You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hive.apache.org by "Thejas M Nair (JIRA)" <ji...@apache.org> on 2015/04/04 08:52:33 UTC
[jira] [Commented] (HIVE-10214) log metastore call timing
information aggregated at query level
[ https://issues.apache.org/jira/browse/HIVE-10214?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14395592#comment-14395592 ]
Thejas M Nair commented on HIVE-10214:
--------------------------------------
Timing information is printed like this -
{code}
2015-04-03 23:40:35,273 INFO [main]: metadata.Hive (Hive.java:logDumpPhase(3272)) - Dumping metastore api call timing information for : execution phase
2015-04-03 23:40:35,273 DEBUG [main]: metadata.Hive (Hive.java:dumpAndClearMetaCallTiming(3252)) - Total time spent in each metastore function (ms): {isCompatibleWith_(HiveConf, )=0, getDatabase_(String, )=3, getTables_(String, String, )=10021}
2015-04-03 23:40:35,274 INFO [main]: metadata.Hive (Hive.java:dumpAndClearMetaCallTiming(3263)) - Total time spent in this metastore function was greater than 1000ms : getTables_(String, String, )=10021
{code}
By default the aggregate timing information for metastore function calls is printed only in DEBUG mode. To enable debug mode for just the necessary class, set the following in hive-log4j.properties file -
log4j.logger.hive.ql.metadata.Hive=DEBUG,DRFA
If the total timing for any function for the query phases (compilation/execution) exceeds 1 sec, it is printed at INFO level as seen above.
> log metastore call timing information aggregated at query level
> ---------------------------------------------------------------
>
> Key: HIVE-10214
> URL: https://issues.apache.org/jira/browse/HIVE-10214
> Project: Hive
> Issue Type: Bug
> Components: Metastore
> Reporter: Thejas M Nair
> Assignee: Thejas M Nair
> Attachments: HIVE-10214.1.patch
>
>
> For troubleshooting issues, it would be useful to log timing information for metastore api calls, aggregated at a query level.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)