Share via

Supported connection properties

  • Article

This article describes the connection properties supported by the Databricks JDBC Driver (OSS).

Authentication and proxy properties

The following authentication and proxy properties are supported by the Databricks JDBC Driver (OSS). Properties are case insensitive.

Property Default value Description
AsyncExecPollInterval 200 The time in milliseconds between each poll for the asynchronous query execution status. Asynchronous refers to the fact that the RPC call used to execute a query against Spark is asynchronous. It does not mean that JDBC asynchronous operations are supported.
Auth_Flow 0 The OAuth2 authentication flow for the driver connection. This property is required if AuthMech is 11.
Auth_JWT_Key_File null The path to the private key file (PEM format) for JWT authentication.
Auth_JWT_Alg RS256 The algorithm for private key JWT authentication. The supported algorithms are: RSA: RS256, RS384, RS512, PS256, PS384, PS512 and EC: ES256, ES384, ES512
Auth_JWT_Key_Passphrase null The passphrase for decrypting an encrypted private key.
Auth_KID null The Key Identifier (KID) required for JWT authentication. This is mandatory when using private key JWT.
AuthMech Required The authentication mechanism, where 3 specifies the mechanism is a Azure Databricks personal access token, and 11 specifies the mechanism is OAuth 2.0 tokens. Additional properties are required for each mechanism. See Authenticate the driver.
CFProxyAuth 0 If set to 1, the driver uses the proxy authentication user and password, represented by CFProxyUID and CFProxyPwd.
CFProxyHost null A string that represents the name of the proxy host to use when UseCFProxy is also set to 1.
CFProxyPort null An integer that represents the number of the proxy port to use when UseCFProxy is also set to 1.
CFProxyUID null A string that represents the username to use for proxy authentication when CFProxyAuth and UseCFProxy are also set to 1.
CFProxyPwd null A string that represents the password to use for proxy authentication when CFProxyAuth and UseCFProxy are also set to 1.
ConnCatalog or catalog SPARK The name of the default catalog to use.
ConnSchema or schema default The name of the default schema to use. This can be specified either by replacing <schema> in the URL with the name of the schema to use or by setting the ConnSchema property to the name of the schema to use.
GoogleServiceAccount null Enables authentication using a Google service account.
GoogleCredentialsFile null The path to the JSON key file for Google Service account authentication.
EnableOIDCDiscovery 1 If set to 1, the OpenID Connect discovery URL is used.
OIDCDiscoveryEndpoint null The OpenID Connect discovery URL for retrieving the OIDC configuration.
Auth_RefreshToken null The OAuth2 refresh token used to retrieve a new access token.
OAuth2ConnAuthAuthorizeEndpoint null The authorization endpoint URL used in an OAuth2 flow.
OAuth2ConnAuthTokenEndpoint null The token endpoint URL for the OAuth2 flow.
ProxyAuth 0 If set to 1, the driver uses the proxy authentication user and password, represented by ProxyUID and ProxyPwd.
ProxyHost null A string that represents the name of the proxy host to use when UseProxy is also set to 1.
ProxyPort null An integer that represents the number of the proxy port to use when UseProxy is also set to 1.
ProxyPwd null A string that represents the password to use for proxy authentication when ProxyAuth and UseProxy are also set to 1.
ProxyUID null A string that represents the username to use for proxy authentication when ProxyAuth and UseProxy are also set to 1.
UseProxy 0 If set to 1, the driver uses the provided proxy settings (for example: ProxyAuth, ProxyHost, ProxyPort, ProxyPwd, and ProxyUID).
UseSystemProxy 0 If set to 1, the driver uses the proxy settings that have been set at the system level. If any additional proxy properties are set in the connection URL, these additional proxy properties override those that have been set at the system level.
UseCFProxy 0 If set to 1, the driver uses the cloud fetch proxy settings if they are provided, otherwise use the regular proxy.
UseJWTAssertion false Enables private key JWT authentication for M2M use cases where client secret authentication is restricted.

SSL trust store configuration properties

The following SSL trust store configuration properties are supported by the Databricks JDBC Driver (OSS). Properties are case insensitive.

