Coding considerations for APPN resources

LU definitions are not needed if you use dynamic independent LUs. See Defining independent LUs.

Coding examples for each of the connection types are included in this section following the general requirements. See Channel connections between APPN nodes, Leased connections between APPN nodes, IBM 3172 Nways Interconnect Controller connections between APPN nodes, and IBM Open Systems Adapter connections between APPN nodes.

If you take advantage of the CPCP and CONNTYPE start option defaults, you only need to consider the following requirements for an APPN connection:

To enable CP-CP sessions over a switched APPN connection, the CPCP start option should be defined as YES or SWITCHED.
The PU definition for the NCP in a composite network node or for the VTAM® in a network node or end node needs to indicate exchange identification support (XID=YES). (For a switched connection, XID cannot be coded but is automatically set to YES.)
Use different names for the CP name and the PU name of a node.
If you choose to allow dynamic definition of your independent logical units by defining the DYNLU start option or operand as YES, you only need to define the independent LUs on their owning CP.
If you choose not to allow dynamic definition of your independent logical units, each network node that needs to connect to the independent logical unit needs a CDRSC definition (or a model CDRSC definition) for the resource. If you chose to code the ALSLIST operand on the CDRSC statement, the ALSLIST operand should specify either the name of an adjacent link station that can be used to connect to the LU or to ISTAPNPU. ISTAPNPU is a generic representation for an APPN adjacent link station. If ISTAPNPU is included in the ALSLIST, VTAM uses any available APPN link station when calculating a route to the LU.
If you chose not to code ALSLIST, VTAM searches for the LU in the APPN network and, if found, adds ISTAPNPU to the ALSLIST dynamically.

Also, unlike a LEN adjacent link station, an APPN adjacent link station is not required to have a NETID that matches the NETID of the independent LU at session establishment.
To establish a statically defined switched APPN connection, either the CPNAME operand or the IDBLK and IDNUM operands must be defined for each switched PU definition. If the CPNAME operand is defined, its value must be the same as the CP name of the node (the SSCPNAME start option for VTAM nodes) that owns the PU. If the values are not the same, VTAM cannot locate the PU and LU definitions. There may be more than one PU related to that same CP.
To establish a statically defined switched LEN connection, either the CPNAME operand or the IDBLK and IDNUM operands must be defined for each switched PU definition. If the CPNAME operand is defined, its value specifically identifies a specific PU and cannot be duplicated.
To establish a dynamically defined switched connection, you can allow the DYNADJCP start option to default to YES and code the DYNPU operand on the GROUP definition statement in the XCA or NCP major node.
You need to know the maximum number of sessions that will be established through each NCP in a composite network node. For NCP type 2.1 connections, this number is defined during NCP generation and cannot dynamically expand. However, for a direct VTAM connection, the session control blocks are held in the BSBUF buffer pool that can expand and contract as necessary.
For an NCP connection, you should also know the maximum number of sessions that any particular type 2.1 node can establish per link station. If you do not specify this limit, a single independent LU in a type 2.1 node can exhaust all of the session control blocks generated in the NCP. For information about specifying this limit, see Limiting sessions.
There is no corresponding control for type 2.1 nodes directly attached to a VTAM. However, because this is a buffer pool, it can expand as needed to accommodate more sessions.
When an independent LU in a type 2.1 node acts as the primary logical unit, NCP must assign another network address to the PLU. This address pool is generated in NCP. Because this is a predefined pool, you should know the maximum number of PLU sessions that all type 2.1 node independent LUs attached to that NCP can establish concurrently. For a direct VTAM connection, the addresses are assigned from the element pool that is managed by VTAM.
Depending upon the specific type of APPN connection and the type of node at each endpoint, High-Performance Routing (HPR) can be used. If required, HPR support can be disabled on a node-wide basis using the HPR=NONE start option, or limited using the HPR=ANR start option. Within a node providing HPR support, HPR can be disabled on individual links by using the HPR operand on the GROUP, LINE, PU or CDRM definition statements. See Limiting HPR support for additional information about the effects of the HPR start option. For further information about HPR, see High-Performance Routing (HPR).