Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Plain View
Hive, mail # dev - how to manage hive query logs in hive.querylog.location


Copy link to this message
-
how to manage hive query logs in hive.querylog.location
Bing Li 2012-12-03, 09:47
Hi, All
Hive keeps its structured query logs in hive.querylog.location which is set
in hive-site.xml.

There are two kinds of log files kept in hive.querylog.location,
- hive_job_log_<random unique number>.txt
Hive history file, which would be created when hive CLI launched each time,
e.g. "hive -e" invoked

- ${user_name}_<random unique number>.pipeout
Temporary files used to store results of non-Hive commands (e.g. set, dfs).
Sometimes it's empty and zero-length, is it expected?

Is there a way to manage above temp log files:
1. Do these files ever get cleaned up?
2. Is there a limit to how many of these files will be created?
3. Is there a configuration that will either limit the creation of these
files or clean them up?
Thanks,
- Bing