jhorstmann commented on issue #1570: URL: https://github.com/apache/arrow-datafusion/issues/1570#issuecomment-1018351513
For hash aggregation I quite like the implementation described in https://dl.acm.org/doi/pdf/10.1145/2588555.2610507 (4.4). The algorithm there optimizes for cache locality by pre-aggregating into fixed-size thread-local hash tables and then overflowing into main memory partitions, but I think it should be easy to extend with also flushing these partitions to disk if they get too large. -- 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]
