Browse Prior Art Database

Automated Large System Recovery Testing

IP.com Disclosure Number: IPCOM000105327D
Original Publication Date: 1993-Jul-01
Included in the Prior Art Database: 2005-Mar-19
Document File: 2 page(s) / 46K

Publishing Venue

IBM

Related People

Ross, MT: AUTHOR [+2]

Abstract

A method is disclosed for the testing of large computer system's recovery handware in a totally automated environment.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 98% of the total text.

Automated Large System Recovery Testing

      A method is disclosed for the testing of large computer
system's recovery handware in a totally automated environment.

      Increasingly more hardware is allocated for fault tolerance.
During normal computer operation most of the hardware devoted for
fault tolerance and recovery will not be used unless a fault occurs.

It is critical to test that portion of hardware during manufacturing
Test and Acceptance at system cycle time in addition to the
self-tests performed at DC cycle times.

      This have been achieved using a series of programs that cause
various faults in the system during manufacturing Test and Acceptance
period.  These programs use built-in hardware to inject various
errors (faults).  These tests are monitored via Computer Aided System
Test (CAST).  Expected results for all the injected faults are
clearly defined.  The results include events seen by the Operating
System and Serviceability data.  Serviceability data is shown in an
Action Query Entry, (AQE) format.  Also, in order to stress the
hardware for fault tolerance even more, all these tests are performed
under frequency and voltage bias.

      The CAST supervisor program invokes the test program to inject
the fault and monitors the system under test for expected results, as
shown in the flow chart.  Each test is designed for a signal Field
Replacement Unit (FRU).  If the result does not match the expected,
that FRU is replaced an...