z/OS Communications Server: SNA Network Implementation Guide
Previous topic | Next topic | Contents | Contact z/OS | Library | PDF


Controlling network data flow using pacing

z/OS Communications Server: SNA Network Implementation Guide
SC27-3672-01

Pacing is a means of controlling the flow of messages in the network to avoid congestion. Data congestion results whenever the rate at which data is going into a network exceeds the capacity of the network. Response times might lengthen and throughput might decrease. Severe or prolonged congestion in one part of a network can affect the other parts, causing overall network efficiency to suffer.

VTAM® monitors traffic and limits congestion in the network with pacing. Pacing controls network flow by limiting the amount of data a transmitter can send before receiving an acknowledgment from the receiver. For example, if the agreed amount of data is three units, after sending three units of data the transmitter must wait for the receiver to send an acknowledgment before sending any more data.

Table 1 lists the two levels of pacing that are involved in the flow-control process.
Table 1. Pacing types
Pacing type Description
Session Session pacing prevents overrun of the buffers of the LUs at the session endpoints. There are two types of session pacing: fixed and adaptive.

See Virtual route pacing for more information.

Virtual Route Virtual route pacing is used by subarea nodes to avoid buffer congestion at the virtual route endpoints.

See Virtual route pacing for more information.

Both pacing mechanisms allow pacing to occur inbound and outbound independently. Pacing values do not have to be the same for the two directions. The term inbound pacing is associated with the pacing window for message units being received by a node. Outbound pacing is associated with the pacing window for messages being sent by the node.

Go to the previous page Go to the next page




Copyright IBM Corporation 1990, 2014