Legacy Azure Storage support#

The Hive connector can be configured to use Azure Data Lake Storage (Gen2). Trino supports Azure Blob File System (ABFS) to access data in ADLS Gen2.

Hive connector configuration for Azure Storage credentials#

To configure Trino to use the Azure Storage credentials, set the following configuration properties in the catalog properties file. It is best to use this type of configuration if the primary storage account is linked to the cluster.

The specific configuration depends on the type of storage and uses the properties from the following sections in the catalog properties file.

For more complex use cases, such as configuring multiple secondary storage accounts using Hadoop’s core-site.xml, see the Advanced configuration options.

ADLS Gen2 / ABFS storage#

To connect to ABFS storage, you may either use the storage account’s access key, or a service principal. Do not use both sets of properties at the same time.

ABFS Access Key#
Property name	Description
`hive.azure.abfs-storage-account`	The name of the ADLS Gen2 storage account
`hive.azure.abfs-access-key`	The decrypted access key for the ADLS Gen2 storage account

ABFS Service Principal OAuth#
Property name	Description
`hive.azure.abfs.oauth.endpoint`	The service principal / application’s OAuth 2.0 token endpoint (v1).
`hive.azure.abfs.oauth.client-id`	The service principal’s client/application ID.
`hive.azure.abfs.oauth.secret`	A client secret for the service principal.

When using a service principal, it must have the Storage Blob Data Owner, Contributor, or Reader role on the storage account you are using, depending on which operations you would like to use.

ADLS Gen1 (legacy)#

While it is advised to migrate to ADLS Gen2 whenever possible, if you still choose to use ADLS Gen1 you need to include the following properties in your catalog configuration.

Note

Credentials for the filesystem can be configured using ClientCredential type. To authenticate with ADLS Gen1 you must create a new application secret for your ADLS Gen1 account’s App Registration, and save this value because you won’t able to retrieve the key later. Refer to the Azure documentation for details.

ADLS properties#
Property name	Description
`hive.azure.adl-client-id`	Client (Application) ID from the App Registrations for your storage account
`hive.azure.adl-credential`	Value of the new client (application) secret created
`hive.azure.adl-refresh-url`	OAuth 2.0 token endpoint url
`hive.azure.adl-proxy-host`	Proxy host and port in `host:port` format. Use this property to connect to an ADLS endpoint via a SOCKS proxy.

WASB storage (legacy)#

WASB properties#
Property name	Description
`hive.azure.wasb-storage-account`	Storage account name of Azure Blob Storage
`hive.azure.wasb-access-key`	The decrypted access key for the Azure Blob Storage

Advanced configuration#

All of the configuration properties for the Azure storage driver are stored in the Hadoop core-site.xml configuration file. When there are secondary storage accounts involved, we recommend configuring Trino using a core-site.xml containing the appropriate credentials for each account.

The path to the file must be configured in the catalog properties file:

hive.config.resources=<path_to_hadoop_core-site.xml>

One way to find your account key is to ask for the connection string for the storage account. The abfsexample.dfs.core.windows.net account refers to the storage account. The connection string contains the account key:

az storage account  show-connection-string --name abfswales1
{
  "connectionString": "DefaultEndpointsProtocol=https;EndpointSuffix=core.windows.net;AccountName=abfsexample;AccountKey=examplekey..."
}

When you have the account access key, you can add it to your core-site.xml or Java cryptography extension (JCEKS) file. Alternatively, you can have your cluster management tool to set the option fs.azure.account.key.STORAGE-ACCOUNT to the account key value:

<property>
  <name>fs.azure.account.key.abfsexample.dfs.core.windows.net</name>
  <value>examplekey...</value>
</property>

For more information, see Hadoop Azure Support: ABFS.

Accessing Azure Storage data#

URI scheme to reference data#

Consistent with other FileSystem implementations within Hadoop, the Azure Standard Blob and Azure Data Lake Storage Gen2 (ABFS) drivers define their own URI scheme so that resources (directories and files) may be distinctly addressed. You can access both primary and secondary storage accounts linked to the cluster with the same URI scheme. Following are example URIs for the different systems.

