IBM Communication Controller Migration Guide

Appendix C. Avoiding VTAM network addressing problems 405

Understanding VTAM network addressing

Network addresses are the architected means for identifying specific resources

in an SNA network, referred to as Network Addressable Units or NAUs. Network

addresses are assigned to SNA resources in a variety of ways (like pre-definition

or dynamic assignment), and are also used in a variety of ways. In general, most

network addresses are assigned to SNA resources that participate in LU-LU

sessions, such as SNA applications and devices. But other types of devices, like

PUs and LINEs, must also be assigned network addresses which are used for

SSCP-PU sessions or as a general means of identification.

SNA sessions

Every LU-LU session that is established must be represented by a unique pair of

network addresses: one network address that is assigned to the primary LU

(PLU) of the session and one that is assigned to the secondary LU (SLU).

Because many SNA application programs support “parallel sessions” (two or

more concurrently active sessions between the same two session partners),

assigning only a single network address to each session-capable SNA resource

is not always sufficient to guarantee network address pair uniqueness for all

sessions. Therefore, it is often necessary to assign more than one network

address to some SNA resources.

Depending on where a resource resides in the network (or attaches to the

network), network addresses may be assigned from any subarea node (typically

NCP or VTAM). However, it is the SSCP component of VTAM that manages the

assignment of network addresses during LU-LU session establishment, even if

one or both of the network addresses being assigned comes from an NCP

network address pool. To optimize the use (and reuse) of network addresses for

LU-LU sessions and to simplify the algorithms used to assign network addresses

to SNA resources, VTAM follows these general guidelines:

򐂰 Each session-capable SNA resource that is capable of being the SLU for a

session is assigned one network address that is used for every session in

which this resource acts as the SLU. For resources that support only one

active session at a time (like dependent LUs) and applications that do not

support parallel sessions, the SLU network address may also be used for

sessions in which the resource acts as the PLU, since network address pair

uniqueness is still guaranteed.

򐂰 Each session-capable SNA resource that is capable of acting as the PLU for a

session may also be assigned one or more network addresses (as needed)

that is only used for sessions in which this resource is acting as the PLU. For

SNA resources that establish parallel sessions, a new PLU network address

is assigned for each parallel session, since the SLU typically uses the same

network address for each parallel session. To conserve network addresses,