hadoop-hdfs-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Prateek Baranwal <deepsilentwat...@gmail.com>
Subject Re: Hadoop Cluster Initialization Fail
Date Mon, 25 Mar 2013 18:47:08 GMT

Look at any of the other data nodes to obtain the directory structure of
You will need to recreate the directories like hadoop-hdfs , hive, hbase
and the likes. Make sure you chown them with the respective user:group as
shown in datanode log directory. This should allow directory access to
write down log files.

Cheers. Let me know if it woks.


On Mon, Mar 25, 2013 at 11:11 PM, rohit sarewar <rohitsarewar@gmail.com>wrote:

> Hi,
> I kind of cleared the logs at /var/logs with rm -rf. I know this was a
> stupid move. Now the cluster does not start throwing an error.:
> Problem accessing /cmf/process/920/logs. Reason:
>  Server returned HTTP response code:500 for URL
> http://xxxxxx:9000/process/920-hdfs-NAMENODE/files/logs/stdout.log
> The server declined access to the page or resource. "
> I am using CDH 4.1
> Any help would be really appreciated.
> Regards,
> Rohit

View raw message