Gabor Kaszab has posted comments on this change. ( http://gerrit.cloudera.org:8080/9943 )
Change subject: IMPALA-5706: Spilling sort optimisations ...................................................................... Patch Set 14: (1 comment) http://gerrit.cloudera.org:8080/#/c/9943/8/be/src/runtime/sorter.cc File be/src/runtime/sorter.cc: http://gerrit.cloudera.org:8080/#/c/9943/8/be/src/runtime/sorter.cc@1676 PS8, Line 1676: num_required_buffers -= sorted_runs_[j]->HasVarLenPages() ? 2 : 1; > Thanks for the heads-up! I count for 2 buffers for an intermediate result t There is an issue with allocation 2 buffers unconditionally for the result run: In case there are no var len pages for any of the runs, then the # of available buffers will be 3 (as a result of ComputeMinReservation()) but if out of those two are taken by the merge result then won't be enough buffers to merge at least two runs. So I have to take it into account that check in Run::Init() when I calculate the number of buffers for the merge result (if (has_var_len_slots_)). -- To view, visit http://gerrit.cloudera.org:8080/9943 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I74857c1694802e81f1cfc765d2b4e8bc644387f9 Gerrit-Change-Number: 9943 Gerrit-PatchSet: 14 Gerrit-Owner: Gabor Kaszab <gaborkas...@cloudera.com> Gerrit-Reviewer: Csaba Ringhofer <csringho...@cloudera.com> Gerrit-Reviewer: Gabor Kaszab <gaborkas...@cloudera.com> Gerrit-Reviewer: Impala Public Jenkins <impala-public-jenk...@cloudera.com> Gerrit-Reviewer: Tim Armstrong <tarmstr...@cloudera.com> Gerrit-Comment-Date: Mon, 21 May 2018 16:56:37 +0000 Gerrit-HasComments: Yes