Prior to Looker 24.20, when setting up OAuth authentication for Google BigQuery connections, Looker created OAuth credentials that allowed the database user to request read and write scope. Starting in Looker 24.20, Looker instead requests OAuth read-only scopes for any new BigQuery OAuth connections, new OAuth authorizations to existing BigQuery OAuth connections, and re-authorizations to existing BigQuery OAuth connections.
Starting on March 1, 2025, Looker will sign out any users who have not re-authorized with OAuth read-only scopes from all corresponding BigQuery connections. This will cause any schedules dependent on these connections to fail. Each of these users will need to reauthorize their OAuth connection credentials to ensure uninterrupted schedule delivery. You can also reassign a schedule to a user who has reauthorized their OAuth connection credentials.
To ensure a seamless transition to the updated OAuth credentials, follow the steps in the proceeding sections:
- Reauthorize your OAuth connection credentials
- Generate a list of all users with potentially affected schedules
- (Optional) Force read-only scope usage across your Looker instance
Reauthorize your OAuth connection credentials
To update your OAuth credentials to use a read-only scope, follow these steps:
- Navigate to the Account page.
- In the OAuth Connection Credentials section, click Reauthorize next to each set of credentials.
- You will be prompted to reauthorize Looker to access BigQuery data. The confirmation screen should list the permission "View your data in Google BigQuery" rather than "View and manage your data in Google BigQuery."
Each user that has OAuth credentials for the BigQuery connection will need to complete these steps.
Generate a list of all users with potentially affected schedules
To generate a list of all users without read-only OAuth credentials who have created schedules on your BigQuery connections, visit the following System Activity Explore, replacing INSTANCE_NAME with the address of your Looker instance (such as https://example.cloud.looker.com).
INSTANCE_NAME/explore/system__activity/scheduled_plan_oauth_events?fields=user.name,count,query.model&f[query.model]=-NULL&f[count]=0&sorts=user.name&limit=500&column_limit=50&query_timezone=America%2FLos_Angeles&vis=%7B%22show_view_names%22%3Afalse%2C%22show_row_numbers%22%3Atrue%2C%22transpose%22%3Afalse%2C%22truncate_text%22%3Atrue%2C%22hide_totals%22%3Afalse%2C%22hide_row_totals%22%3Afalse%2C%22size_to_fit%22%3Atrue%2C%22table_theme%22%3A%22white%22%2C%22limit_displayed_rows%22%3Afalse%2C%22enable_conditional_formatting%22%3Afalse%2C%22header_text_alignment%22%3A%22left%22%2C%22header_font_size%22%3A12%2C%22rows_font_size%22%3A12%2C%22conditional_formatting_include_totals%22%3Afalse%2C%22conditional_formatting_include_nulls%22%3Afalse%2C%22x_axis_gridlines%22%3Afalse%2C%22y_axis_gridlines%22%3Atrue%2C%22show_y_axis_labels%22%3Atrue%2C%22show_y_axis_ticks%22%3Atrue%2C%22y_axis_tick_density%22%3A%22default%22%2C%22y_axis_tick_density_custom%22%3A5%2C%22show_x_axis_label%22%3Atrue%2C%22show_x_axis_ticks%22%3Atrue%2C%22y_axis_scale_mode%22%3A%22linear%22%2C%22x_axis_reversed%22%3Afalse%2C%22y_axis_reversed%22%3Afalse%2C%22plot_size_by_field%22%3Afalse%2C%22trellis%22%3A%22%22%2C%22stacking%22%3A%22%22%2C%22legend_position%22%3A%22center%22%2C%22point_style%22%3A%22none%22%2C%22show_value_labels%22%3Afalse%2C%22label_density%22%3A25%2C%22x_axis_scale%22%3A%22auto%22%2C%22y_axis_combined%22%3Atrue%2C%22ordering%22%3A%22none%22%2C%22show_null_labels%22%3Afalse%2C%22show_totals_labels%22%3Afalse%2C%22show_silhouette%22%3Afalse%2C%22totals_color%22%3A%22%23808080%22%2C%22type%22%3A%22looker_grid%22%2C%22defaults_version%22%3A1%2C%22series_types%22%3A%7B%7D%2C%22hidden_fields%22%3A%5B%22count%22%5D%7D&filter_config=%7B%22query.model%22%3A%5B%7B%22type%22%3A%22%21null%22%2C%22values%22%3A%5B%7B%7D%2C%7B%7D%5D%2C%22id%22%3A0%7D%5D%2C%22count%22%3A%5B%7B%22type%22%3A%22%3D%22%2C%22values%22%3A%5B%7B%22constant%22%3A%220%22%7D%2C%7B%7D%5D%2C%22id%22%3A1%7D%5D%2C%22__%21internal%21__%22%3A%5B%22OR%22%2C%5B%5B%22AND%22%2C%5B%5B%22FILTER%22%2C%7B%22field%22%3A%22query.model%22%2C%22value%22%3A%22-NULL%22%2C%22type%22%3A%22%21null%22%7D%5D%2C%5B%22FILTER%22%2C%7B%22field%22%3A%22count%22%2C%22value%22%3A%220%22%7D%5D%5D%5D%5D%5D%7D&dynamic_fields=%5B%7B%22category%22%3A%22measure%22%2C%22expression%22%3Anull%2C%22label%22%3A%22Count%22%2C%22value_format%22%3Anull%2C%22value_format_name%22%3Anull%2C%22based_on%22%3A%22event_attribute.value%22%2C%22_kind_hint%22%3A%22measure%22%2C%22measure%22%3A%22count%22%2C%22type%22%3A%22count_distinct%22%2C%22_type_hint%22%3A%22number%22%2C%22filters%22%3A%7B%22event_attribute.value%22%3A%22%25%2Fauth%2Fbigquery.readonly%25%22%7D%7D%5D&origin=share-expanded
Each of these users will need to reauthorize their OAuth connection credentials to ensure uninterrupted schedule delivery.
(Optional) Force read-only scope usage across your Looker instance
To sign out all your users who have OAuth credentials that allow read and write scope from any of your BigQuery connections, follow these steps:
- Navigate to the Admin settings - General settings page.
- Toggle the Force BigQuery Readonly Scope usage setting to "Enabled" and click Update.
This process does not sign your users back in to BigQuery. Your users will be prompted to sign in to BigQuery the next time they run a query that is based on a model with a BigQuery connection. Any schedules that are dependent on these connections will fail until the user signs in. You can also reassign a schedule to yourself or another user who has already reauthorized their OAuth connection credentials.