What is a Cluster?

• enVector Cluster is a fully managed environment where you can store and search encrypted vector data without dealing with infrastructure details.
• Every cluster has a dedicated endpoint and data space—so you can connect directly with enVector SDK and handle your requests smoothly.

Notes

• You can view your cluster details and the indexes created within them.
• Indexes are available only when the cluster is running.

How to connect (enVector SDK)

1. Install the enVector SDK version that matches the cluster version.

   pip install pyenvector=={envector_version}

2. Connect to the cluster with your API Key.

   import pyenvector as ev
   
   ev.init(
       address="<YOUR_CLUSTER_ENDPOINT>",
       access_token="<YOUR_API_KEY>",
       key_path="./path/to/sdk-keys",
       key_id="name-of-sdk-key",
   )

   • address: cluster endpoint
   • access_token: API Key value
   • key_path: path to load SDK keys for encryption/operations
   • key_id: name of the SDK keys

Features - Cluster Info

• Cluster ID: Unique identifier of the cluster.
• Created At: Cluster creation time.
• Cloud: CSP selected at the time of cluster creation.
• Cloud Region: CSP region selected the time of cluster creation.
• enVector Version: The latest available enVector SDK version. The cluster you create is compatible with this SDK version.
• Endpoint: The endpoint address to connect to this cluster.
• Cluster Status

  Running	the cluster is in a running state.
  Provisioning	the cluster is in a provisioning state.
  Starting	the cluster is restarting from a stopped state.
  Deleting	the cluster is being deleted.
  Stopping	the cluster is stopping from a running state.
  Stopped	the cluster is stopped.
  Error	an error occurred in the cluster.
  pending	the cluster is in a pending state.

Features - Indexes

• Index Name: The name assigned when the index was created.
• Created At: Index creation time.
• Index Type: The method used to store and search vector data.
• Dim: Vector dimension.
• Row Count: Total number of vector rows stored.