External tables for Cloud Storage

Console

1. Go to the BigQuery page.

   Go to BigQuery

2. In the left pane, click explore Explorer:

   

   If you don't see the left pane, click last_page Expand left pane to open the pane.

3. In the Explorer pane, expand your project, click Datasets, and then click a dataset.

4. Click Create table. The Create table pane opens.

5. In the Source section, specify the following details:

   1. For Create table from, select Google Cloud Storage.

   2. Provide the path to the folder, using wildcards. For example, my_bucket/my_files*. The folder must be in the same location as the dataset that contains the table you want to create, append, or overwrite.

   3. From the File format list, select the file type.

   4. Select the Source data partitioning checkbox, and then specify the following details:

      1. For Select Source URI Prefix, enter the URI prefix. For example, gs://my_bucket/my_files.
      2. Optional: To require a partition filter on all queries for this table, select the Require partition filter checkbox. Requiring a partition filter can reduce cost and improve performance. For more information, see Requiring predicate filters on partition keys in queries.

      3. In the Partition inference mode section, select one of the following options:

         • Automatically infer types: set the partition schema detection mode to AUTO.
         • All columns are strings: set the partition schema detection mode to STRINGS.
         • Provide my own: set the partition schema detection mode to CUSTOM and manually enter the schema information for the partition keys. For more information, see Provide a custom partition key schema.

6. In the Destination section, specify the following details:

   1. For Project, select the project in which you want to create the table.
   2. For Dataset, select the dataset in which you want to create the table.
   3. For Table, enter the name of the table that you want to create.
   4. For Table type, select External table.
   5. Select the Create a BigLake table using a Cloud Resource connection checkbox.
   6. For Connection ID, select the connection that you created earlier.

7. In the Schema section, enable schema auto-detection by selecting the Auto detect option.

8. To ignore rows with extra column values that don't match the schema, expand the Advanced options section and select Unknown values.

9. Click Create table.

SQL

Use the CREATE EXTERNAL TABLE DDL statement:

1. In the Google Cloud console, go to the BigQuery page.

   Go to BigQuery

2. In the query editor, enter the following statement:

   CREATE EXTERNAL TABLE `PROJECT_ID.DATASET.EXTERNAL_TABLE_NAME`
   WITH PARTITION COLUMNS
   (
     PARTITION_COLUMN PARTITION_COLUMN_TYPE,
   )
   WITH CONNECTION {`PROJECT_ID.REGION.CONNECTION_ID` | DEFAULT}
   OPTIONS (
     hive_partition_uri_prefix = "HIVE_PARTITION_URI_PREFIX",
     uris=['FILE_PATH'],
     max_staleness = STALENESS_INTERVAL,
     metadata_cache_mode = 'CACHE_MODE',
     format ="TABLE_FORMAT"
   );

   Replace the following:

   • PROJECT_ID: the name of your project in which you want to create the table—for example, myproject
   • DATASET: the name of the BigQuery dataset that you want to create the table in—for example, mydataset
   • EXTERNAL_TABLE_NAME: the name of the table that you want to create—for example, mytable
   • PARTITION_COLUMN: the name of the partitioning column
   • PARTITION_COLUMN_TYPE: the type of the partitioning column
   • REGION: the region that contains the connection—for example, us
   • CONNECTION_ID: the connection ID—for example, myconnection

     When you view the connection details in the Google Cloud console, the connection ID is the value in the last section of the fully qualified connection ID that is shown in Connection ID—for example projects/myproject/locations/connection_location/connections/myconnection.

     To use a default connection, specify DEFAULT instead of the connection string containing PROJECT_ID.REGION.CONNECTION_ID.

   • HIVE_PARTITION_URI_PREFIX: hive partitioning uri prefix–for example, gs://mybucket/
   • FILE_PATH: path to the data source for the external table that you want to create—for example, gs://mybucket/*.parquet
   • STALENESS_INTERVAL: specifies whether cached metadata is used by operations against the BigLake table, and how fresh the cached metadata must be in order for the operation to use it. For more information about metadata caching considerations, see Metadata caching for performance.

     To disable metadata caching, specify 0. This is the default.

     To enable metadata caching, specify an interval literal value between 30 minutes and 7 days. For example, specify INTERVAL 4 HOUR for a 4 hour staleness interval. With this value, operations against the table use cached metadata if it has been refreshed within the past 4 hours. If the cached metadata is older than that, the operation retrieves metadata from Cloud Storage instead.

