The VAX system at Triumph Components has been in use since about 1996, and information systems manager Dan Blackshear couldn’t be happier with it. There’s just one problem: He’s got to scrap it. The system is fast, works well with Windows and “fits into the modern environment very cleanly” at the El Cajon, aerospace parts maker, said Blackshear. “But it’s a dinosaur, and eventually it has to go”.
VAX Users See the Writing on the Wall
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Yep, it will cost a bunch. Since no one uses a Q-bus then I wouldn’t count on all vax machines going away too soon.
I bet they will go for Linux, Linux is fast, works well with Windows and fits into the modern environment very cleanly and is free.
everyone sees the writing on the wall, except HP’s Fiorina…
Customers dont migrate from dec/Compaq to HP, the will move on to some other company.
Anyone seen the last numbers of Itanium-sales? IBM don’t even advertise their IA64-boxen and are about to take over the lead.
Quite funny, that the HP VAX website encourages you to migrate to the Alpha platform… you know, that platform that lost its FreeBSD Tier 1 status because it has been retired for so long.
Btw. that VAX emulator, on what platform does it run? OpenVMS for Alpha? HP-UX? Linux?
And now that I know nothing about VAX and big systems in the enterprise:
– Are those 150.000 VAX computers in use mostly desktops like the VaxStation 3100, or ones more like the VAX 10000?
– If they are the latter, what would be a logical upgrade, Alpha excluded? Something big from Sun? A cheap Linux box? Something completely else?
Read somewhere (don’t remember where) that they were going for OpenVMS on an Itanium box.
I don’t think the cost of the OS really adds into the equation when your buying systems in the market that OpenVMS is aimed at. OpenVMS on Itanium is a more likely port, no doubt there will be supplied software from HP that allows the VAX code run on Itanium (like FX32! on Alpha Windows NT for i386 binaries)
They should bring back VAX manufacturing, even if it’s age old architecture, if customers like it and the software’s good on it, why not? If you could couple that architecture with the best of breed latest technologies and offer this to customers to add into their clusters with support contracts etc, wouldn’t it be great? What a business that would be! I’m sure that Carly would fork over the know how for a few jet planes and stewardesses that knew how to cut cigars.
Digital’s business was hacked to pieces by the coming of personal computing, also because they built this awesome architecture called Alpha. Customers would have this one box that they could buy to do the work that a group of VAX performed. VAX service contracts were the sweet spot, Alpha was the disruptive technology.
And even still, not all customers migrated to Alpha, plenty are and were happy with VAX. Support was decent with Compaq until the coming of Itanium. So once again, the champions of personal computing have come to wreak havoc, all fear the mighty Intel even bearning gifts.
Enterprise computing is not necessarily about 64 bit monster machines with high throughput, the later VAXes only made it to 32–clustering, stability, support, and software made it a great Enterprise solution. So this is a great hommage to the engineering teams that built the damned thing, three cheers!
Now that brings back memories — I ended up with a free vax back when, and ended up converting it to a pool-side bar, using the hard drive tray for the beer cooler. It worked great till the florida weather rusted it out… A good use for a 1 MIPS computer
The 8 in hard drive platters are still going strong though – they make excellent wind chimes. Or maybe I should call them hurricane chimes.
…all the VAX and OpenVMS comments have been good. I was an OpenVMS engineer and I always thought the OS was fantastic, but also always preferred VAX architecture over Alpha with VMS. The VAX was simply rock solid, and VMS was written to it. You could boot with 4 mb of ram and still serve multiple users (on terminals). When the workstation came into being, you could run it and X/motif with 16 mb ram and you were in heaven if you had 32. Sweet OS on a sweet platform. We are in the process (in the company I work at now) of porting all of our VAX/VMS apps to Linux/Java. I don’t get into the lab often (they try to keep me out
so I am not sure what it is (7000 maybe!? I don’t know) but I wish I could take it home with me and stick it in my home office. Heck, it would heat my basement in the winter. 
But seriously, I left DEC about 7 years ago and there were still so many VAX/VMS customers that OpenVMS was never touched during all the layoffs. I wish them all well. Maybe they WILL try and port it. I hope.
Mike
Vax is nice. I hope the porting goes well and it’s recognition for stability is not undermined.
I’m curious to know how the stability was on the Alpha considering it’s faulty hard-wired memory management
I guess I am not sure what you mean? Can you elaborate?
