andygrove edited a comment on pull request #8473: URL: https://github.com/apache/arrow/pull/8473#issuecomment-710090984
> A `stream` of record batches will only return the next batch _after_ the previous one was (async) computed. Therefore, we can't parallelize, as we are blocked by the execution of the previous batch. That sounds like exactly what we need. We compute partitions in parallel and each partition produces an ordered set of batches. In some cases, it is critical that the order is deterministic. The benefit of this PR is that we don't need to wait for each operator to complete before the next one starts (once we remove the use of `collect` everywhere). ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected]
