Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Synchronization of Processors in Distributed Processing Networks

IP.com Disclosure Number: IPCOM000046688D
Original Publication Date: 1983-Aug-01
Included in the Prior Art Database: 2005-Feb-07
Document File: 2 page(s) / 51K

Publishing Venue

IBM

Related People

Lanier, CS: AUTHOR [+3]

Abstract

When data processing stations are linked in a multi-accessible network structure (via rings, loops, etc.), in a peer-to-peer mode, controls are required to regulate data transmissions in order to avoid inefficient use of the available bandwidth. For instance, assume that a station desiring to transmit data (origin station) begins to send data before the intended destination station becomes ready to receive the data. The communications adaptation equipment at the destination station, in any practical environment, will have only a small amount of buffer storage capacity.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Synchronization of Processors in Distributed Processing Networks

When data processing stations are linked in a multi-accessible network structure (via rings, loops, etc.), in a peer-to-peer mode, controls are required to regulate data transmissions in order to avoid inefficient use of the available bandwidth. For instance, assume that a station desiring to transmit data (origin station) begins to send data before the intended destination station becomes ready to receive the data. The communications adaptation equipment at the destination station, in any practical environment, will have only a small amount of buffer storage capacity. Hence, if the processing software at the destination station has not readied a read (input) command prior to reception of the data, the data cannot be stored, and such transmissions must be rejected, requiring the origin station and its software to repeat these transmissions. In one possible scenario, an origin station repeatedly attempts to send a first packet or frame of data, and the destination station repeatedly returns refusal indications because its receiving sub-channel (host memory) is unprepared for reception of the data. This results in inefficient usage of the shared communication link medium. A more efficient way of handling this is for the receiving station at the link adaptation to remember the hardware level that it has signaled a subchannel busy (refusal) condition to the transmitting station, and preferably to remember the address of the transmitting station. The origin station does not retry the transmission voluntarily. But upon a sub-channel being prepared by its processing software, the receiving station...