• Please review our updated Terms and Rules here

How were VAX/VMS systems deployed during their heyday in the 1980's?

tsm75

New Member
Joined
Jul 21, 2023
Messages
3
I've been playing around with VAX/VMS systems as a hobbyist for more than 20 years (I have owned MicroVAX II, and 3800/3900, and others), but I am not very familiar with how they were originally deployed. Besides from being used as industrial control systems, I have heard them described as "departmental servers" which is pretty vague. But I am interested in what actual applications they ran, and how users interfaced to them. There must have been some VAX/VMS installations which were large, are there any examples? What industries used them and who were some of the large VMS customers?

When you read about history of business computing in the 80's it is almost exclusively about IBM PC and I never understood how VAX/VMS fit in and what exactly it was used for aside from a few niche applications, but all indications were that it was enormous money maker for DEC so there must have been a lot customers. I am well aware of Unix (Ultrix or BSD) VAX installations (especially in research, academia, engineering)--usage models which I am thoroughly familiar with), but it is the VMS side of things which I am much less familiar with how it was used commercially. But as far as I understand, VMS accounted for vast majority of VAX sales.
 
VAX/VMS provided the "company mainframe" for a lot of businesses. It took a while before PCs were big enough to run company-wide or departmental databases, etc.

An employer of mine used a VAXcluster for many business functions, including Manufacturing Resource Planning (and probably payroll), at least until the mid-90s when I left. I ran a team developing software to implement customer deliverable systems, so I only needed access to keep up with what the hardware folks were ordering/building to run my team's software on. I also could use it for general VAX/VMS software activities, but my disk quota was small. I remember being impressed that the node I was logged into one day had a gigabyte of RAM (when 486s maxed out at 64MB).

My team used MicroVAX 3100 systems for customer deliverable systems for a while, so we had 'em in-house for development and testing. And we ran cross-development tools for our microprocessor based embedded systems we delivered.
 
Last edited:
When you read about history of business computing in the 80's it is almost exclusively about IBM PC

Seriously? IBM "International BUSINESS Machines" Had many different product lines. The PC was the least powerful of them.
So much so that their initial intended use was as peripherals for their "real" computers.

VAXen fit into the niche of "Superminicomputers". The result of the industry moving from 16 bit (PDP-11) to 32 bit machines.
The original name of the VAX was the VAX-11.
 
The VAX had three major markets:
1) The comparatively cheap mainframe for smaller companies. Stick all your software on it and run lots of terminals.
2) The department server. Hang a lot more storage off a small VAX than any PC derived system could until the PC derived designs caught up.
3) The scientific workstation. Just like the rest, effectively a much faster PC where reducing time is more important than getting cheaper computers.

Unfortunately, improvements to micros especially CPUs hollowed out 2 and 3 while IBM was doing a lot of price cutting on its mainframes reducing the high end. DEC's internal competitiveness played a bit too; it was often easier and cheaper to install a non-DEC product than the proper DEC product to interface with another DEC product.

DEC is Dead; Long Live DEC is perhaps the easiest history of the VAX glory years and how it imploded. Big Blues is the tale of IBM's decline in the 80s which also covers how IBM was slashing the mainframe pricing making DEC's effort to undercut IBM challenging.
 
In the early 80s, (1983 or so) we ran a 11/750 setup on the second floor of an office building. Landlord made us build a floor reinforcement overlay and we put in 208 3-phase. I don't recall too much about the configuration; multiport cards for a bunch of CIT-101E terminals. I used my 5150 with a homebrew 15" mono monitor--far more useful than a terminal. I think the setup had a Cipher 1600 bpi streamer tape, a Fuji Eagle and a CDC Hawk cartridge drive. We also had a couple of fast CDC line printers (Dataproducts interface, I think).

We ran BSD Unix on it. It wasn't too many years before microcomputers eclipsed the lower VAXen. I was even offered an 11/730 free gratis (operational) if I would just haul it away. I declined.
 
In 1988 I was working for a large state college bureaucracy which was undergoing a sort of merger with another state college bureaucracy. About a dozen of folks from the other bureaucracy came to work in our office building and stuff that came with them included some DECmate II and DECmate III systems, a VAXmate, some VT220s, a VT240, a couple LN03s, some DECserver 100s and 200s, and a MicroVAX II running first MicroVMS 4.7 and then VMS 5.0. And ALL-IN-1, and WPS, which were what most of 'em used.

