I know others have had this same problem[1], and now I have as well. I have tried all the suggested troubleshooting techniques on that question. To summarize:
- This is a new Firehose to Redshift
- S3 objects are predictably appearing with 100% success in CloudWatch
- Redshift delivery is showing as 0% success, so it must be trying
- I'm seeing that Firehose is making connections to Redshift so the firewall rules must be correct
- I am using JSON formatted entries with an external column mapping file.
- The Firehose and Redshift cluster are in the us-west-2 region, but the bucket is in US Standard (us-east-1) so I'm using the WITH REGION option.
- Following in the path of others, I have tried deleting and recreating the firehose to no avail.
- I also tried doing the COPY from the redshift cluster manually and found that it worked perfectly.
- There doesn't appear to be anything in the Redshift error tables, nor in the errors section of the bucket.
I'm about to give up on this. Anyone have suggestions for where to find the error logs before I admit failure?