Error handling

By default, Cassandra drivers don’t need any special exception handling.

Every exception raised by a driver can be classified as a client exception, server exception, or a timeout.

Server errors are returned directly from the coordinator to the driver and are identical across all of the drivers. Client errors are specific to issues that occur in the driver itself. While all drivers share a driver timeout, other client error vary from driver to driver.

Exceptions are raised in Cassandra drivers by methods such as the Java driver’s CqlSession.execute() or the Python driver’s session.execute().

Cassandra drivers include automatic retry policies for certain error conditions as well as functionality to extend retry policies or implement manual retries. For more information, see the following: