Lars Volker has posted comments on this change. ( http://gerrit.cloudera.org:8080/13550 )
Change subject: WIP IMPALA-8484: Run queries on disjoint executor groups ...................................................................... Patch Set 7: (1 comment) http://gerrit.cloudera.org:8080/#/c/13550/7/be/src/scheduling/admission-controller.cc File be/src/scheduling/admission-controller.cc: http://gerrit.cloudera.org:8080/#/c/13550/7/be/src/scheduling/admission-controller.cc@496 PS7, Line 496: const int64_t max_requests = pool_cfg.max_requests; > This code looks weird in that most other code now uses GetMaxRequestsForPoo I agree that this still needs some thought. Should we use GetMaxRequestsForPool(group_size) here instead? I generally feel that admitting more concurrent queries to larger clusters might not increase the throughput beyond a certain point. To model this, we could also add a field "admit_num_limit" to TBackendDescriptor, make it default to num_cores and expose it via a flag. -- To view, visit http://gerrit.cloudera.org:8080/13550 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I8a1d0900f2a82bd2fc0a906cc094e442cffa189b Gerrit-Change-Number: 13550 Gerrit-PatchSet: 7 Gerrit-Owner: Lars Volker <l...@cloudera.com> Gerrit-Reviewer: Andrew Sherman <asher...@cloudera.com> Gerrit-Reviewer: Bikramjeet Vig <bikramjeet....@cloudera.com> Gerrit-Reviewer: Impala Public Jenkins <impala-public-jenk...@cloudera.com> Gerrit-Reviewer: Lars Volker <l...@cloudera.com> Gerrit-Reviewer: Tim Armstrong <tarmstr...@cloudera.com> Gerrit-Comment-Date: Fri, 28 Jun 2019 00:54:44 +0000 Gerrit-HasComments: Yes