Skip to content

Configuring custom Privacera DataServer port in Databricks Clusters with FGAC

When you are using Boto3 from Databricks Cluster with FGAC, Privacera enables Databricks' cluster security features which restricts non-standard outbound connections. If you are using a custom Privacera DataServer port (other than 443), you need to configure the Databricks cluster to allow outgoing connections on that port. This is done by modifying the Databricks init script to include the necessary iptables configuration.

Setup

  • Modify the Databricks init script:
    • Add the necessary iptables configuration to allow outgoing connections on the specified port.
    • Example: To enable boto3 access on port 8282, run the following command:
      Bash
      1
      sudo iptables -I OUTPUT 1 -p tcp -m tcp --dport 8282 -j ACCEPT
  • Restart the Databricks cluster to apply the changes.

The connector does not configure a DataServer in FGAC mode, but Boto3 calls from a notebook still go to the DataServer endpoint, and the connector's init script turns on the same Python process isolation that blocks non-standard outbound ports. So a custom DataServer port needs the same iptables rule:

  1. Add the rule to a separate init script, for example allow_dataserver_port.sh, and attach it after the connector's init script. Do not edit the connector's script - downloading the bundle again replaces it.
    Bash
    1
    sudo iptables -I OUTPUT 1 -p tcp -m tcp --dport 8282 -j ACCEPT
  2. Restart the Databricks cluster to apply the changes.