Csaba Ringhofer has posted comments on this change. ( http://gerrit.cloudera.org:8080/21284 )
Change subject: IMPALA-12990: Fix impala-shell handling of unset rows_deleted ...................................................................... Patch Set 4: (3 comments) http://gerrit.cloudera.org:8080/#/c/21284/2/shell/impala_client.py File shell/impala_client.py: http://gerrit.cloudera.org:8080/#/c/21284/2/shell/impala_client.py@980 PS2, Line 980: self._check_hs2_rpc_status(resp.status) > This seems right given how it's used in impala_shell.py. About TestHS2FaultInjection::test_close_dml: The "return" was never reached there as the expectation is that an exception is thrown during the RPC. Updated the code as it was quite misleading. about ImpalaBeeswaxClient::close_dml: thanks for catching it, updated the beeswax case StrictHS2Client::close_dml: TDmlResult is only returned in Impala, so the strict client doesn't use it. Hive has a different way to report number of modified rows, which is not used in Impala. If interested, IMPALA-12647 has details about the Impala and Hive solution >Also the comment describing ImpalaClient::close_dml The comment I found was about row errors, not deleted rows, and for row errors this is still true. Updated the comment as it was stale and didn't mention deleted rows. http://gerrit.cloudera.org:8080/#/c/21284/2/tests/shell/test_shell_commandline.py File tests/shell/test_shell_commandline.py: http://gerrit.cloudera.org:8080/#/c/21284/2/tests/shell/test_shell_commandline.py@925 PS2, Line 925: def test_iceberg_dml_reporting(self, vector, unique_database): > Reusing the Kudu DML sequence makes this new test compact but a bit complex Changed the function to make the Kudu / Iceberg differences more explicit and added some comments. If you think that it is still too complex, I can split the function to two. http://gerrit.cloudera.org:8080/#/c/21284/2/tests/shell/test_shell_commandline.py@932 PS2, Line 932: def _test_dml_reporting(self, vector, create_sql, db, is_kudu): > Do these get run with both HS2 and Beeswax? Yes ... but as I learned the Python3 + beeswax combination is only run in exhaustive tests so the tests I ran didn't catch it. This is a bit unlucky situation, I think that we should run all protocols with Python2 and Python3 in core tests. -- To view, visit http://gerrit.cloudera.org:8080/21284 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I5812b8006b9cacf34a7a0dbbc89a486d8b454438 Gerrit-Change-Number: 21284 Gerrit-PatchSet: 4 Gerrit-Owner: Csaba Ringhofer <csringho...@cloudera.com> Gerrit-Reviewer: Csaba Ringhofer <csringho...@cloudera.com> Gerrit-Reviewer: Impala Public Jenkins <impala-public-jenk...@cloudera.com> Gerrit-Reviewer: Michael Smith <michael.sm...@cloudera.com> Gerrit-Reviewer: Peter Rozsa <pro...@cloudera.com> Gerrit-Comment-Date: Thu, 11 Apr 2024 11:16:23 +0000 Gerrit-HasComments: Yes
