Distributed Name Service (DNS) Management Operations Manual

The DNS Solution
Introducing Distributed Name Service
1–4 31258 Tandem Computers Incorporated
The DNS Solution DNS is the Tandem solution to problems involved in managing object names in a
distributed network. With DNS, you and other users can:
Record and retrieve information about specific objects. For example, for a given
object, you can record the name of the subsystem to which the object is defined,
the name of the management process that controls the object, the object type, and
other information.
Assign one or more aliases to an object, so users need not refer to objects by their
subsystem-controlled names. Thereafter, you can enter an alias interactively or
use it in a program to identify the object without knowing the name of the
subsystem that controls or manages that object.
In addition, you can refer to objects by different alias names. For instance, in your
computer room, operators can refer to a line as LINE46. People who troubleshoot
line problems at your site would probably refer to the line by its telephone
company circuit number, 06KDF0056A. DNS gives you the flexibility to assign
multiple names to objects.
Define groups of related objects and refer to the collection of objects by a group
name. Given the group name, a person or program can retrieve names of the
members; given the name of a member, a person or program can find out to which
group or groups the name belongs.
Define composite objects. A composite object is one that is controlled and known
to multiple subsystems, possibly by multiple names. Given the name of the
composite object, a person or program can find out which subsystems control the
object and what name the object has in each subsystem. Given the name by which
an object is known to one subsystem (or an alias for that name), a person or
program can find out whether the object is a composite object, what other
subsystems control the object, and what those subsystems call the object.
Distribute name definitions to remote nodes, so users on other systems (nodes)
can refer to an object without having to know where it resides. This offers fast
name resolution since the name database is not centralized. Therefore, you are
not dependent on one system for all your name definitions. By defining each
object to DNS at the node where the object resides, you can reduce the effect on
DNS caused by node loss or network partitioning.
Support multiple DNS configurations. At sites where it is necessary to support
different kinds of environments (for example, research, testing, and production),
DNS can support multiple configurations simultaneously. Each organization can
create a custom DNS configuration, along with its own DNS processes and
databases.