Columbia University Computing History   

CU20A Epitaph

30-May-87 12:49:06-EDT,2456;000000000000
Mail-From: SY.KEN created at 30-May-87 12:49:04
Date: Sat 30 May 87 12:49:04-EDT
From: Ken Rossman 
Subject: The demise of CU20A
Address: 715 Watson Lab, 612 W 115th St, NY NY 10025
Phone: (212) 280-4876
Message-ID: <12306533774.9.SY.KEN@CU20B.COLUMBIA.EDU>
CU20A, CUCCA's longest running and uptime record setting DEC-20, has reached the end of its lifetime here at Columbia. It will be retired on Friday, June 12th, 1987, making it nearly 10 years old. Academic users wishing to continue to use DEC-20 resources should use CU20D from now on. CU20A was installed in 1977 to supplement the DEC timesharing computing resources then offered by a PDP-11/50 running the RSTS operating system. At the time, CU20A ran TOPS-20 version 3A, used a single RP06 disk drive (76,000 pages total for both system and user files), had 256K words of core memory (actual magnetic *core* memory) to run everything in, it ran as a DEC-2040, which didn't have any fast cache memory, and it had no networking resources whatsoever.

Since that time, CU20A has seen many upgrades and improvements, both in hardware and in software. It has seen many versions of the TOPS-20 operating system come and go (versions 3, 4, 5, and the current 6.1), along with many CUCCA enhancements to the operating system and working tools. It has had many more RP06 disks added to it to expand its online storage capacity, and most recently, with the advent the Common File System, RA81 Winchester-like storage media was added to further increase the online storage capacity. It has seen its memory increased and upgraded from 256K words of core memory to 2M words of MOS memory. It was upgraded from a 2050 to a 2060, bringing things like fast caching memory and extended addressing into the picture, increasing both speed and functionality. It has had DECnet and TCP based networking (as well as other network protocols and services) added to it in various different forms over the years, vastly increasing its flexibility in a variety of ways.

In short, CU20A has seen just about all the possible changes and improvements any DEC-20 anywhere might have seen, and has (hopefully) served its user community well in the process.

See the May 1987 issue of the CUCCA Newsletter for more details on CU20A's demise, and the future of central mainframe computing at CUCCA in general.

12-Jun-87 17:24:46-EDT,857;000000000001
Mail-From: SY.KEN created at 12-Jun-87 17:24:38
Date: Fri 12 Jun 87 17:24:38-EDT
From: Ken Rossman 
Subject: Demise of CU20A
To: CCnet-Managers@CU20B.COLUMBIA.EDU
Address: 715 Watson Lab, 612 W 115th St, NY NY 10025
Phone: (212) 280-4876
Message-ID: <12309991811.25.SY.KEN@CU20B.COLUMBIA.EDU>

CU20A, our first, and longest running DEC-20 (nearly 10 years!) was shut down forever at noon today. You might want to change any references to it in host tables, and either remove it, or perhaps better, make CU20A an alias for CU20D, since those two machines are CFS mirror images, and CU20D is still around (at least for another few minutes anyway -- naw, at least another half a year, actually).

CU20A's node number will *probably* go to CUNIXA, the new 8700 coming in during July sometime. FYI, /Ken

Columbia University Computing History Frank da Cruz / This page created: May-June 1987 Last update: 3 April 2021