The IT department that furnished the equipment openly told us we were making the wrong choice moving our word-processing workload off the HP3000 to PCs, and that we should get rid of the PCs and get a VAXcluster to run ALL-IN-1 and other important stuff. And a bunch of VT220s. We had kinda been there and done that with HP's word processing software (TDP/3000 and HPWORD) and what we saw as the coming thing that was worth keeping on the 3000 was electronic mail (HPDeskManager).

(The IT department that furnished the equipment also had a spreadsheet product on their VAXcluster, something called 20/20. We had Visicalc on one of our 3000s and found it got in the way of other processing until auto recalc was turned off. Both we and our customers had found we were happier with Lotus 1-2-3 on HP150s and then HP Vectras, even if we had to run around with HP 9142 tape drives and back all the PCs up.)

Having to work with VMS, I thought it involved too much typing (especially DCL), and I was coming to this from a background working with classic-3000 MPE which wasn't known for Unix's brevity.
 
You would find VAXen almost everywhere. They had the great advantage that whilst some things are quirky it was perceived and sold as being more user-friendly than IBM Mainframes, which if by IBM Mainframes you meant MVS it was (IBM also had VM/CMS which was much simpler for users). So when I worked in the Science Research community in the UK we had both VAXen and IBM Mainframes running VM/CMS. In that environment it was both to store and manipulate data, as well as to perform calculations. We had both to ensure access to the widest range of software.

But you would also find them in office automation environments. The All-In-One product was ground breaking in offering Messaging, Time Management and Word Processing in one package.

For me their one limitation was lack of scalability due to the character mode terminals, so when doing forms or screen editing every character must be processed and echoed by the central CPU . On an IBM Mainframe there are separate intelligent screen controller which store the form and associated input data and only pass it to the host program when the whole form is complete, resulting in many fewer IO interuptions.

So a friend who workde in a big bank bemoaned the fact that for the higher management All-in-One was an option, the great unwashed were stuck with the IBM PROFS product when he deemed inferior. (I prefered PROFS)
 
Last edited:
Thanks everybody for all of the replies. This helps a lot. But I have some more questions.

The VAX had three major markets:
1) The comparatively cheap mainframe for smaller companies. Stick all your software on it and run lots of terminals.

From a hardware point of view, which models would that have encompassed? Would this have been mainly the larger VAX models (6000, 8000)? Would earliest models (11/780) fit into this category? Several replies mentioned ALL-IN-ONE which I was not familiar with before but sounds like it was a major driver for these machines. What were some other widely used applications?

Any idea on how many terminals(users) per machine there would have been? Order of magnitude--tens, hundreds, more? How far from the physical mainframe location would users be?

2) The department server. Hang a lot more storage off a small VAX than any PC derived system could until the PC derived designs caught up.

For this type of server, what would have been the client? PC's? And what would have been the server applications?

And again, what HW class would have been used? Would MicroVAX II or the later and bigger MicroVAX 3x00 have been used for this?

When we talk about VAX clusters, would those have been more of a mainframe usage model, or used for department servers also?

3) The scientific workstation. Just like the rest, effectively a much faster PC where reducing time is more important than getting cheaper computers.

So, the VAXstation line--running DECwindows on user's desktop, and developing and running scientific and engineering applications. This usage model I'm familiar with.
 
I worked at a company developing atmospheric radar systems and we had the following machines:

uVAX II w/Ultrix for engineering
uVAX II w/VMS for corporate stuff, primarily a Manufacturing Information System (parts list, BOM, inventory control)
VAXStation 2000 (or two?) for documentation (Framemaker)

The microvaxen were all controlled with VT220s through DEC terminal servers. I think the site only had a few PCs, but I don't remember how they connected to the network, probably PathWorks.

Customer systems were Vaxstation II/GPX with VMS for data display (frontend was a 68020 to do the collection and crunching).
We also shipped a couple Codar ruggedized uVAX II/GPX systems as customer units as well.

CW
 
ALLIN1 was Office Automation. It did e-mail that was fancier than the standard VMS mail, it included WPS for word processing, I'm not sure what else it did or why a third-party spreadsheet (20/20) that was neither Visicalc nor 1-2-3 was desirable. If you had your VMS systems connected via DECnet it could send e-mail between 'em.

The university where I worked had selected the HP 3000 mini ca. 1980 as a platform for financial applications (budget, financial accounting, payroll) that were to be distributed across its campuses (previously they had been centralized on a Univac 1108 at one of the larger campuses). My recollection is that the other university in the merger used their VAX/VMS systems for similar purposes too. The office automation applications had come later for us and may have for them too.

