Bharath Vissapragada has posted comments on this change. ( http://gerrit.cloudera.org:8080/10543 )
Change subject: IMPALA-6119: Fix issue with multiple partitions sharing same location ...................................................................... Patch Set 3: (1 comment) http://gerrit.cloudera.org:8080/#/c/10543/2/fe/src/main/java/org/apache/impala/catalog/HdfsTable.java File fe/src/main/java/org/apache/impala/catalog/HdfsTable.java: http://gerrit.cloudera.org:8080/#/c/10543/2/fe/src/main/java/org/apache/impala/catalog/HdfsTable.java@1485 PS2, Line 1485: if (partitions == null) { > I checked how the partitionIds_ member is used: The only way this set is ac Digging into the history of these files, it appears that we maintain multiple data structures for partition pruning performance (read commit message) [1]. Gabor's approach of using keySet() on partitionMap_ makes sense to me but I guess we need to be careful to not regress something else (I'm hoping keySet().get() is still an O(1) and keySet() is just a reference and is O(1) too, it appears so based on my reading of code). My opinion is that we should make these changes in a different patch so that it is easy to revert incase something else blows up. [1] https://github.com/apache/impala/commit/0348a36b49558f2d94475e97bd62281daa4a5949#diff-982e729f15d73d5ed29b1a9969fefcf4 -- To view, visit http://gerrit.cloudera.org:8080/10543 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I2a54bc8224bcefe65b83de2df58bb84629f2aa4a Gerrit-Change-Number: 10543 Gerrit-PatchSet: 3 Gerrit-Owner: Gabor Kaszab <[email protected]> Gerrit-Reviewer: Bharath Vissapragada <[email protected]> Gerrit-Reviewer: Gabor Kaszab <[email protected]> Gerrit-Reviewer: Sailesh Mukil <[email protected]> Gerrit-Reviewer: Zoltan Borok-Nagy <[email protected]> Gerrit-Comment-Date: Mon, 04 Jun 2018 20:06:15 +0000 Gerrit-HasComments: Yes
