Ngone51 edited a comment on issue #24467: [SPARK-27568][CORE] Fix readLock leak 
while calling take()/first() on a cached rdd
URL: https://github.com/apache/spark/pull/24467#issuecomment-490136001
 
 
   > For this particular case, I think using a task completion listener is good 
enough?
   
   Block level read/write lock mechanism has a basic assumption that *all block 
locks should be released when a task finished*. And that's why we check the 
leaked locks after the task finished. As task completion listener also would be 
triggered after task finished, so I think using it may not take big difference.
   
   Actually, *The process of checking leaked locks*(by calling 
`releaseAllLocksForTask(taskId)`) is just like a `close` operation of database 
iterator, which releases all unlocked blocks for the task. So, I agree with 
@jiangxb1987 and prefer not to fix it if we could not have a better way against 
to this hacky way by now.

----------------------------------------------------------------
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.
 
For queries about this service, please contact Infrastructure at:
[email protected]


With regards,
Apache Git Services

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to