   • CACHE_MODE: specifies whether the metadata cache is refreshed automatically or manually. For more information about metadata caching considerations, see Metadata caching for performance.

     Set to AUTOMATIC for the metadata cache to be refreshed at a system-defined interval, usually somewhere between 30 and 60 minutes.

     Set to MANUAL if you want to refresh the metadata cache on a schedule you determine. In this case, you can call the BQ.REFRESH_EXTERNAL_METADATA_CACHE system procedure to refresh the cache.

     You must set CACHE_MODE if STALENESS_INTERVAL is set to a value greater than 0.

   • TABLE_FORMAT: the format of the table that you want to create—for example, PARQUET

3. Click play_circle Run.

For more information about how to run queries, see Run an interactive query.

Examples

The following example creates a BigLake table over partitioned data where:

• The schema is autodetected.
• The metadata cache staleness interval for the table is 1 day.
• The metadata cache refreshes automatically.

CREATE EXTERNAL TABLE `my_dataset.my_table`
WITH PARTITION COLUMNS
(
  sku STRING,
)
WITH CONNECTION `us.my-connection`
OPTIONS(
  hive_partition_uri_prefix = "gs://mybucket/products",
  uris = ['gs://mybucket/products/*'],
  max_staleness = INTERVAL 1 DAY,
  metadata_cache_mode = 'AUTOMATIC'
);

The following example creates a BigLake table over partitioned data where:

• The schema is specified.
• The metadata cache staleness interval for the table is 8 hours.
• The metadata cache must be manually refreshed.

CREATE EXTERNAL TABLE `my_dataset.my_table`
(
  ProductId INTEGER,
  ProductName STRING,
  ProductType STRING
)
WITH PARTITION COLUMNS
(
  sku STRING,
)
WITH CONNECTION `us.my-connection`
OPTIONS(
  hive_partition_uri_prefix = "gs://mybucket/products",
  uris = ['gs://mybucket/products/*'],
  max_staleness = INTERVAL 8 HOUR,
  metadata_cache_mode = 'MANUAL'
);

bq

First, use the bq mkdef command to create a table definition file:

bq mkdef \
--source_format=SOURCE_FORMAT \
--connection_id=REGION.CONNECTION_ID \
--hive_partitioning_mode=PARTITIONING_MODE \
--hive_partitioning_source_uri_prefix=GCS_URI_SHARED_PREFIX \
--require_hive_partition_filter=BOOLEAN \
--metadata_cache_mode=CACHE_MODE \
 GCS_URIS > DEFINITION_FILE

Replace the following:

• SOURCE_FORMAT: the format of the external data source. For example, CSV.
• REGION: the region that contains the connection—for example, us.

• CONNECTION_ID: the connection ID—for example, myconnection.

  When you view the connection details in the Google Cloud console, the connection ID is the value in the last section of the fully qualified connection ID that is shown in Connection ID—for example projects/myproject/locations/connection_location/connections/myconnection.

  To use a default connection, specify DEFAULT instead of the connection string containing PROJECT_ID.REGION.CONNECTION_ID.

• PARTITIONING_MODE: the Hive partitioning mode. Use one of the following values:

  • AUTO: Automatically detect the key names and types.
  • STRINGS: Automatically convert the key names to strings.
  • CUSTOM: Encode the key schema in the source URI prefix.

• GCS_URI_SHARED_PREFIX: the source URI prefix.

• BOOLEAN: specifies whether to require a predicate filter at query time. This flag is optional. The default value is false.

• CACHE_MODE: specifies whether the metadata cache is refreshed automatically or manually. You only need to include this flag if you also plan to use the --max_staleness flag in the subsequent bq mk command to enable metadata caching. For more information on metadata caching considerations, see Metadata caching for performance.

  Set to AUTOMATIC for the metadata cache to be refreshed at a system-defined interval, usually somewhere between 30 and 60 minutes.

  Set to MANUAL if you want to refresh the metadata cache on a schedule you determine. In this case, you can call the BQ.REFRESH_EXTERNAL_METADATA_CACHE system procedure to refresh the cache.

  You must set CACHE_MODE if STALENESS_INTERVAL is set to a value greater than 0.

• GCS_URIS: the path to the Cloud Storage folder, using wildcard format.

• DEFINITION_FILE: the path to the table definition file on your local machine.

If PARTITIONING_MODE is CUSTOM, include the partition key schema in the source URI prefix, using the following format:

--hive_partitioning_source_uri_prefix=GCS_URI_SHARED_PREFIX/{KEY1:TYPE1}/{KEY2:TYPE2}/...

