My favorites | Sign in
Project Home Downloads Wiki Issues Source
READ-ONLY: This project has been archived. For more information see this post.
Search
for
ProgrammingGuideConfigurationList  
SSS Mapreduce Programming Guide - Configuration List
en, ja
Updated Feb 27, 2013 by tatsuhik...@gmail.com

Configuration list

The settings refered by the system of SSS Mapreduce are shown below.

|| explanation
name type default value
job.force_skip_mode boolean false When this is true, combiner is not used.
job.new_group.keyformat "userkey", "thread" or "order" "thread" key format of new TupleGroup
job.new_group.numslot int 16 number of slot of new TupleGroup
job.output_log boolean true Whether JobEngine print job information to the job information file or not.
sidedata.default.type "ondemand" or "inadvance" "inadvance" Default side data type
count.input.size boolean true Whether Mapper/Reducer counts input tuple size.
count.output.size boolean true Whether Mapper/Reducer counts output tuple size.
client.jobinfo.output boolean true Whether SssClient writes job information or not.
client.jobinfo.filename String "sssjob-{RANDOM_UUID}.log" File name of job information
client.jobinfo.append boolean true Job information file is opend with append-mode or not.
data_getter.cache.nbytes long 16384 DataGetter cache size in each thread.
data_getter.queue.nbytes long 262144 DataGetter queue size to communicate between user thread and reading threads.
put.output_queue.local.limit_nbytes long 131072 DataPutter thread local buffer size
put.output_queue.limit_nbytes long 2097152 DataPutter buffer size.
heartbeat.period long 60 The interval at which client sends heartbeat to server.(seconds)
marge_reducer.handle_one_side_only boolean false MergeReducers handle only tuples with the key that exists in the both two input tuple groups. This is set to 'true', Merge reducers handle all tuples in two input tuple groups.
map.output.keyformat "userkey", "thread" or "order" the value of job.new_group.keyformat The key format of new TupleGroup of map output created by SimpleJob.
reduce.output.keyformat "userkey", "thread" or "order" the value of job.new_group.keyformat The key format of new TupleGroup of reduce output created by SimpleJob.
put.output.keyformat "userkey", "thread" or "order" the value of job.new_group.keyformat The key format of new TupleGroup created by DataPutter.

Powered by Google Project Hosting