Hi there, you probably want to see this.. http://hbase.apache.org/book.html#splitter
... as well as this... http://hbase.apache.org/book.html#regions.arch.locality ... as the latter describes data locality. On 4/4/12 7:41 AM, "sdnetwork" <sdnetw...@gmail.com> wrote: > >Hello, > >I started working with hadoop / HBase and I have a question about the >distribution of map / reduce on a htable through the different nodes of >the >cluster. > >If I understand the map is subdivided by region (TableInputFormat) and >each >map are executed on the node taht containing the region. > >But a row is always stored on a single region so if I implements a custom >org.apache.hadoop.mapreduce.InputFormat that split a row and one column >family in parameter, the job will be executed on a single node regardless >the number of column qualifier? > >if this is true i must change my data schema. or maybe i can manually >distribute the job through the cluster. > >I can not find documentation that clearly explains how the map are >distributed across the cluster. >maybe somebody have it ? > >thanks in advance. >-- >View this message in context: >http://old.nabble.com/hbase-map-reduce-questions-tp33554779p33554779.html >Sent from the HBase User mailing list archive at Nabble.com. > >