back to FAQ2a. about
defining data-subsetting or partitioning rules
Q: how does Data Safety
Corporation define 'data sub-setting' or 'partitioning rules'
and why are they necessary in a distributed database system?
A: each database in a distributed
system contains some defined portion of the total data in the
distributed system. The server database contains all of the data
that is in the distributed system. Remote clients each contain
some portion of the data the server contains. This portion of
the data that each client database contains is referred to as a
sub-set. The act of defining sub-sets is also referred to as
defining "partitioning rules" for the database, or sometimes
simply "partitioning".
Most non-distributed databases have not been designed to provide
the partitioning rules as part of the schema of the database.
Since all data resides in a single database, there is no need to
specify partitioning rules. Each user typically works only with
the data they are interested in. Sometimes the application will
be able to filter information based on the current user, and
this provides a form of partitioning for the data. For
distributed databases, partitioning rules are needed to limit
the amount of data that needs to be transmitted to the remote
user, due to transmission cost and time issues. Additionally,
partitioning rules provide a secondary data security benefit, in
the event of loss or theft, because only a small portion of the
corporate database will be compromised
