Skip to main content

Cassandra

Cassandra is a NoSQL, row-oriented, highly scalable and highly available database.

CassandraByteStore needs the cassio package to be installed:

%pip install --upgrade --quiet  cassio

The Store takes the following parameters:

  • table: The table where to store the data.
  • session: (Optional) The cassandra driver session. If not provided, the cassio resolved session will be used.
  • keyspace: (Optional) The keyspace of the table. If not provided, the cassio resolved keyspace will be used.
  • setup_mode: (Optional) The mode used to create the Cassandra table (SYNC, ASYNC or OFF). Defaults to SYNC.

CassandraByteStore​

The CassandraByteStore is an implementation of ByteStore that stores the data in your Cassandra instance. The store keys must be strings and will be mapped to the row_id column of the Cassandra table. The store bytes values are mapped to the body_blob column of the Cassandra table.

from langchain_community.storage import CassandraByteStore
API Reference:CassandraByteStore

Init from a cassandra driver Session​

You need to create a cassandra.cluster.Session object, as described in the Cassandra driver documentation. The details vary (e.g. with network settings and authentication), but this might be something like:

from cassandra.cluster import Cluster

cluster = Cluster()
session = cluster.connect()

You need to provide the name of an existing keyspace of the Cassandra instance:

CASSANDRA_KEYSPACE = input("CASSANDRA_KEYSPACE = ")

Creating the store:

store = CassandraByteStore(
    table="my_store",
    session=session,
    keyspace=CASSANDRA_KEYSPACE,
)

store.mset([("k1", b"v1"), ("k2", b"v2")])
print(store.mget(["k1", "k2"]))
[b'v1', b'v2']

Init from cassio​

It's also possible to use cassio to configure the session and keyspace.

import cassio

cassio.init(contact_points="127.0.0.1", keyspace=CASSANDRA_KEYSPACE)

store = CassandraByteStore(
    table="my_store",
)

store.mset([("k1", b"v1"), ("k2", b"v2")])
print(store.mget(["k1", "k2"]))

Usage with CacheBackedEmbeddings​

You may use the CassandraByteStore in conjunction with a CacheBackedEmbeddings to cache the result of embeddings computations.

from langchain.embeddings import CacheBackedEmbeddings
from langchain_openai import OpenAIEmbeddings

cassio.init(contact_points="127.0.0.1", keyspace=CASSANDRA_KEYSPACE)

store = CassandraByteStore(
    table="my_store",
)

embeddings = CacheBackedEmbeddings.from_bytes_store(
    underlying_embeddings=OpenAIEmbeddings(), document_embedding_cache=store
)

Was this page helpful?


You can leave detailed feedback on GitHub.