Namespace support makes many tasks easy - security, quota, etc Suppose your table is a.tsv, it would be stored under default namespace. On hdfs, you would see something like:
/apps/hbase/data/data/default/a.tsv When you put the table under namespace a (a:tsv), the layout would be: /apps/hbase/data/data/a/tsv Please also read namespace related content under: http://hbase.apache.org/book.html#hbase.accesscontrol.configuration Thanks On Tue, Jan 17, 2017 at 7:24 AM, Henning Blohm <[email protected]> wrote: > Hi, > > I am considering to run the same application for two different tenants on > the same HBase cluster. Instead of changing table names, I would prefer to > make use of the namespace feature. Ideally, I would be able to only change > client connectivity configuration (*-site.xml properties) and leave table > names untouched as those are referenced in code. > > From what I read in the documentation it seems however that the only way > to reference a table in a non-default namespace is to prepend its name with > a colon separated namespace qualifier. > > Is that true? > > If it is, and tenants are originally meant to be used for multi-tenancy, > what makes them more useful than any other table name prefix? > > Thanks, > > Henning > > >
