BigQuery: Restore from PiTR

Who can use this feature

Anyone with these permissions: Restore resources

This article explains how to restore a BigQuery dataset from a PiTR backup. You can restore all tables in the dataset, or select specific tables to restore.

PiTR restores use BigQuery’s native time travel capability. Views, materialized views, routines, and tags aren’t restored. These must be recreated manually after the restore completes.

To restore from a standard Eon snapshot instead, see BigQuery: Restore a Dataset.

Restore a BigQuery dataset from PiTR

Before you start

Before you restore a BigQuery dataset from a PiTR backup, make sure the restore account has restore permissions version 1.3.4 or later.

Find the dataset and choose a restore point

In your inventory, select the BigQuery dataset you want to restore. The Resource details panel opens.
Click the Eon snapshots tab. You’ll see a calendar view with colored dots on the days when recovery points are available.

When you click a day, the available recovery points for that day are displayed below the calendar.
Click the BigQuery PiTR tab. You’ll see the available recovery points for the selected day.
Select the point in time you want to restore to, and click Restore. The Restore BigQuery wizard opens.

Choose the restore account

Choose the GCP project you want to restore to, and then click Next.

Configure the restore

In the Configuration step, set the following:

Under Restore granularity, choose whether to restore All tables or only Selected tables.

If you choose Selected tables, set the tables you want to restore.
Enter a Destination Dataset ID for the restored dataset.

Make sure the dataset ID you enter doesn’t have tables with the same names as those you’re restoring. If a table with the same name already exists in the destination dataset, the restore job will fail.
Choose a Location for the restored dataset.

You can set a specific region, or set a multi-region dataset by choosing US or EU.
Click Restore to start the asynchronous restore job.

Monitor the restore job

You can see your list of restore jobs on the Restore jobs tab in the Jobs page.

After the job is completed, the status changes to SUCCESS, and you can work with the restored dataset as normal.

For more information about restore job statuses, see Restore Job Statuses.