[
https://issues.apache.org/jira/browse/STORM-211?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14012998#comment-14012998
]
ASF GitHub Bot commented on STORM-211:
--------------------------------------
Github user revans2 commented on a diff in the pull request:
https://github.com/apache/incubator-storm/pull/128#discussion_r13207661
--- Diff: external/storm-hdfs/README.md ---
@@ -0,0 +1,342 @@
+# Storm HDFS
+
--- End diff --
Fair enough, it is just that a lot of the configs are really hard to get
right, especially for security, unless you copy the configs from a gateway.
Also I could be wrong about this but in the past for us there is enough
about Hadoop and its RPC layer that is static, that you cannot talk to two
different clusters that are configured incompatibly. But a client can talk to
many different compatible clusters. Either way we can adjust the documentation
as we get more experience with users trying to do this.
> Add module for HDFS integration
> -------------------------------
>
> Key: STORM-211
> URL: https://issues.apache.org/jira/browse/STORM-211
> Project: Apache Storm (Incubating)
> Issue Type: Sub-task
> Reporter: P. Taylor Goetz
>
> Add a module with generic components (storm, trident) for interacting with
> HDFS:
> - Write to regular and sequence files
> - Core bolts, and Trident state implementation.
> - Integrate with secure (kerberos-enabled) HDFS
--
This message was sent by Atlassian JIRA
(v6.2#6252)
