You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Matthias J. Sax (JIRA)" <ji...@apache.org> on 2015/09/03 10:49:45 UTC

[jira] [Commented] (FLINK-2606) Extend StormFileSpout to handle HDFS

    [ https://issues.apache.org/jira/browse/FLINK-2606?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14728711#comment-14728711 ] 

Matthias J. Sax commented on FLINK-2606:
----------------------------------------

It just came to my mind, that there exists a storm-hdfs module already: https://github.com/apache/storm/tree/master/external/storm-hdfs

Thus, I am not sure if Flink should provide an own HdfsSpout. It's not the purpose of the compatibility layer to provide spout/bolt implementations with all kind of features. The currently included spouts/bolt are actually provided for testing and showcase purpose only.

Maybe we should just close this ticket?

> Extend StormFileSpout to handle HDFS
> ------------------------------------
>
>                 Key: FLINK-2606
>                 URL: https://issues.apache.org/jira/browse/FLINK-2606
>             Project: Flink
>          Issue Type: New Feature
>          Components: Storm Compatibility
>            Reporter: Matthias J. Sax
>            Priority: Minor
>
> Currently, `StormFileSpout` uses a simple `FileReader` internally to read local files. Thus, it cannot read data from HDFS.
> With this ticket, either `StormFileSpout` should be extended to be able to handle HDFS files or a new class (eg, `StormHdfsSpout`) should be introduced for it.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)