After you create the table definition file, use the bq mk command to create the BigLake table:

bq mk --external_table_definition=DEFINITION_FILE \
--max_staleness=STALENESS_INTERVAL \
DATASET_NAME.TABLE_NAME \
SCHEMA

Replace the following:

• DEFINITION_FILE: the path to the table definition file.

• STALENESS_INTERVAL: specifies whether cached metadata is used by operations against the BigLake table, and how fresh the cached metadata must be in order for the operation to use it. If you include this flag, you must have also specified a value for the --metadata_cache_mode flag in the preceding bq mkdef command. For more information on metadata caching considerations, see Metadata caching for performance.

  To disable metadata caching, specify 0. This is the default.

  To enable metadata caching, specify an interval value between 30 minutes and 7 days, using the Y-M D H:M:S format described in the INTERVAL data type documentation. For example, specify 0-0 0 4:0:0 for a 4 hour staleness interval. With this value, operations against the table use cached metadata if it has been refreshed within the past 4 hours. If the cached metadata is older than that, the operation retrieves metadata from Cloud Storage instead.

• DATASET_NAME: the name of the dataset that contains the table.

• TABLE_NAME: the name of the table you're creating.

• SCHEMA: specifies a path to a JSON schema file, or specifies the schema in the form field:data_type,field:data_type,.... To use schema auto-detection, omit this argument.

Examples

The following example uses AUTO Hive partitioning mode, and also sets the metadata cache to have a 12 hour staleness interval and to get refreshed automatically:

bq mkdef --source_format=CSV \
  --connection_id=us.my-connection \
  --hive_partitioning_mode=AUTO \
  --hive_partitioning_source_uri_prefix=gs://myBucket/myTable \
  --metadata_cache_mode=AUTOMATIC \
  gs://myBucket/myTable/* > mytable_def

bq mk --external_table_definition=mytable_def \
  --max_staleness=0-0 0 12:0:0 \
  mydataset.mytable \
  Region:STRING,Quarter:STRING,Total_sales:INTEGER

The following example uses STRING Hive partitioning mode:

bq mkdef --source_format=CSV \
  --connection_id=us.my-connection \
  --hive_partitioning_mode=STRING \
  --hive_partitioning_source_uri_prefix=gs://myBucket/myTable \
  gs://myBucket/myTable/* > mytable_def

bq mk --external_table_definition=mytable_def \
  mydataset.mytable \
  Region:STRING,Quarter:STRING,Total_sales:INTEGER

The following example uses CUSTOM Hive partitioning mode:

bq mkdef --source_format=CSV \
  --connection_id=us.my-connection \
  --hive_partitioning_mode=CUSTOM \
  --hive_partitioning_source_uri_prefix=gs://myBucket/myTable/{dt:DATE}/{val:STRING} \
  gs://myBucket/myTable/* > mytable_def

bq mk --external_table_definition=mytable_def \
  mydataset.mytable \
  Region:STRING,Quarter:STRING,Total_sales:INTEGER

API

To set Hive partitioning using the BigQuery API, include the hivePartitioningOptions object in the ExternalDataConfiguration object when you create the table definition file. To create a BigLake table, you must also specify a value for the connectionId field.

If you set the hivePartitioningOptions.mode field to CUSTOM, you must encode the partition key schema in the hivePartitioningOptions.sourceUriPrefix field as follows: gs://BUCKET/PATH_TO_TABLE/{KEY1:TYPE1}/{KEY2:TYPE2}/...

To enforce the use of a predicate filter at query time, set the hivePartitioningOptions.requirePartitionFilter field to true.

Terraform

This example creates a BigLake Table on partitioned data.

To authenticate to BigQuery, set up Application Default Credentials. For more information, see Set up authentication for client libraries.

