zhli1142015 commented on pull request #28769:
URL: https://github.com/apache/spark/pull/28769#issuecomment-641652164
> I think this 'works', though I do wonder how many iterators are open at
once. It can't be like thousands, right?
>
> I guess we don't otherwise have a guess about what is leaking iterators?
if they were reliably closed this wouldn't be a big deal, right
for my cases (i use spark history server), caller is "AppStatusStore.scala".
Most iterators would be closed either by "closeableIterator" or GC
("finalize"). **But one thing here is if DB is already closed, iterators can
not be closed, otherwise it will rigger crash**. This is the reason i want to
close all iterators in LevelDB's close method.
level db's iterator would keep file handle, if iterator is leaked, it
actually would cause issue for me (mentioned in attached jira).
----------------------------------------------------------------
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
