completed operations, including bulk delete operations:

gcloud firestore operations list

Check operation status

gcloud

Use the operations describe command to show the status of a bulk delete operation.

gcloud firestore operations describe [OPERATION_NAME]

Estimate the completion time

A request for the status of a long-running operation returns the metrics workEstimated and workCompleted. Each of these metrics is returned in both number of bytes and number of documents:

  • workEstimated shows the estimated total number of bytes and documents an operation will process. Firestore might omit this metric if it can't make an estimate.

  • workCompleted shows the number of bytes and documents deleted so far. After the operation completes, the value shows the total number of bytes and documents that were actually processed, which might be larger than the value of workEstimated.

Divide workCompleted by workEstimated for a rough progress estimate. This estimate might be inaccurate, because it depends on delayed statistics collection.

Cancel an operation

gcloud

Use the operations cancel command to stop an operation in progress:

gcloud firestore operations cancel [OPERATION_NAME]

Cancelling a running operation doesn't undo the operation. A cancelled bulk delete operation will not recover the deleted documents.

Delete an operation

Use the gcloud firestore operations delete command to remove a completed operation from the list of recent operations. To cancel a running operation, use the earlier cancellation operation.

gcloud firestore operations delete [OPERATION_NAME]

Billing and pricing for bulk delete operations

You are required to enable billing for your Google Cloud project before you use the managed bulk delete service.

Bulk delete operations are charged for document reads and deletes at the rates listed in Firestore pricing. Bulk delete operations incur one index entry read per document found and one delete operation per document deleted. You are charged one read operation for up to 1000 index entries read. For example, for a bulk delete operation that deleted 1500 documents, you will be charged with 2 documents reads and 1500 document deletes.

Note that Firestore charges on the actual work done. If the operation is cancelled or failed due to user error, you will be charged with progress made. Firestore won't charge read or delete for the documents that are not eventually deleted (for example, documents that are modified after the delete operation started). The cost will be attributed on the day of operation completion.

Bulk delete operations will not trigger your Google Cloud budget alerts until after completion. Similarly, reads and deletes performed during a bulk delete operation are applied to your free quota/usage after the operation is complete. Bulk delete operations will not affect the usage shown in the usage section of the console.

View bulk delete costs

Bulk delete operations apply the goog-firestoremanaged:bulkdelete label to billed operations. In the Cloud Billing reports page, you can use this label to view costs related to bulk delete operations.

Best practices

Avoid skipping over deleted data which might slow down the queries.