concurrently with other protocols
mainframe
what is known as a network node, usually defined as a Cisco router. The network node is a PU
2.1 in IBM terminology. Network nodes typically have control points (CPs) that communicate
with neighboring network node control points and exchange information. There are also end
nodes, which could be PU 2.1. End nodes provide end-user services and are usually located on
the outskirts of the APPN network. A third type of node is called the low entry node (LEN),
which represents dumb nodes. LENs were developed before APPN, and they need to have
resources defined to them. The fourth and final type of node is the mainframe, which is called
a composite network node.
defined.
router to router? Two routing protocols are used. The first one, developed by IBM, was
intermediate session routing (ISR). The new development is high-performance routing (HPR).
session routing (ISR). A single session is divided into stages, as shown in Figure 13-15. Each
session has a unique identifier, called the Local-Form Session Identifier (LFSID). In Figure
13-15, there is a session 123 between the Token Ring PC and the router. There is also a unique
session between the two routers, labeled 456. As a message traverses the routers, the entire
message must be buffered. Error detection, error correction, flow control, and resegmentation
might occur at each point, not at the end stations. At each router, the incoming unique LFSID
is swapped for the outgoing unique LFSID. So the only things that must be remembered are the
label and port in both directions. The major drawback of ISR is that it is very much like IP static
routing and cannot reroute around failures. In addition, much checking and verifying are done
at each stop rather than just at the connection's endpoints.