> On June 7, 2015, 6:08 p.m., Jiangjie Qin wrote: > > Is it still possible that in tryCompleteElseWatch after we get the watcher > > from watchersForKey() but before we call watchers.watch(), that watcher get > > removed from purgatory? If so, is it possible that operation got lost?
Ack. > On June 7, 2015, 6:08 p.m., Jiangjie Qin wrote: > > core/src/main/scala/kafka/server/DelayedOperation.scala, lines 258-265 > > <https://reviews.apache.org/r/35187/diff/1/?file=980465#file980465line258> > > > > Can we put this entire part into tryCompleteWatched/purgeComplete? This > > code block can be simplified to: > > inWriteLock(removeWatchersLock) { > > if (watchersForKey.get(key) == this) > > watchersForKey.remove(key) > > } I feel it is better to keep it in a separate function as it is referred in two places, because when we change the logic in the future we will not forget either one of the functions. - Guozhang ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/35187/#review86972 ----------------------------------------------------------- On June 7, 2015, 4:49 a.m., Guozhang Wang wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/35187/ > ----------------------------------------------------------- > > (Updated June 7, 2015, 4:49 a.m.) > > > Review request for kafka. > > > Bugs: KAFKA-2253 > https://issues.apache.org/jira/browse/KAFKA-2253 > > > Repository: kafka > > > Description > ------- > > Double check on whether the key is still correlated to the watchers in > removeKey function > > > Diffs > ----- > > core/src/main/scala/kafka/server/DelayedOperation.scala > 123078d97a7bfe2121655c00f3b2c6af21c53015 > > Diff: https://reviews.apache.org/r/35187/diff/ > > > Testing > ------- > > > Thanks, > > Guozhang Wang > >
