Alex Behm has posted comments on this change. Change subject: IMPALA-5412: Fix scan result with partitions on same file ......................................................................
Patch Set 5: (3 comments) http://gerrit.cloudera.org:8080/#/c/7625/5/be/src/exec/hdfs-scan-node-base.h File be/src/exec/hdfs-scan-node-base.h: Line 365: typedef std::unordered_map<PartitionFileKey , HdfsFileDesc*, pair_hash> FileDescMap; extra space after PartitionFileKey http://gerrit.cloudera.org:8080/#/c/7625/5/tests/metadata/test_partition_metadata.py File tests/metadata/test_partition_metadata.py: Line 52: def test_multiple_partitions_same_location(self, vector, unique_database): I think this test could use some cleanup, but I'm ok to accept this patch if you feel the cleanup is too cumbersome. * instead of limiting the file formats and relying on allow_unsupported_formats, we could use existing alltypes data (create table like, then create partitions pointing to known locations with data) * split up the read and write tests; it's good to have coverage of the write path, but none of the JIRAs mentioned here were bugs in the write path Line 103: # check if using num_nodes=1 has the same behaviour # force all scan ranges to be on the same node -- To view, visit http://gerrit.cloudera.org:8080/7625 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: Ie74b305377248045c0d87b911943e1cabb7223e9 Gerrit-PatchSet: 5 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Gabor Kaszab <[email protected]> Gerrit-Reviewer: Alex Behm <[email protected]> Gerrit-Reviewer: Attila Jeges <[email protected]> Gerrit-Reviewer: Gabor Kaszab <[email protected]> Gerrit-Reviewer: Laszlo Gaal <[email protected]> Gerrit-Reviewer: Matthew Jacobs <[email protected]> Gerrit-Reviewer: Tim Armstrong <[email protected]> Gerrit-HasComments: Yes
