Databricks Lakeflow Connector for PostgreSQL on GCP Cloud

[David_M]

Lakeflow connection for PosgresHi all,

I hope this message finds you well.

I am currently trying to create a Lakeflow connection in Databricks for a PostgreSQL database hosted on Google Cloud Platform (GCP). However, when testing the connection, I am encountering a connection failure error.

If anyone has experience or insights on connecting Databricks Lakeflow to PostgreSQL on GCP, I would greatly appreciate your guidance on troubleshooting this issue.

[Isi]

Hello @David_M 

To better support you we’d need to clarify a few points:

1. PostgreSQL location

   Is this PostgreSQL deployed inside a private VPC in GCP or is it exposed through a public IP accessible from the internet? This is key to understand what type of connectivity is possible. Databricks Docs networking  Google Docs 

2. Databricks cluster/compute type

   • If the Postgres is in a private VPC, and you are trying to connect from a serverless SQL Warehouse, you likely won’t have direct access since serverless runs in the Databricks-managed cloud outside your network.

   • In that case, you should try using a PRO/Classic cluster, which is deployed inside your own VPC, though we would also need to consider any limitations of the service with your current resource setup.

3. Firewall rules in GCP

   Have you opened the Postgres firewall to allow connections from Databricks egress IPs? It’s important to confirm whether the worker nodes (VMs) of your cluster can reach port 5432.

With these details we can better advise whether you should enable connectivity via public IP, configure firewall rules, or set up private access between clouds.

Could you share a bit more context on how you have it configured?

Thanks 🙂

Isi