Networking on z/OS
|
Previous topic |
Next topic |
Contents |
Glossary |
Contact z/OS |
PDF
Control point (CP-CP) sessions Networking on z/OS |
|
|
To perform directory services and topology and route selection services, adjacent APPN nodes throughout the APPN network use the pair of CP-CP sessions to exchange network and control information. CP-CP sessions are always logical unit type 6.2 (LU 6.2) sessions. Using this session type, a contention situation could arise if both session partners attempted to allocate a conversation and exchange data at the same time. This situation is resolved by defining one of the sessions the contention-winner (often called the conwinner) session and the other the contention-loser (or conloser) session. The primary session partner refers to its session as the contention-winner session, and the secondary session partner refers to that same session as the contention-loser session. The contention winner side of the session is the one that initiates the BIND. CP-CP sessions are only established between adjacent APPN nodes and always use the CPSVCMG logon mode name and APPN class of service. End node CP-CP sessions An end node establishes CP-CP sessions with an adjacent network node. The network node that has a CP-CP session with the end node is referred as the network node server (NNS). End nodes can have active links to many network node or end nodes at the same time, but can establish a CP-CP session pair with only one network node server at a time. If a CP-CP session fails, an end node can immediately choose another adjacent network node to act is its network node server. Note the following points:
Network node CP-CP sessions To perform directory services, topology and route selection services, adjacent APPN nodes throughout the APPN network use the set of two CP-CP sessions to exchange network and control information. A network node or composite network node can establish CP-CP sessions with any network node or composite network node to which it has an APPN direct link that supports CP-CP sessions. CP-CP sessions between two network nodes are used to perform searches for resources, exchange topology information, and can be used to register resources with a central directory server. After an APPN connection has been established, identification information is exchanged between the nodes, and CP-CP sessions can be established between the control points in the directly attached nodes. After the CP-CP sessions are established, the two nodes exchange CP capabilities, which indicate the level of network services provided by the control point. NETID considerations Each APPN node has one network ID and one CP name assigned. The network ID identifies the network to which the node belongs, and the CP name is unique within that network. The network ID and CP name are defined in the APPN node at the time of system definition. Within an APPN network, all interconnected network nodes share a common network ID. CP-CP sessions are allowed only between network nodes that have the same NETID, unless one or both network nodes are defined as extended border nodes. An end node can use the same network ID as its network node server, or it can use a different network ID. |
Copyright IBM Corporation 1990, 2010 |