Connection timed out - using sqlalchemy to access AWS usaspending data

Question

I created an instance of the usaspending.gov database in my AWS RDS. A description of this database can be found here: https://aws.amazon.com/public-datasets/usaspending/

The data are available as a PostgreSQL snapshot, and I would like to access the database using Python's sqlalchemy package within a Jupyter notebook within Amazon SageMaker.

I tried to set up my database connection with the code below, but I'm getting a Connection timed out error. I'm pretty new to AWS and Sagemaker, so maybe I messed up my sqlalchemy engine? I think my VPC security settings are OK (it looks like they accept inbound and outbound requests).

Any ideas what I could be missing?

engine = create_engine('postgresql://root:password@[my endpoint]/[DB instance]

Answer 1

(I'm a member of the SageMaker team.)

Thanks for using Amazon SageMaker!

Can you check the VPC settings on your side? It might be related to an issue another AWS customer saw for Redshift, https://forums.aws.amazon.com/thread.jspa?threadID=270111&tstart=0 . The issue might be that DNS not being resolved to private IP address of the RDS instance. If following the forum post https://forums.aws.amazon.com/thread.jspa?threadID=270111&tstart=0 , your issue doesn't get resolved, you can start a forum thread with AWS SageMaker. We would help you debug the issue.