Property Default value Description
AllowSelfSignedCerts 0 If set to 1, the driver allows connections to servers with self-signed SSL certificates.
CheckCertificateRevocation 0 If set to 1, the driver checks whether the SSL certificate has been revoked.
SSL 1 Whether the connector communicates with the Spark server through an SSL-enabled socket.
SSLTrustStore null The path to the trust store file for SSL certificate validation.
SSLTrustStorePassword null The password for the trust store file, if it is password-protected.
SSLTrustStoreType JKS The type of the trust store, for example, JKS or PKCS12. If not specified, the driver defaults to JKS trust store. Valid types are JKS, PKCS12, and BCFKS.
UseSystemTrustStore 0 If set to 1, the driver uses the system's default trust store for SSL certificate verification.

Trust store types

The JDBC driver supports the following SSL modes and trust store types.

Self-signed certificate mode

To use self-signed certificate mode, set the connection property AllowSelfSignedCerts=1. This mode uses a trust-all socket factory that accepts any certificate.

Custom trust store

To use a custom trust store, specify a custom trust store file in the SSLTrustStore connection property. This trust store is loaded directly from the specified path and uses the certificates for SSL certificate validation. It can be in JKS, PKCS12, or other supported formats.

You must specify the following additional connection properties:

  • SSLTrustStore: Path to the trust store file
  • SSLTrustStorePassword: Password for the trust store (if needed)
  • SSLTrustStoreType: Type of trust store (optional, defaults to JKS if not specified)

Java system property trust store

To use the system property trust store, set UseSystemTrustStore=1 and make sure that you do not specify a custom trust store. Instead, specify a trust store using the Java system property javax.net.ssl.trustStore. This property is set at the JVM level using the -D flag, for example:

java -Djavax.net.ssl.trustStore=/path/to/truststore.jks -Djavax.net.ssl.trustStorePassword=changeit ...

The JDBC driver first checks for the Java system property javax.net.ssl.trustStore. If it is set, it uses this trust store file instead of the JDK's default. If no system property is set, uses the JDK's default trust store (cacerts), which is ocated at $JAVA_HOME/lib/security/cacerts or similar path.

JDK default trust store (cacerts)

The JDK comes with a built-in trust store called cacerts which contains certificates from well-known Certificate Authorities, which allows verification of certificates issued by those CAs. This trust store is typically located at $JAVA_HOME/lib/security/cacerts with a default password "changeit" or "changeme".

To use the JDK default trust store, set UseSystemTrustStore=1 and make sure that you do not specify a custom trust store or a Java system property trust store. If a trust store is also specified using the Java system property javax.net.ssl.trustStore, it is ignored, which ensures the driver only uses certificates from the default JDK trust store.

Trust store order of precedence

The driver uses the following order of precedence to determine which trust store to use:

  1. The custom trust store specified in the SSLTrustStore connection property
  2. The trust store specified in the Java system property javax.net.ssl.trustStore (when UseSystemTrustStore=1)
  3. The JDK's default trust store (cacerts)

Security recommendations

To keep your connection secure, Databricks recommends the following:

  • For production environments:

    • Do not use self-signed certificate mode (AllowSelfSignedCerts=1).
    • Use official CA-signed certificates.
    • Use UseSystemTrustStore=1 unless you need a custom trust store.

  • For custom trust stores:

    • Use when connecting to servers with certificates not in the default trust store.
    • Ensure the trust store contains the complete certificate chain.
    • Protect trust store files with appropriate permissions.

SQL configuration properties

The following SQL configuration properties are supported by the Databricks JDBC Driver (OSS). These are also described in Configuration parameters. Properties are case insensitive.

Property Default value Description
ansi_mode TRUE Whether to enable strict ANSI SQL behavior for certain functions and casting rules.
enable_photon TRUE Whether to enable the Photon vectorized query engine.
legacy_time_parser_policy EXCEPTION The methods used to parse and format dates and timestamps. Valid values are EXCEPTION, LEGACY, and CORRECTED.
max_file_partition_bytes 128m The maximum number of bytes to pack into a single partition when reading from file based sources. The setting can be any positive integer and optionally include a measure such as b (bytes), k or kb (1024 bytes).
read_only_external_metastore false Controls whether an external metastore is treated as read-only.
statement_timeout 172800 Sets a SQL statement timeout between 0 and 172800 seconds.
timezone UTC Set the local timezone. Region IDs in the form area/city, such as America/Los_Angeles or zone offsets in the format (+|-)HH, (+|-)HH:mm or (+|-)HH:mm:ss, e.g -08, +01:00 or -13:33:33. Also, UTC is supported as an alias for +00:00
use_cached_result true Whether Databricks SQL caches and reuses results whenever possible.

