Github user liurenjie1024 commented on the issue:
https://github.com/apache/storm/pull/1825
I think the behavior of consumer's automatic partition management is
complicated and may lead to unpredictable behavior, which is quite important
for many purposes like debugging, performance tuning, etc, especially in
critical systems like billing system in our case. In fact, even the author of
kafka consumer suggests us to use manual partition management when we use
streaming process frameworks (please refer to the Manual Partition Assignment
part of [kafka consumer doc
(http://kafka.apache.org/0101/javadoc/index.html?org/apache/kafka/clients/consumer/KafkaConsumer.html).
Other streaming process frameworks like flink, apex, spark streaming, old
storm kafka client also uses manual partition management.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
