Browse Prior Art Database

Recovery Procedure for Data Channel

IP.com Disclosure Number: IPCOM000104905D
Original Publication Date: 1993-Jun-01
Included in the Prior Art Database: 2005-Mar-19
Document File: 2 page(s) / 82K

Publishing Venue

IBM

Related People

Casper, DF: AUTHOR [+4]

Abstract

An I/O (input/output) subsystem in a data processing system is made up of I/O devices and a system of busses and processors that transfer data between the devices and central processor memory. A magnetic disk storage device is a familiar example of an I/O device. In some I/O subsystems, processors called the channel processors (or channels) handle data transfers, and a processor called the IOP communicates with the operating system in the central processor to begin an operation and later to report that the transfer has completed successfully or that it has failed. A processor called the process controller (PC) handles other situations such as errors. When an error is detected by the channel, the channel reports separately to the IOP and the PC, and the IOP reports to the operating system.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Recovery Procedure for Data Channel

      An I/O (input/output) subsystem in a data processing system is
made up of I/O devices and a system of busses and processors that
transfer data between the devices and central processor memory.  A
magnetic disk storage device is a familiar example of an I/O device.
In some I/O subsystems, processors called the channel processors (or
channels) handle data transfers, and a processor called the IOP
communicates with the operating system in the central processor to
begin an operation and later to report that the transfer has
completed successfully or that it has failed.  A processor called the
process controller (PC) handles other situations such as errors.
When an error is detected by the channel, the channel reports
separately to the IOP and the PC, and the IOP reports to the
operating system.

      This channel recovery procedure begins when a latch is set
signifying that an error has occurred on the channel bus (an
interface check) or in the channel processor (a channel check).  When
a channel check occurs, the clocks in the channel are stopped and
normal channel operations end.  When an interface check occurs, for
example, a parity error on the bus from the device, the clocks in the
channel are not stopped, and normal channel operation can continue
with other devices on the bus.  The general recovery procedure has
the following steps: (1) the channel signals the device to disconnect
from the channel; (2) the channel puts data into a channel memory
called the trace array and sends an interrupt to the PC and the PC
reads the trace array and begins its part of the recovery procedure;
(3) the channel reports to the IOP and the IOP reports to the
operating channel; and (4) the operating system issues instructions
to the channel (Halt, Clear Subchannel) to retry or restart the
operation.

      Sometimes a second error occurs before the PC has read the
trace data for the first error.  In this recovery procedure, the
channel checks the status of the first trace data and it does not
enter the trace data into the trace array in this situation.  The
channel proceeds with the re...