Logging properties

The following logging properties are supported by the Databricks JDBC Driver (OSS). Properties are case insensitive.

Property Default value Description
LogLevel OFF The logging level, which is a value 0 through 6:
  • 0: Disable all logging.
  • 1: Enable logging on the FATAL level, which logs very severe error events that will lead the connector to abort.
  • 2: Enable logging on the ERROR level, which logs error events that might still allow the connector to continue running.
  • 3: Enable logging on the WARNING level, which logs events that might result in an error if action is not taken.
  • 4: Enable logging on the INFO level, which logs general information that describes the progress of the connector.
  • 5: Enable logging on the DEBUG level, which logs detailed information that is useful for debugging the connector.
  • 6: Enable logging on the TRACE level, which logs all connector activity.

Use this property to enable or disable logging in the connector and to specify the amount of detail included in log files.
LogPath To determine the default path for logs, the driver uses the value set for these system properties, in this priority order:
  • user.dir
  • java.io.tmpdir
  • the current directory, in other words .
 The full path to the folder where the connector saves log files when logging is enabled, as a string. To ensure that the connection URL is compatible with all JDBC applications, escape the backslashes (\) in your file path by typing another backslash.
If the LogPath value is invalid, the connector sends the logged information to the standard output stream (System.out).
LogFileSize No maximum The maximum allowed log file size, specified in MB
LogFileCount No maximum The maximum number of allowed log files

Enable and configure logging

The JDBC driver supports the Simple Logging Facade for Java (SLF4J) and java.util.logging (JUL) frameworks. The driver uses the JUL logging framework by default.

To enable and configure logging for the JDBC driver:

  1. Enable the logging framework that you want to use:

    • For SLF4J logging, set the system property -Dcom.databricks.jdbc.loggerImpl=SLF4JLOGGER and provide the SLF4J binding implementation (compatible with SLF4J version 2.0.13 and above) and corresponding configuration file in the classpath.
    • For JUL logging, set the system property -Dcom.databricks.jdbc.loggerImpl=JDKLOGGER. This is the default.

  2. Set the LogLevel property on the connection string to the desired level of information to include in log files.

  3. Set the LogPath property on the connection string to the full path to the folder where you want to save log files.

    For example, the following connection URL enables logging level 6 and saves the log files to the C:temp folder:

    jdbc: databricks://localhost:11000;LogLevel=6;LogPath=C:\\temp

  4. Restart your JDBC application and reconnect to the server to apply the settings.

Other feature properties

The following properties enable features in the Databricks JDBC Driver (OSS). Properties are case insensitive.

Property Default value Description
EnableComplexDatatypeSupport 0 If set to 1, support for complex data types (ARRAYs, STRUCTs, MAPs) as native Java objects instead of strings is enabled.
EnableTelemetry 0 If set to 1, telemetry is enabled. See Telemetry.
UserAgentEntry browser The User-Agent entry to be included in the HTTP request. This value is in the following format: [ProductName]/[ProductVersion] [Comment]
UseThriftClient 1 Whether the JDBC driver should use the Thrift client or Statement Execution APIs.
VolumeOperationAllowedLocalPaths `` The comma separated list of allowed local paths for downloading and uploading of UC Volume Ingestion files. The paths include sub-directories as well. See Manage files using volumes.
VolumeOperationRetryableHttpCode 408,502,503,504 The comma separated list of retryable HTTP codes for Unity Catalog volume ingestion.
VolumeOperationRetryTimeout 15 The retry timeout in minutes for Unity Catalog volume ingestion HTTP requests.

Telemetry collection

Telemetry enables Databricks to streamline debugging and provide timely troubleshooting by collecting:

  • Client environment details (driver version, runtime, OS details)
  • JDBC connection configurations (excludes any PII data)
  • Operation latency measurements
  • Execution result format (inline JSON, Arrow, etc.)
  • Operation types (execution query, metadata query, volume operations)
  • Error classification data
  • Retry counts

Note

Databricks maintains strict privacy standards ensuring no collection of query content, results, or personally identifiable information (PII).