Amdahl Remembrances

Created by Lori Lesnefski September 26, 1997

This machine has served the engineering community for 10 years now.  It has quite a legacy.  It was put in service on April 16, 1987 at a hefty lease price of $73,000 a month.  The configuration was four domains: two UTS and two VM, 192MB of memory, 10 reel to reel tape drives, Amdahl and IBM DASD, and Amdahl 4705 Front End Processors.  The operating system release was UTS 1.1 which was installed by Nick Maniscalco.  The total three-year lease amounted to 1.8 million dollars.  It was then purchased in 1994 for $47,000.

It was in 1987 where Bob Salvatore learned the ropes as the UTS Release Manager packaging the new and improved 1.2 release.  Luckily for Bob, Al Goddard was good at untying the knots that Bob got himself into with this release.  All product and operating system testing was done on acpy02.  When the testing concluded on acpy02, everything was then frozen on the SOE machine, acpy01.  This release was distributed to the Field Support Group sites for installations and remained there a couple of years.  During Bob's reign as the UTS Release Manager he delivered three ull releases (2.0.1, 2.1.2, and 2.1.3) and several maintenance releases.

The SRDB database made it's home on acpy01.  The first recorded service request was on July 21, 1987.  Untill SRDB was forced to retire on October 15, 1996, it recorded 15,400 service requests.  By severity, there was 837 severity one, 3,311 seveirity two, 9,665 severity three, 1,183 severity four and 399 severity five requests.  The statistics show that nearly one third or 4,558 of those requests were Amdahl related.

In June of 1988, Bob Salvatore was assigned to be my mentor as the UTS system administrator.  During this time, I endured the product testing of BX.25, TCP/IP, Datakit, AFT, and many, many database benchmarks.  Where I really earned my stripes was during the spring months.  Typically, this rainy season brought us lots of electrical storms.  There is not a doubt in my mind, that at the first sight of lightening, many of us had the same thought: The Amdahl is going to crash!  I was fortunate to have the assistance of the operating staff, Barbara Barbano and Nancy Mircoff to help recover the system quickly, call the vendors and restore data.  One spring, I recorded restoring 7 GB of data due to the abrupt losses of power and to a sticky-head syndrome that plagued the IBM single-density DASD.

While we are on the subject of file system restores; there was this incident back in 1989 tha those of us with tenure will certainly never forget.  Jose S. was debugging a problem with Ingres backups and restores on acpy01.  Jose wrote a cron job that was to remove files within his working directory.  This script caused such havoc on the system due to it's "less than perfect programming" that it eventually worked it's eay up to the root file system and removed everything on acpy01 in just a matter of minutes!  Jose did not use the absolute path.  What a disaster!  Al Goddard and Nich Maniscalco were in the middle of debugging a severity 1 system crash for the USS application.  Al and Nick had to take a prop plane to Fairhaven so they could continue debugging the problem on their system.  It was 3am when they noticed that the same script started to do it's thing on the Fairhaven test system.  What a hilarious site it must have been to see 3 men, Bill Bannister, Al and Nick fly across the console room to press the hard STOP button on the processor to prevent another disaster.  It took almost a week to restore our local system back to its original state.  Does anyone know where Jose is today?  Despite this catastrophe that Jose caused, he was the first Ingres backup guru that DP&CT had.  One morning Bill Postel was awakened at 1am and informed that the Denver Data Center backups were failing.  Bill knew that Jose was the only one that knew the software.  Bill didn't have Jose home number so he called his number at work.  To his surprise, Jose answered the office phone and had the problem resolved in a matter of minutes.

This machine has surpassed many of the powerful leaders that tried closing DOTC and wanted to move the Amdahl to another data center.  The Amdahl sure did outlive those powerful leaders.  It is unfortunate that the legacy must come to an end due to the economic conditions of the business and stagnation of UTS in the Business unit.  I am happy to say that the donation agreement forms have been approved by the Public Relations Department and our senior management to donate the Amdahl 5890 to the Information Age Technology Museum as a functioning example of the 1980's mainframe technology.  It will serve its purpose well.  This museum is located in Wall Township, NJ at the US Army Signal Corp. research site, Camp Evans.

In conclusion I would like to share these words with you from Alexander Graham Bell:

"When one door closes, another opens.  But so often we look so long and so regretfully upon the closed door that we do not see the one which has been opened for us."

AT this time,  I am going to power the Power Distribution Unit to power off the machine for the last time.   As the machine shuts down, I would like everyone to sign the inside of the door panel as a remembrance of those who were associated with the machine.

Page updated December 29, 2003  page created September 02, 2000 Copyright©  Infoage 1998-2003 InfoAge. All rights reserved.