ABFS URI:

abfs[s]://<file_system>@<account_name>.dfs.core.windows.net/<path>/<path>/<file_name>

ADLS Gen1 URI:

adl://<data_lake_storage_gen1_name>.azuredatalakestore.net/<path>/<file_name>

Azure Standard Blob URI:

wasb[s]://<container>@<account_name>.blob.core.windows.net/<path>/<path>/<file_name>

Querying Azure Storage#

You can query tables already configured in your Hive metastore used in your Hive catalog. To access Azure Storage data that is not yet mapped in the Hive metastore, you need to provide the schema of the data, the file format, and the data location.

For example, if you have ORC or Parquet files in an ABFS file_system, you need to execute a query:

-- select schema in which the table is to be defined, must already exist
USE hive.default;

-- create table
CREATE TABLE orders (
     orderkey BIGINT,
     custkey BIGINT,
     orderstatus VARCHAR(1),
     totalprice DOUBLE,
     orderdate DATE,
     orderpriority VARCHAR(15),
     clerk VARCHAR(15),
     shippriority INTEGER,
     comment VARCHAR(79)
) WITH (
     external_location = 'abfs[s]://<file_system>@<account_name>.dfs.core.windows.net/<path>/<path>/',
     format = 'ORC' -- or 'PARQUET'
);

Now you can query the newly mapped table:

SELECT * FROM orders;

Writing data#

Prerequisites#

Before you attempt to write data to Azure Storage, make sure you have configured everything necessary to read data from the storage.

Create a write schema#

If the Hive metastore contains schema(s) mapped to Azure storage filesystems, you can use them to write data to Azure storage.

If you don’t want to use existing schemas, or there are no appropriate schemas in the Hive metastore, you need to create a new one:

CREATE SCHEMA hive.abfs_export
WITH (location = 'abfs[s]://file_system@account_name.dfs.core.windows.net/<path>');

Write data to Azure Storage#

Once you have a schema pointing to a location where you want to write the data, you can issue a CREATE TABLE AS statement and select your desired file format. The data will be written to one or more files within the abfs[s]://file_system@account_name.dfs.core.windows.net/<path>/my_table namespace. Example:

CREATE TABLE hive.abfs_export.orders_abfs
WITH (format = 'ORC')
AS SELECT * FROM tpch.sf1.orders;

Migration to native Azure implementation#

SEP includes a native Azure implementation that does not rely on the legacy Hadoop implementation. Starburst recommends upgrading existing deployments to the new native implementation at your earliest convenience.

The following considerations apply to legacy configurations moving to the native implementation:

Legacy storage types (WASB and ADL) are not supported.
Azure Government is not supported.

To migrate a catalog to use the native filesystem implementation for Azure, make the following edits to your catalog configuration:

Set the fs.hadoop.enabled=false catalog configuration property.
Add the fs.native-azure.enabled=true catalog configuration property.
Add the azure.auth-type catalog configuration property. Set to ACCESS_KEY for access key authentication or OAUTH for OAuth 2.0 authentication. Leave unconfigured for default authentication.
Refer to the following table to rename your existing legacy catalog configuration properties to the corresponding native configuration properties. Supported configuration values are identical unless otherwise noted.

Legacy property	Native property	Notes
`hive.azure.abfs-access-key`	`azure.access-key`
`hive.azure.abfs.oauth.endpoint`	`azure.oauth.endpoint`	Also see `azure.oauth.tenant-id` in OAuth 2.0 authentication.
`hive.azure.abfs.oauth.client-id`	`azure.oauth.client-id`
`hive.azure.abfs.oauth.secret`	`azure.oauth.secret`
`hive.azure.abfs.oauth2.passthrough`	`azure.use-oauth-passthrough-token`

Remove the following legacy configuration properties if they exist in your catalog configuration:
- hive.azure.abfs-storage-account
- hive.azure.wasb*
- hive.azure.asl*
- warp-speed.config.internal-communication.shared-secret
- warp-speed.use-http-server-port
- warp-speed.config.http-rest-port-enabled