[
https://issues.apache.org/jira/browse/FLUME-2061?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13675331#comment-13675331
]
Ambud Sharma commented on FLUME-2061:
-------------------------------------
timestamping is due to a feature in Flume called file roll/rolling and can be
adjusted using the hdfs.rollInterval configuration in your conf.
However I am not certain as to why you would want to remove it and create one
large file that stores all logs; such a file would be really hard to index or
run map/reduce on even if you create splits.
> HDFS Sink File Naming
> ----------------------
>
> Key: FLUME-2061
> URL: https://issues.apache.org/jira/browse/FLUME-2061
> Project: Flume
> Issue Type: Task
> Components: Sinks+Sources
> Environment: CDH4 Cluster
> Reporter: kasa v varun
> Labels: HDFS, flume, sink
>
> While using HDFS location as a sink and the files get transfered to that sink
> location, the name of the final file which gets stored in HDFS would be
> something like <HDFSFilePrefix>-<timestamp>.
> Is there any way in which we can give custom file names to the final files
> getting stored in HDFS?
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
