[
https://issues.apache.org/jira/browse/HIVE-16345?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15956807#comment-15956807
]
Peter Vary commented on HIVE-16345:
-----------------------------------
About the test results:
- HIVE-15776 - Flaky test: TestMiniLlapLocalCliDriver vector_if_expr
- org.apache.hive.jdbc.TestJdbcDriver2.testResultSetMetaData failing for a
while now
- TestTextFileHCatStorer errors are not reproducible locally
[~ychena], [~kgyrtkirk]: Could you please review?
[~aihuaxu], [~stakiar]: Your two below patches were reviewed, and tested
parallel, and combined they cause TestBeeLineDriver fails, which are handled
here:
- HIVE-16297: Improving hive logging configuration variables
- HIVE-16308: PreExecutePrinter and PostExecutePrinter should log to INFO level
instead of ERROR
[~sershe], [~vikram.dixit]: I have changed the IOContext key and you were the
ones who modified this part of the code at the last time. Could you please
review, at least this small change?
{code:title=Utilities.java}
public static void setMapRedWork(Configuration conf, MapredWork w, Path
hiveScratchDir) {
String useName = conf.get(INPUT_NAME);
if (useName == null) {
useName = "mapreduce:" + hiveScratchDir;
}
conf.set(INPUT_NAME, useName);
setMapWork(conf, w.getMapWork(), hiveScratchDir, true);
if (w.getReduceWork() != null) {
conf.set(INPUT_NAME, useName);
setReduceWork(conf, w.getReduceWork(), hiveScratchDir, true);
}
}
{code}
Thanks,
Peter
> BeeLineDriver should be able to run qtest files which are using default
> database tables
> ---------------------------------------------------------------------------------------
>
> Key: HIVE-16345
> URL: https://issues.apache.org/jira/browse/HIVE-16345
> Project: Hive
> Issue Type: Improvement
> Components: Testing Infrastructure
> Reporter: Peter Vary
> Assignee: Peter Vary
> Attachments: HIVE-16345.2.patch, HIVE-16345.patch
>
>
> It would be good to be able to run the default clientpositive tests.
> Currently we can not do that, since we start with a specific database. We
> should filter the query input and replace the table references
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
