IBM System/370 Guide

The dynamic device reconfiguration (DDR) routine permits the operator
to move a demountable volume from one device to another of the same
type when a permanent hardware error occurs and provides repositioning
of the volume so that the failing I/O operation can be retried. A
volume can also be demounted. so that device cleaning procedures can
be performed. and then remounted on the same device. The DDR option
also supports demountable system residence devices and unit record
equipment. DDR is entered from a device-dependent ERP after a permanent
error. either channel or device. occurs on a demountable device. Task
termination occurs if the error cannot be corrected and a user-written
permanent error-handling routine is not present.
ADVANCED CHECKPOINT/RESTART AND WARM START FACILITIES
If the RMS and I/O RMS routines fail in their attempt to correct
a hardware error and the error is one that causes a program or system
termination. the automatically provided advanced checkpoint/restart
and warm start facilities of OS can be employed to minimize the impact
of these terminations on system operation. The automatic restart
facility can be used to cause terminated programs to be rescheduled
immediately without resubmission of their job control. so that a minimum
of operator intervention is required. The operator must authorize
all automatic job step restarts. If a permanent I/O device or channel
failure caused the program termination. the device or channel can be
varied offline. This will insure allocation of a different device
when the program step is reinitiated.
The warm start facilities of the control program provide automatic
saving of SYSIN and SYSOUT data sets and input and output work queues
so that processed work is not lost when a system termination occurs.
The operator is informed of the status of jobs in execution when the
system terminated. and these jobs should be restarted automatically
from the beginning or from a checkpoint if the type of processing
involved permits such a restart. System design should include planned
restart procedures for unscheduled terminations of individual programs
and the system.
50:15 REPAIR FEATURES
The programmed repair features supplied are designed to minimize
the impact of malfunction diagnosis and repair on system availability.
Fault location and repair time should be reduced by:
1. Improved error recording. Both intermittent and solid hardware
and I/O errors are logged at the time of failure if Model 165
RMS routines are present. More status information will be
available than was recorded previously for Models 65 and 75.
2. Online error diagnosis,. Error diagnosis and repa1r can be
performed concurrently with system operation in a
multiprogramming environment to avoid total system or direct
access facility unavailability.
3. CE diagnostic routines. More exacting diagnostic routines will
be available.
The following maintenance and diagnostic routines will be provided:
• Online Test Executive program (OLTEP) and Online Tests (OLT's)
for operation under OS control to test malfunctioning I/O devices
concurrently with system operation. OLT's are provided for most
Model 165 I/O devices.
83

IBM System/370 Specifications

General

Brand	IBM
Model	System/370
Category	Server
Language	English