COPY into Redshift from S3 access point

Question

I am attempting to load S3 data info Redshift using an S3 access point (as opposed to a bucket). When I perform the COPY command, I receive an invalid bucket error. It works fine to load from a bucket directly, but when I use an access point ARN as a bucket, I get the error. I'm guessing that it's simply not supported, but hopefully there's something I can do.

Answer 1

I suspect that this probably won't work.

My reasoning is that I know that Amazon Redshift can load data from Amazon S3 even when the Redshift cluster is in a private subnet and there is no NAT server. Thus, Redshift has its "own connection" to S3 in the backplane, rather than going through the VPC.

Since the S3 Access Point exists only in a VPC, Redshift would not be able to use the Access Point.

(I look forward to being corrected if anyone knows better!)

Answer 2

It looks like at the current time, access points are not supported for copying data into Redshift.