Cassandra Connector

This package provides a helpful wrapper around the cassandra-driver package with the following features:

1. References environment variables for seamless connections to Apache Cassandra® and DataStax® Astra DB
2. ConnectionManager maintains a dictionary of cluster connections, facilitating connections to multiple clusters in a single place.
3. Automatically downloads Astra secure connect bundles, based on provided endpoint.

Installing

pip install cassandra-connector

Using

Connecting to Astra

Connecting to Astra with Environment Variables

The simplest approach is to set appropriate environment variables:

ASTRA_DB_APPLICATION_TOKEN=AstraCS:<your token here>
ASTRA_DB_API_ENDPOINT=https://<your endpoint here>

and then:

from cassandra_connector import CassandraConnectorManager

cm = CassandraConnectorManager()
astra = cm.get_connector('env_astra')
session = astra.session

which will get a driver Session object based on the ASTRA environment variables.

Connecting to Astra Directly

You can bypass the CassandraConnectorManager and use the CassandraConnector directly by passing a keyword dict to the astra parameter, including both token and endpoint parameters:

from cassandra_connector import CassandraConnector

astra = CassandraConnector(astra={"token": "AstraCS:<your token here>", "endpoint": "https://<your endpoint here>"})
session = astra.session

Connecting to Cassandra

Connecting to a Cassandra, DataStax Enterprise (DSE), or any other Cassandra-compatible cluster is similar to the driver connection, but as a single step invocation.

There are two parameters introduced with the CassandraConnector; these are used to construct the authentication provider object that is passed into the Cluster constructor:

• authProviderClass is a string representing the Python package of the provider (it must be import-able)
• authProviderArgs is a dict of keyword arguments that are passed to the provider class

All other arguments will be passed directly to the Cluster constructor.

Note: If you do not need an auth provider to connect, you may omit these parameters. This is generally not advised for any production enviroment as it means that anybody with access to your cluster can access the data within the cluster.

Connecting to Cassandra with Environment Variables

If you want to use env_cassandra, represent them as a single JSON document in the CASSANDRA_CONNECTION variable, for example:

CASSANDRA_CONNECTION={"authProviderClass": "cassandra.auth.PlainTextAuthProvider", "authProviderArgs": {"username": "cassandra", "password": "cassandra"}, "contact_points": ["localhost"], "port": 9042}

Once set, you can get a Session:

from cassandra_connector import CassandraConnectorManager

cm = CassandraConnectorManager()
cassandra = cm.get_connector('env_cassandra')
session = cassandra.session

Connecting to Cassandra Directly

You can provide these same arguments directly to the CassandraConnector:

from cassandra_connector import CassandraConnector

cassandra = CassandraConnector(
    authProviderClass="cassandra.auth.PlainTextAuthProvider",
    authProviderArgs={"username": "cassandra", "password": "cassandra"},
    contact_points=["localhost"],
    port=9042)
session = cassandra.session

The CassandraConnectorManager Object

In addition to being able to create CassandraConnector objects from environment variables, you can use the get_connector function along with a dict key of your choosing (other than the above env_* keys). The first time a key is used, you must pass connection parameters as you would connecting "directly" as documented above, but subsequently you can simply pass the key and the CassandraConnector will be returned.

The CassandraConnector Object

As seen above, this object has a .session property which is the native driver Session object. The object also has a .cluster property which gives access to the native driver Cluster object.

The session() function has two boolean parameters:

• replace will close the existing session and replace it with a new session; you may wish to do this if you have changed something in the underlying Cluster object for example.
• new will create a new (and detached) Session object.

For More Details

Docstrings are the best reference for more detailed usage for the connection arguments.

Contributing

Contributions welcome, just fire off a pull request :)