Users who attempt to run queries are reporting the following error message: error: Spectrum Scan Error: Access throttledWhich solution will resolve this error?
Check file sizes of fact tables in Amazon S3, and look for large files. Break up large files into smaller files of equal size between 100 MB and 1 GB
Reduce the number of queries that users can run in parallel.
Check file sizes of fact tables in Amazon S3, and look for small files. Merge the small files into larger files of at least 64 MB in size.
Review and optimize queries that submit a large aggregation step to Redshift Spectrum.
Explanations:
Large files are not the main issue. The error is related to throttling, which is often caused by having too many small files, not large files.
Throttling is typically due to excessive small file usage in Redshift Spectrum, not the number of queries. Reducing parallel queries won’t address this.
Small files cause throttling in Redshift Spectrum. Merging small files into larger ones (at least 64 MB) reduces overhead and improves performance.
Optimizing queries may help with performance, but the error is due to throttling caused by small files, not by inefficient queries.
I organize that the answer is:
Check file sizes of fact tables in Amazon S3, and look for small files. Merge the small files into larger files of at least 64 MB in size.