---
url: /v6.2/en-us/connecting-spark-sql.md
---
# Connecting to Spark SQL

### Operation Steps

Please follow the steps below to connect to a `Spark SQL` data source.

1. On the data connection page, click "New Data Connection" in the upper right corner.

   ![](/images/data-connection/connection-001.png)

2. In the data source types, select the `Spark SQL` data source.
   ![](/images/data-connection/connecting-sparksql-001.png)

3. Fill in the required parameters for the data source connection as prompted.
   ![](/images/data-connection/connecting-sparksql-002.png)

   **Connection Configuration Information Description**

   | Field                                  | Description                                                                 |
   | ----------- | --------------------------------------------------- |
   | Name                                   | The name of the connection. Required and unique within the user.            |
   | Host Address                           | The address of the database. If the URL field is filled, the URL takes precedence. |
   | Port                                   | The port of the database. If the URL field is filled, the URL takes precedence.    |
   | Username                               | The username for the database.                                              |
   | Password                               | The password for the database.                                              |
   | Database                               | The name of the database.                                                   |
   | Max Connections                        | The maximum number of connections in the connection pool.                   |
   | Encoding                               | The encoding setting for the database connection.                           |
   | Prefer using database comment as dataset title | Choose whether to display the table name or the table comment. When enabled, the title is shown; when disabled, the table comment is shown. |
   | URL                                    | The JDBC URL of the database.                                               |
   | Hierarchical loading of schema and tables | Disabled by default. When enabled, schemas and tables are loaded hierarchically. Only schemas are loaded during connection, and you need to click the schema to load the tables under it. |
   | Query Timeout (seconds)                 | Default is 600. For large data volumes, you can increase the timeout as needed. |
   | Show only tables under the specified database/schema | When this option is selected and the database field is not empty, only tables under the specified database are shown. |

   4. After filling in the parameters, click the `Validate` button to get the validation result (this checks the connectivity between HENGSHI SENSE and the configured data connection; you cannot add the connection if validation fails).

   5. After validation passes, click `Execute Preset Code` to pop up the preset code for this data source, then click the execute button.

   6. Click the `Add` button to add the `Spark SQL` connection.

> \[!CAUTION] Please Note
>
> 1. Parameters marked with \* are required; others are optional.
> 2. When connecting to a data source, you must execute the preset code. Failure to do so will result in certain functions being unavailable during data analysis. In addition, when upgrading from a version prior to 4.4 to 4.4, you need to execute the preset code for existing data connections in the system.

***

### Supported `Spark SQL` Versions

`2.3.0` and above

### Data Connection Preview Support

Supports all tables that can be listed by `show tables`.

### SQL Dataset Support for SQL

Only `SELECT statements` are supported, along with all SELECT-related features supported by the connected Spark instance. Users need to ensure that the syntax complies with the `Spark SQL` standard.

### Supported Authentication Methods for Connecting to Spark Thrift Server

Supports username and password authentication; SSL is not supported.

### Unsupported Field Types

The following data types in Spark cannot be processed correctly:

* BINARY
* arrays
* maps
* structs
* union