In general, when you try to serve contents off a data storage, the storage
needs to be mounted so that the apache web server itself or the module in
charge (in your case probably the php module) can find it using standard
system libraries and system calls.
HDFS itself is not a POSIX file system that can be directly mounted. There
is a fuse implementation that sort of bridges the gap, but it is provided
only for convenience, not as a something that turns HDFS into a general
purpose clustered file system.
The focus of HDFS design and development has not been on online serving, so
what you are trying probably is not the right direction. If you explain why
you are using hadoop/hdfs and what you what to achieve, I am sure you will
get good advices from the community.
To reach the relevant audience, please post your usage questions to one of
the *-user maling lists.
On 9/21/11 6:30 AM, "mariaprabudass" <[EMAIL PROTECTED]> wrote:
> Hi all,
> I have copied my files using hadoop in the following path
> "/user/hadoop/input" path. Now i need to open the "index.php" file which
> located in the above path, so i configure that path in apache. Then i tried
> in Ubuntu using vim 000-default and edit the below two lines.
> DocumentRoot /user/hadoop/input
> <Directory /user/hadoop/input>
> But my "index.php" not working because file(index.php) not viewed, Just
> throw an following error.
> The requested URL /input/index.php was not found on this server.
> Kindly check and advise me how to solve this issue.
> View this message in context:
> Sent from the Hadoop core-dev mailing list archive at Nabble.com.