Thanks Ted for the pointer. A follow-up question. Will the region server always write data to its local hdfs? I am seeing logs that said the region server is trying to get data from another data nodes. Under what scenario will a region server needs to get data from a non-local data node?
thanks On Fri, Sep 26, 2014 at 7:21 AM, Ted Yu <yuzhih...@gmail.com> wrote: > Region server should be installed on server which is a data node. > > Region server count may be lower than data node count. So some data nodes > would not have region server running. > > See http://hbase.apache.org/book.html#regions.arch.locality > > Cheers > > On Fri, Sep 26, 2014 at 7:02 AM, Thomas Kwan <thomas.k...@manage.com> wrote: > >> Hi there, >> >> To get good read performance, should the region server be installed on >> every data node? (Thinking about data locality here) >> >> thanks >> thomas >>
