There are three different ways to provide input for a job:
Distributed File System (DFS): Specify a path to a preexisting file on the DFS that the WebMapReduce server is associated with. This option is usually used for large datasets.
Upload: Upload data from the local machine to use as job input. This option should be used for small- to medium-sized input.
Direct Input: Input text directly into the form to use as job input. This is the quickest and easiest way to provide small amounts of data for jobs, and is well-suited for test jobs.