> On April 24, 2014, 12:44 a.m., Jun Rao wrote: > > core/src/main/scala/kafka/log/Log.scala, lines 275-276 > > <https://reviews.apache.org/r/20616/diff/4/?file=566280#file566280line275> > > > > We probably should use validMessages here. > > Guozhang Wang wrote: > I actually thought about this. The thing is that the bytes that gets > trimmed is not recorded anywhere. So when we reject these messages in the > re-check after de-recompression, it would be better to record the whole > message bytes just to account those trimmed bytes.
That's fine. Perhaps we can add a comment to explain this? - Jun ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/20616/#review41250 ----------------------------------------------------------- On April 24, 2014, 11:26 p.m., Guozhang Wang wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/20616/ > ----------------------------------------------------------- > > (Updated April 24, 2014, 11:26 p.m.) > > > Review request for kafka. > > > Bugs: KAFKA-1409 > https://issues.apache.org/jira/browse/KAFKA-1409 > > > Repository: kafka > > > Description > ------- > > Update comments > > > Incorporate Jun's comments round three > > > Fix KAFKA-1409: Address Jun's comments round two > > > Diffs > ----- > > core/src/main/scala/kafka/cluster/Partition.scala > c08eab0bc35aa2b0caa0dfc1987b4face4315d3b > core/src/main/scala/kafka/log/Log.scala > 46df8d99d977a3b010a9b9f4698187fa9bfb2498 > core/src/main/scala/kafka/message/MessageSet.scala > a1b5c63b4d9deaa2556427d995c41841838cf581 > core/src/main/scala/kafka/server/KafkaApis.scala > bb0359d933356b8fcc2c8df22ce6ea9cb838d6a2 > core/src/main/scala/kafka/server/KafkaRequestHandler.scala > f11f6e2c82b016f4dd3e92466ae1d2ede19feb30 > > Diff: https://reviews.apache.org/r/20616/diff/ > > > Testing > ------- > > > Thanks, > > Guozhang Wang > >
