AstraDB

Configure your own AstraDB instance as data sink.

Configure via UI

Configure via API / Client

Python Client

from llama_cloud.types import CloudAstraDBVectorStore

ds = {
  'name': '<your-name>',
  'sink_type': 'ASTRA_DB', 
  'component': CloudAstraDBVectorStore(
      token='<astra-db-application-token>',
      api_endpoint='<astra-db-api-endpoint>',
      collection_name='<collection-name>',
      embedding_dimension=1536,  # Length of embedding vectors
      keyspace='<keyspace-name>',  # optional (default: 'default_keyspace')
  )
}
data_sink = client.data_sinks.create_data_sink(request=ds)

TypeScript Client

const ds = {
    'name': 'astradb',
    'sinkType': 'ASTRA_DB', 
    'component': {
        'token': '<astra-db-application-token>',
        'api_endpoint': '<astra-db-api-endpoint>',
        'collection_name': '<collection-name>',
        'embedding_dimension': 1536,  // Length of embedding vectors
        'keyspace': '<keyspace-name>',  // optional (default: 'default_keyspace')
    }
}

data_sink = await client.dataSinks.createDataSink({
  projectId: projectId,
  body: ds
})

Configuration Parameters

Parameter	Type	Required	Description
token	string	Yes	The Astra DB Application Token to use for authentication
api_endpoint	string	Yes	The Astra DB JSON API endpoint for your database
collection_name	string	Yes	Collection name to use. If not existing, it will be created
embedding_dimension	integer	Yes	Length of the embedding vectors in use (e.g., 1536 for OpenAI)
keyspace	string	No	The keyspace to use. If not provided, 'default_keyspace' will be used

Prerequisites

Before configuring AstraDB as a data sink, ensure you have:

1. AstraDB Database: A running AstraDB database instance
2. Application Token: An AstraDB Application Token with appropriate permissions
3. API Endpoint: The JSON API endpoint URL for your database
4. Keyspace: A keyspace in your database (optional, will use 'default_keyspace' if not specified)

Getting Started with AstraDB

1. Create an AstraDB Database

• Visit the AstraDB Console
• Create a new database or use an existing one
• Note down your database's API endpoint

2. Generate an Application Token

• In the AstraDB Console, navigate to your database
• Go to the "Connect" tab
• Generate an Application Token with the necessary permissions
• Save the token securely

3. Configure the Data Sink

Use the token and API endpoint in your data sink configuration as shown in the examples above.

Filter Syntax

When using AstraDB as a data sink, you can apply filters using standard MongoDB-style query operators:

Filter Operator	Description
$eq	Equals
$ne	Not equal
$gt	Greater than
$lt	Less than
$gte	Greater than or equal
$lte	Less than or equal
$in	Value is in a list
$nin	Value is not in a list

These filters can be applied to metadata fields when querying your AstraDB collection to refine search results based on specific criteria.