You are viewing a plain text version of this content. The canonical link for it is here.
Posted to hdfs-dev@hadoop.apache.org by "Vinayakumar B (JIRA)" <ji...@apache.org> on 2014/09/19 18:17:33 UTC
[jira] [Created] (HDFS-7099) Support polling for more data from a
write-in-progress file using same DFSInputStream
Vinayakumar B created HDFS-7099:
-----------------------------------
Summary: Support polling for more data from a write-in-progress file using same DFSInputStream
Key: HDFS-7099
URL: https://issues.apache.org/jira/browse/HDFS-7099
Project: Hadoop HDFS
Issue Type: Improvement
Reporter: Vinayakumar B
Assignee: Vinayakumar B
Currently DFSInputStream can read only till the bytes written which were written by the time of Opening it for reading.
But, later there can be more bytes written to same file.
If client expects more data to be available, then client can poll it continuously for more data without re-opening the stream.
One simple example use case for this is tailing logs files.
Currently LocalFileSystem supports this.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)