HI dear all

My scenario as:

I am collecting the data from sensor devices on window base machine and
store on local disk. I have hadoop cluster on ubuntu OS. Is it possible to
use the flume (spool directory option) on window system to  push data
directly to cluster? My window machine is not part of the cluster.
If it is possible then how I can make the setup.

Many thanks

You could build Flume on Windows:
https://cwiki.apache.org/confluence/display/FLUME/Build+Flume+1.3.x+up+on+Windows

And use the tail or spooldir exec. You have to setup the path in Windows URI style.

Other way could be to setup a syslog agent or avro source on windows which will tail the file to a flume agent.

- Alex

On Dec 29, 2012, at 5:52 PM, kashif khan <[EMAIL PROTECTED]> wrote:

> HI dear all
>
> My scenario as:
>
> I am collecting the data from sensor devices on window base machine and
> store on local disk. I have hadoop cluster on ubuntu OS. Is it possible to
> use the flume (spool directory option) on window system to  push data
> directly to cluster? My window machine is not part of the cluster.
> If it is possible then how I can make the setup.
>
> Many thanks