Hi Everyone, Question about hashed partition in compaction. I have datasource with multi tenant information with customer_id column. Currently we have enabled dynamic partitioning. However, if we enable customer_id as partitioning with hashed option, then the segments that are saved in s3 are bucketd with number of partitions as well. If we run MSQ with customer_id as filter, does it pull all the segment files across partition buckets or only the buckets which contain customer_id from S3 and process only this data in MSQ ?

I think MSQ partition elimination only works with multi-dim (i.e. range) partitioning, not with hash.

thanks @Will Xu