Faunus Graph Configuration

This section describes the various configuration parameters in Faunus. Each InputFormat/OutputFormat as well as the numerous MapReduce steps have specific configurations that are described in their respective sections of the documentation. Note that all configurations go into the global Configuration that is used throughout Hadoop. Therefore, Hadoop specific configurations can be provided in the Faunus configuration files. These configurations can be retrieved via FaunusGraph.getConf() or FaunusGraph.getConf(String prefix). Finally, they can be set via FaunusGraph.getConf().set(...).

Input Configuration

Option	Description	Value	Default
faunus.graph.input.format	the format of the input graph data	class	none
faunus.input.location	the location to read the graph data if in HDFS	string	none
faunus.graph.input.vertex-query-filter	the vertex-query filter to use on input graph	string	none
faunus.graph.input.edge-copy.direction	copy direction edges to opposite direction	direction	none

Output Configuration

Option	Description	Value	Default
faunus.graph.output.format	the format of the output graph data	class	none
faunus.sideeffect.output.format	the format of the output side-effect data	class	none
faunus.output.location	the location to write of the graph and side-effect data	string	none
faunus.output.location.overwrite	whether to override the output location data	boolean	false

Global Configuration

Option	Description	Value	Default
faunus.pipeline.map-spill-over	The size of in-memory mapper combiners before writing to output	int	500
mapred.jar	The job Java jar — native faunus-x-job.jar or developer jar w/ Faunus dependencies	string	faunus-x-job.jar