Dan Hecht has posted comments on this change. ( http://gerrit.cloudera.org:8080/8707 )
Change subject: IMPALA-4835: Part 2: Allocate scan range buffers upfront ...................................................................... Patch Set 24: (1 comment) http://gerrit.cloudera.org:8080/#/c/8707/24/be/src/runtime/io/disk-io-mgr.h File be/src/runtime/io/disk-io-mgr.h: http://gerrit.cloudera.org:8080/#/c/8707/24/be/src/runtime/io/disk-io-mgr.h@308 PS24, Line 308: int64_t max_bytes); > To guarantee progress, the scanner definitely has to have enough reservatio I'm not arguing for one way or the other at this point. I'm just trying to understand the change and why various parts are the way they are. Regarding about still have the same logic for block/unblock ranges: Why is that? Doesn't the existing code already limit each range to 3 buffers? i.e. why isn't it sufficient to just say that each scanner's reservation is 3 * buffer_size? -- To view, visit http://gerrit.cloudera.org:8080/8707 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Ia243bf8b62feeea602b122e0503ea4ba7d3ee70f Gerrit-Change-Number: 8707 Gerrit-PatchSet: 24 Gerrit-Owner: Tim Armstrong <tarmstr...@cloudera.com> Gerrit-Reviewer: Bikramjeet Vig <bikramjeet....@cloudera.com> Gerrit-Reviewer: Dan Hecht <dhe...@cloudera.com> Gerrit-Reviewer: Michael Ho <k...@cloudera.com> Gerrit-Reviewer: Tianyi Wang <tw...@cloudera.com> Gerrit-Reviewer: Tim Armstrong <tarmstr...@cloudera.com> Gerrit-Comment-Date: Tue, 13 Feb 2018 21:51:52 +0000 Gerrit-HasComments: Yes
