Boaz Ben-Zvi
Boaz Ben-Zvi
@weijietong can you add a detailed description of your work - the hash functions are a sensitive part of the execution, and possibly more people would like to look into...
Adding the hash32() method to the ValueVector is useful; however picking up algorithms just based on a paper or being famous may not be good enough. At my previous employer...
@weijietong - how about postponing this work to 1.17.0 ? It is a change in a critical part of the engine, and we need to get a better understanding of...
DRILL-6845: Semi-Hash-Join to skip incoming build duplicates, automatically stop skipping if too few
Commit added ( 6fb890c ) - The number of spilled partitions is taken from partitionStatSet (not passed via `initialize()`), and the size of the probe side is also considered when...
DRILL-6845: Semi-Hash-Join to skip incoming build duplicates, automatically stop skipping if too few
Seems that this PR was closed by mistake; re-opening now. @ilooner - do you have more comments or suggestions ? we are trying to finish and commit this work soon....