Aborted query notification in Redshift about aws-lambda-redshift-loader HOT 4 OPEN

Yes, if you reprocess the batch it will use the configuration entry that you have set when the reprocess occurs, which if unchanged would load both clusters. You could create a single cluster configuration that you link to the prefix and change the old one if you want to change the load target.

from aws-lambda-redshift-loader.

The s3Prefix is a primary key in config table, so it won't let me to create two separate load configurations, with different Redshift clusters, but for the same s3 prefix.
I had to trick the config a bit and created two separate entries with the single Redshift cluster config for each; one prod and one dr. I hope it is going to work:
1. bucket/source/schema/table_name/year=*/month=*/day=*/hour=*
2. bucket/source/schema/table_name/year=*/month=*/day=*/*

from aws-lambda-redshift-loader.

I think in that case it will potentially use both, or may just select the first one. I would instead change the prefix to a dummy value for the single cluster, and then flip it back to the dual cluster configuration once the reprocess is done.

from aws-lambda-redshift-loader.

I used your suggestion, Ian. The approach that I described above did not work, and as you said, it picked up the first matching prefix configuration.

from aws-lambda-redshift-loader.