[
https://issues.apache.org/jira/browse/STORM-1015?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14910110#comment-14910110
]
ASF GitHub Bot commented on STORM-1015:
---------------------------------------
Github user choang commented on the pull request:
https://github.com/apache/storm/pull/705#issuecomment-143655914
I recommend making your abstraction at the state store, so you would have:
```
public interface StateStore {
public void write(Partition p, long offset);
public long read(Partition p);
}
public class ZkStateStore implements StateStore {
...
}
public class KafkaStateStore implements StateStore {
...
}
/**
Some topologies read from LATEST after a restart, so only memory state is
needed.
*/
public class MemoryStateStore implements StateStore {
...
}
```
You would not need different PartitionStateManager for different stores.
You would just have:
```
public class PartitionStateManager {
public PartitionStateManager (..., StateStore store) { ... }
public void writeState(...) {
store.write(...);
}
}
```
> Store Kafka offsets with Kafka's consumer offset management api
> ---------------------------------------------------------------
>
> Key: STORM-1015
> URL: https://issues.apache.org/jira/browse/STORM-1015
> Project: Apache Storm
> Issue Type: Improvement
> Components: storm-kafka
> Affects Versions: 0.11.0
> Reporter: Hang Sun
> Priority: Minor
> Labels: consumer, kafka, offset
> Original Estimate: 72h
> Remaining Estimate: 72h
>
> Current Kafka spout stores the offsets (and some other states) inside ZK with
> its proprietary format. This does not work well with other Kafka offset
> monitoring tools such as Burrow, KafkaOffsetMonitor etc. In addition, the
> performance does not scale well compared with offsets managed by Kafka's
> built-in offset management api. I have added a new option for Kafka to store
> the same data using Kafka's built-in offset management capability. The change
> is completely backward compatible with the current ZK storage option. The
> feature can be turned on by a single configuration option. Hope this will
> help people who wants to explore the option of using Kafka's built-in offset
> management api.
> References:
> https://cwiki.apache.org/confluence/display/KAFKA/Committing+and+fetching+consumer+offsets+in+Kafka
> https://cwiki.apache.org/confluence/display/KAFKA/A+Guide+To+The+Kafka+Protocol#AGuideToTheKafkaProtocol-OffsetCommit/FetchAPI
> -thanks
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
