[ https://issues.apache.org/jira/browse/FLINK-22170?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17324665#comment-17324665 ]
Shengkai Fang commented on FLINK-22170: --------------------------------------- Thanks for [~lirui] detailed guidance. I have verified with the at least one query for each test method. I meet three problems now: - For hive 2.x, it doesn't support table with PK. But the exception message just is {{NOT NULL constraints not supported until 3.0.0}}. I think we can just teil users don't support PK constraints is much better. - In streaming mode, some queries doesn't work. I have opened a ticket about this. Plerase take a look at FLINK-22334 - The module order may influence the loading of the function, e.g. ARRAY. > Manual test for hive dialect > ---------------------------- > > Key: FLINK-22170 > URL: https://issues.apache.org/jira/browse/FLINK-22170 > Project: Flink > Issue Type: Sub-task > Components: Connectors / Hive > Reporter: Rui Li > Assignee: Shengkai Fang > Priority: Major > Fix For: 1.13.0 > > > This is to test the HiveParser introduced by > [FLIP-152|https://cwiki.apache.org/confluence/display/FLINK/FLIP-152%3A+Hive+Query+Syntax+Compatibility] > The test can be conducted as follow: > * Setup Flink and Hive. Add {{flink-sql-connector-hive}} and hadoop jars to > classpath. > * Start SQL client and run test cases in {{HiveDialectITCase}} and > {{HiveDialectQueryITCase}}. Choose at least one test case from each test > method. Query test can be found in [this > folder|https://github.com/apache/flink/tree/master/flink-connectors/flink-connector-hive/src/test/resources/query-test]. > * Test switching between hive and default dialects. > * Make sure to place HiveModule first in the module list when testing DQL and > DML with hive dialect. > * Perform the test with at least hive 1.x and 2.x. -- This message was sent by Atlassian Jira (v8.3.4#803005)