[ http://issues.apache.org/jira/browse/HADOOP-250?page=all ]
Devaraj Das updated HADOOP-250:
-------------------------------
Attachment: http_browsing_dfs.patch
The problems here were mostly setup related (admit I should have run the unit
tests before submitting the patch!).
Since I start a StatusHttpServer in FSNameSystem class, the "build" directory
should be included in the classpath (so webapps resource can be found) and
hadoop.log.dir should be set, otherwise, StatusHttpServer cannot be constructed
properly [updated patch patches build.xml file with these settings].
Also, the stop method of StatusHttpServer should be called every time the
namenode is stopped, otherwise a subsequent test trying to create a NameNode
object may fail since StatusHttpServer fails to open the port.
Another thing that causes some failures to happen (once in a while) is the
sleep/timeout values in MiniDFSCluster.java. A NameNode is started and after
sleeping for 1 sec, a datanode is started. But if the namenode hasn't fully
come up by then, then RPCs will fail (the ipc.client.timeout of 1 sec seems to
be too small). So I changed all 1000 msec to 2000 msec in MiniDFSCluster.java
and the tests seem to be passing consistently after that. Updated patch is
attached.
> HTTP Browsing interface for DFS Health/Status
> ---------------------------------------------
>
> Key: HADOOP-250
> URL: http://issues.apache.org/jira/browse/HADOOP-250
> Project: Hadoop
> Type: New Feature
> Components: dfs
> Versions: 0.3.0
> Reporter: Devaraj Das
> Fix For: 0.4.0
> Attachments: dfshealth.pdf, http_browsing_dfs.new.patch,
> http_browsing_dfs.patch, http_browsing_dfs.patch, http_browsing_dfs.patch
>
> A web interface to view the DFS health/status (name and data nodes) is to be
> created. User can connect to the webserver on the namenode and a web page
> will be displayed. The web page will give some details about that namenode
> (startup time and the total cluster capacity). The web page will contain a
> table of 'live' and 'dead' datanodes. Each live datanode will be a link to
> the complete details of the datanode as given by DatanodeInfo (also see
> DataNodeReport).
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see:
http://www.atlassian.com/software/jira
