Security and Authentication

To protect data from unauthorized access, some Spark data stores require connections to be authenticated with user credentials or encrypted using the SSL protocol. The Simba Apache Spark ODBC Connector provides full support for these authentication protocols.

Note:

In this documentation, "SSL" refers to both TLS (Transport Layer Security) and SSL (Secure Sockets Layer). The connector supports TLS 1.0, 1.1, and 1.2. The SSL version used for the connection is the highest version that is supported by both the connector and the server.

The connector provides mechanisms that enable you to authenticate your connection using the Kerberos protocol, the OAuth 2.0 protocol, an API signing key, token-based authentication, your Spark user name only, or your Spark user name and password. You can also authenticate a connection to an HDInsight distribution on Windows Azure. You must use the authentication mechanism that matches the security requirements of the Spark server. For information about determining the appropriate authentication mechanism to use based on the Spark server configuration, see Authentication Mechanisms. For detailed connector configuration instructions, see Configuring AuthenticationConfiguring Authentication.

Additionally, the connector supports the following types of SSL connections:

  • No identity verification
  • One-way authentication
  • Two-way authentication

It is recommended that you enable SSL whenever you connect to a server that is configured to support it. SSL encryption protects data and credentials when they are transferred over the network, and provides stronger security than authentication alone. For detailed configuration instructions, see Configuring SSL VerificationConfiguring SSL Verification.