[GitHub] [hive] zabetak commented on pull request #2432: HIVE-24590: Operation logging still leaks log4j appenders

[GitBox <gi...@apache.org>]

zabetak commented on pull request #2432:
URL: https://github.com/apache/hive/pull/2432#issuecomment-951288032


   > Ya, nothing about this solution feels ideal. How was the default of 60s chosen? That seems pretty short to me if there's a stall in the query engine (or anywhere else).
   
   Even if there is a big delay and the 60sec window passes there is no problem. The appender will be closed and it will reopen again when the next log event arrives. Having a purge policy guarantees that there will be no leak no matter what happens.
   
   Something that may become problematic is the arrival of many (in the order of thousands) queries in the 60 sec window. This will lead to the creation of many appenders and usage of many file descriptors. However, this might never be a problem for Hive, at least not before other parts of the system need to be fixed first.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: gitbox-unsubscribe@hive.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: gitbox-unsubscribe@hive.apache.org
For additional commands, e-mail: gitbox-help@hive.apache.org