Port 6666 is used by Databricks for secure cluster connectivity, specifically when using PrivateLink for enhanced network security.

What is PrivateLink?

PrivateLink is a technology that allows you to connect to Databricks workspaces privately through your VPC (Virtual Private Cloud) without exposing traffic to the public internet. This provides additional security and control over your data and network communication.

Why is Port 6666 Needed?

When PrivateLink is enabled, Databricks uses port 6666 to communicate securely between the cluster nodes and the Databricks control plane. This ensures that the data transfer and communication between the components remain within your private network, protected from external access.

Configuration Requirements

If you are using PrivateLink with Databricks, you need to ensure the following:

• Network ACLs (Access Control Lists): Allow inbound and outbound TCP traffic on port 6666 between the Databricks workspace subnets and any other relevant subnets or security groups.
• Security Groups: Ensure that the security groups associated with your Databricks resources allow inbound and outbound traffic on port 6666.
• Firewall Rules: If additional firewalls are in place, verify that they permit traffic on port 6666.