[
https://issues.apache.org/jira/browse/STORM-1219?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15060365#comment-15060365
]
ASF GitHub Bot commented on STORM-1219:
---------------------------------------
Github user arunmahadevan commented on the pull request:
https://github.com/apache/storm/pull/893#issuecomment-165185880
@harshach with 15 secs, the throughput was very low since the spout could
not emit more tuples after hitting max.spout.pending. The low value (1s)
default also makes sure that the value is not more than
topology.message.timeout.secs in case user changes the default timeout (if the
value is more it would cause unnecessary duplicates). There seems to be no way
to get the message timeout value in getComponentConfiguration and adjust the
tick tuple interval accordingly.
The 1s is only the default, the user can change it via HiveOptions if
compaction is a concern.
> Fix HDFS and Hive bolt flush/acking
> -----------------------------------
>
> Key: STORM-1219
> URL: https://issues.apache.org/jira/browse/STORM-1219
> Project: Apache Storm
> Issue Type: Bug
> Reporter: Arun Mahadevan
> Assignee: Arun Mahadevan
>
> HDFS and Hive bolts is setting the default tick tuple interval in the
> prepare() method, which is not taking effect. This needs to be fixed so that
> the tuples are acked on time.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
