gene-bordegaray commented on PR #21931: URL: https://github.com/apache/datafusion/pull/21931#issuecomment-4364552279
I like this direction as a default, seems less brittle and cleaner baseline. This approach will definitely provide some value for `preserved_file_partitions` (#21207 ) š Iād still like to discuss the possibility of partition-aware filters path when DataFusion can prove both sides share the same partition mapping and as discussed in #21207 the proper infrastructure for expressing more types of partitioning is in place. For use cases we are seeing the partition-specific filters selectivity is quite nice. I imagine that if we provided information from partitioning that two partitioning spaces (build and probe sides) are compatible (such as the same range partitioning) we would not only choose from `IN (SET)` or `multi_hash_lookup`, but a partition routed lookup instead with the global as a fallback. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