# This creates a bucket in the US region named "my-bucket" with a pseudorandom
# suffix.
resource "random_id" "default" {
  byte_length = 8
}
resource "google_storage_bucket" "default" {
  name                        = "my-bucket-${random_id.default.hex}"
  location                    = "US"
  force_destroy               = true
  uniform_bucket_level_access = true
}

resource "google_storage_bucket_object" "default" {
  # This creates a fake message to create partition locations on the table.
  # Otherwise, the table deployment fails.
  name    = "publish/dt=2000-01-01/hr=00/min=00/fake_message.json"
  content = "{\"column1\": \"XXX\"}"
  bucket  = google_storage_bucket.default.name
}

# This queries the provider for project information.
data "google_project" "default" {}

# This creates a connection in the US region named "my-connection".
# This connection is used to access the bucket.
resource "google_bigquery_connection" "default" {
  connection_id = "my-connection"
  location      = "US"
  cloud_resource {}
}

# This grants the previous connection IAM role access to the bucket.
resource "google_project_iam_member" "default" {
  role    = "roles/storage.objectViewer"
  project = data.google_project.default.id
  member  = "serviceAccount:${google_bigquery_connection.default.cloud_resource[0].service_account_id}"
}

# This makes the script wait for seven minutes before proceeding. This lets IAM
# permissions propagate.
resource "time_sleep" "default" {
  create_duration = "7m"

  depends_on = [google_project_iam_member.default]
}

# This defines a Google BigQuery dataset with default expiration times for
# partitions and tables, a description, a location, and a maximum time travel.
resource "google_bigquery_dataset" "default" {
  dataset_id                      = "my_dataset"
  default_partition_expiration_ms = 2592000000  # 30 days
  default_table_expiration_ms     = 31536000000 # 365 days
  description                     = "My dataset description"
  location                        = "US"
  max_time_travel_hours           = 96 # 4 days

  # This defines a map of labels for the bucket resource,
  # including the labels "billing_group" and "pii".
  labels = {
    billing_group = "accounting",
    pii           = "sensitive"
  }
}

# This creates a BigQuery table with partitioning and automatic metadata
# caching.
resource "google_bigquery_table" "default" {
  dataset_id = google_bigquery_dataset.default.dataset_id
  table_id   = "my_table"
  schema     = jsonencode([{ "name" : "column1", "type" : "STRING", "mode" : "NULLABLE" }])
  external_data_configuration {
    # This defines an external data configuration for the BigQuery table
    # that reads Parquet data from the publish directory of the default
    # Google Cloud Storage bucket.
    autodetect    = false
    source_format = "PARQUET"
    connection_id = google_bigquery_connection.default.name
    source_uris   = ["gs://${google_storage_bucket.default.name}/publish/*"]
    # This configures Hive partitioning for the BigQuery table,
    # partitioning the data by date and time.
    hive_partitioning_options {
      mode                     = "CUSTOM"
      source_uri_prefix        = "gs://${google_storage_bucket.default.name}/publish/{dt:STRING}/{hr:STRING}/{min:STRING}"
      require_partition_filter = false
    }
    # This enables automatic metadata refresh.
    metadata_cache_mode = "AUTOMATIC"
  }


  # This sets the maximum staleness of the metadata cache to 10 hours.
  max_staleness = "0-0 0 10:0:0"

  depends_on = [
    time_sleep.default,
    google_storage_bucket_object.default
  ]
}

To apply your Terraform configuration in a Google Cloud project, complete the steps in the following sections.

Prepare Cloud Shell

1. Launch Cloud Shell.

2. Set the default Google Cloud project where you want to apply your Terraform configurations.

   You only need to run this command once per project, and you can run it in any directory.

   export GOOGLE_CLOUD_PROJECT=PROJECT_ID

   Environment variables are overridden if you set explicit values in the Terraform configuration file.

Prepare the directory

Each Terraform configuration file must have its own directory (also called a root module).

1. In Cloud Shell, create a directory and a new file within that directory. The filename must have the .tf extension—for example main.tf. In this tutorial, the file is referred to as main.tf.

   mkdir DIRECTORY && cd DIRECTORY && touch main.tf

2. If you are following a tutorial, you can copy the sample code in each section or step.

   Copy the sample code into the newly created main.tf.

   Optionally, copy the code from GitHub. This is recommended when the Terraform snippet is part of an end-to-end solution.

3. Review and modify the sample parameters to apply to your environment.
4. Save your changes.
5. Initialize Terraform. You only need to do this once per directory.

   terraform init

   Optionally, to use the latest Google provider version, include the -upgrade option:

   terraform init -upgrade

Apply the changes

1. Review the configuration and verify that the resources that Terraform is going to create or update match your expectations:

   terraform plan

   Make corrections to the configuration as necessary.

2. Apply the Terraform configuration by running the following command and entering yes at the prompt:

   terraform apply

   Wait until Terraform displays the "Apply complete!" message.

3. Open your Google Cloud project to view the results. In the Google Cloud console, navigate to your resources in the UI to make sure that Terraform has created or updated them.