Imagine bridged Ethernet across the state so you could connect to a VAX somewhere else from the VT220 on your desk which is connected to a DECserver 100 or 200. DEC's LAT protocol would let you do this. In the early 1990s, PCs running Doom over IPX would make it look like the really bad idea that it was.

VAXclusters were about resource sharing, increased computational capacity, and low-end fault tolerance: if one of the VAXes in the cluster went down, CI-attached disk drives meant other VAXes in the cluster could continue providing services; but running processes didn't migrate from a failed system.
 
Back in 1988-1990 I worked at DEC in Littleton, MA (LJ02).

http://www.bitsavers.org/pdf/dec/in...igitalOrganizations_1988-1989Edition_1988.pdf
9.4 PERSONAL COMPUTING SYSTEMS GROUP
The Personal Computing Systems Group (PCSG) is responsible for the integration of industry-standard PCs and Apple Macintosh systems into the Digital environment. PCSG provides products which leverage upon Digital's strengths in VAX, VMS, LAVc, DECwindows, and DECnet technologies. PCSG is responsible for integration strategies pertinent to these markets and for proposing future PC integration products/systems. The group's goal is to assure that Digital becomes the #1 integrator of PCs in the industry.
That was one of the smaller sites at the time, but must have had at least 100 or so people there. Hard to remember the details now. In the main computer room in the center of the building behind the large glass windows the main system for the site was a VAX VMS cluster. From the timeframe it was most likely an 8200, 8600, or 8800 class system. If I remember correctly it was a dual node system, with hosts MOSIAC and RAINBO.

While there was probably some VT220 use connected to the VMS cluster though DECservers, it was probably most common to connect to the VMS cluster through DECnet or LAT from a PC on your desk, since the focus there was on PC integration. Every few cubicles there was an AUI cable dropping down from the ceiling connected to a DELNI or DEMPR, with the PCs connected to those.

While we were doing software development for PCs, the source code revision control system was on VMS, maybe that was CMS (Code Management System)? At the time maybe most people did most of their source code editing on the VMS cluster. There must have been other software development applications we used on the VMS cluster, such as bug reporting databases, and development schedule tracking. It was also used as a disk server to the PCs over the LAD protocol. There might have been some file server use too, I don't remember.

The other main use of the VMS cluster was of course company email. All of the printers (PrintServer 40, LN03) were connected to the VMS cluster. Another thing I remember using on the VMS cluster was the company wide bulletin board / discussion forums hosted on VMS Notes.
 
That VAX 11/730 was used at Sorcim, along with a CompuPro 86/88 system and an AT system of some sort to build many of the Sorcim products, such as SuperCalc, SuperProject, etc... Martin Herbach remained with the company after it was acquired by CA because, apparently, he was one of the few who knew how to build the products.
 
So, the VAXstation line--running DECwindows on user's desktop, and developing and running scientific and engineering applications. This usage model I'm familiar with.
I think that is possibly the least common. When the VAX started, in 1977, the IBM PC was 10 years away. In fact on checking, even the VT100 was a year away so all users got a dumb terminal. I think most would have a VT52 or VT100 video terminal or a DecWriter printing terminal. If you were really lucky a Tek40xx graphics storage screen. On the original VAX 11/780 which was a 1Mip machine you might run up to 20 terminals I guess, but it would depend on the work. On a big VAX perhaps a couple of hundred....

... also remember networking and communications were evolving over this period, so in the beginning most terminal were local. remote access was almost exclusively by dial-up, initially 300 baud. As that evolved more remote working was possible. .....

...not sure how many user we got on our big vax? 100 say?
 
We used to have a large VAX system at our CEGB headquarters in London. I do not know how many local users it had - but we had a number (me included) distributed around the UK at various CEGB sites on the end of a dial-up line (to start with) and then a dedicated line (later on).

We mostly used the VAX as either a modelling machine (e.g. FORTRAN programming) or document processing (Latex/Runoff etc.). It was also an email and messaging server.

It was also our personal storage machine for the 'office'.

As VAXes reduced in price, each CEGB centre had it's own VAX cluster.

We also used VAXes to develop software, and monitoring for our power stations.

Dave
 
On the original VAX 11/780 which was a 1Mip machine you might run up to 20 terminals I guess, but it would depend on the work.

When I was in college there was regularly 50+ students on the 780, and it wouldn't start getting too slow until it broke 100. I can remember spending close to an hour one night just logging in when a big project was due, but I don't think I wasted a command to see how many people were logged in. [type a command, then hit <Space> every few minutes to keep the terminal server from disconnecting you]. I don't recall the machine configuration, but i assume it was fully loaded with memory (8MB?), and a bunch of disks.

CW
 
Back
Top