forked from D-Net/dnet-hadoop
Claudio Atzori 7a3f8085f7 | ||
---|---|---|
.. | ||
src | ||
README.md | ||
pom.xml |
README.md
Description of the Module
This module defines a collector worker application that runs on Hadoop.
It is responsible for harvesting metadata using different plugins.
The collector worker uses a message queue to inform the progress of the harvesting action (using a message queue for sending ONGOING messages) furthermore, It gives, at the end of the job, some information about the status of the collection i.e Number of records collected(using a message queue for sending REPORT messages).
To work the collection worker need some parameter like:
- hdfsPath: the path where storing the sequential file
- apidescriptor: the JSON encoding of the API Descriptor
- namenode: the Name Node URI
- userHDFS: the user wich create the hdfs seq file
- rabbitUser: the user to connect with RabbitMq for messaging
- rabbitPassWord: the password to connect with RabbitMq for messaging
- rabbitHost: the host of the RabbitMq server
- rabbitOngoingQueue: the name of the ongoing queue
- rabbitReportQueue: the name of the report queue
- workflowId: the identifier of the dnet Workflow
##Plugins
- OAI Plugin
Usage
TODO