Mike
that someone who is up with the play on both VMS and WinNT and their interrelationship:
Windows NT and VMS: The Rest of the Story
http://www.winnetmag.com/Windows/Articles/ArticleID/4494/pg/2/2.htm…
will combine with work that’s being done with cloning the WinNT OS
ReactOS User/Developer Site
http://www.reactos.com/
and the various efforts at cloning OpenVMS
FreeVMS
A free but Very Much Strange operating system
http://freevms.free.fr/indexGB.html
Welcome to FreeVMS
http://www.freevms.org/
FreeVMS – Save the World’s Best O.S.!
http://www.djesys.com/vms/freevms/
And use the ReactOS kernel as the basis for a combined FreeVMS effort. There’s bits and pieces of FreeVMS already up-and-running, it just all needs to be unified.
I’m curious to know how the stability was on the Alpha considering it’s faulty hard-wired memory management
You have it backwards, my poor deluded chap. Alpha actually has lots of its memory management done by the system software ie. the opposite of hardwired. I guess this is a logical extension of the RISC philosophy, and is by no means faulty.
“IT managers interviewed about VAX and its OpenVMS operating system said the machines rarely fail. Geoffrey Ive manages 150 VAX systems, including one bought in 1986 and others purchased in 1990 and 1995 by his South African company, Eskom Transmission. Some of the systems have run for more than six years without a reboot. Reliability is “extremely high,” he said.”
Why would anyone want to retire these machines? Fast, reliable and so forth. I havent used a VAX since my college days, however, can you argue this stability with any other OS, commercial or not.
Seems like todays users and admins are so used to doing the three button salute. Tis a shame.
ofcourse HP is pushing theirs Tru64 Unix version to the market
The VAX architecture is one of the best I have ever worked on but the reality is the hardware is at end-of-life.
Maintenance costs for the hardware become more and more expensive as the years progress.
If it is too expensive to port the applications to Alpha OpenVMS, the pending Itanium OpenVMS, or an alternate OS, there are two viable solutions.
The first option is to use CHARON VAX.
The CHARON VAX emulator is a good solution but it is is expensive and relies on a Windows 2K or XP base OS.
The second option is to use the open-source SIMH VAX emulator which can run under both Linux and Windows 2000/XP.
It was very easy to install/configure VAX 7.3 under SIMH using Fedora CORE1 and Fedora CORE2.
I highly recommend anyone facing the port/retirement issue to give SIMH a try.
Btw. that VAX emulator, on what platform does it run? OpenVMS for Alpha? HP-UX? Linux?
It runs on linux and windows at a minimum, and probably most unix and mac. you probably need (win)libpcap. the one i happened across is called simh. simh had a cap of 64 meg on guest vax, now i think it’s 512 meg. i’m thinking about just getting an alpha since they cheap now, $100 or so.
there’s also charon-vax emu, but there is no longer a hobbyist program. you can get VMS (vax or alpha) for $30 and Tru64 (alpha) for $100 in the hobbyist program. big thanks to HP for continuing this legacy program.
anyone wanting to tinker with VMS can go the Deathrow Cluster. there’s also some shells in belgium, but they require international postcard as validation. i also know of several semi-private alpha and vax.
i’ve heard of 12 year uptime on one vax! i was in high school using a 286 when that thing started service.
And use the ReactOS kernel as the basis for a combined FreeVMS effort. There’s bits and pieces of FreeVMS already up-and-running, it just all needs to be unified.
Last time i checked the freevms list, there was only one developer. Seems bizzare given the big interest in hobbyist vms. and simh is popular, so a lot are running it on i386 in a sense.
as far as using reactos as kernel, that’s an idea. depends on how developed the freevms kernel already is. last i saw there was a bit of kernel and some user space. wish i had a test box to try it on.
ofcourse HP is pushing theirs Tru64 Unix version to the market
HP (compaq) is pushing linux so much i don’t know. i don’t even know if HP has a distro like SGI does. but they are promoting OSS.
it seems like both vms AND tru 64 are being neglected. that said, there have been some new ad campaigns for them.
OpenVMS gets ported to Itanium and True64 will be merged with HP-UX.
I am so happy to hear news like this. I had the misfortune of working with VAX/VMX when I started out. It was the most antequated backward pieve of crap I ever set eyes on. All the people I had to work with were also backward.
It nearly cost me my career. I hope its one of the last things I hear about VAX and or VMS.
Its a world inhabited with ‘why change it, it will not get any better or when i was a kid’ sorts of people.
Imagine where we would be if people refused to stop writing on stone.
It nearly cost me my career. I hope its one of the last things I hear about VAX and or VMS.
I hated VMS when I first used it in ’97. I borrowed a friend’s shell at his uni. I used it for a few things, but liked solaris far more.
Years later, I learned of VMS’s superiour stability, clustering, and security. Technically, it’s very very good. So are Multics and some other “dinosaur” systems.
Don’t judge it purely by DCL, that’s very unfair.
“…..True64 will be merged with HP-UX.”
Which is going to be cool to get the awesome clustering that Tru64 has in HP-UX..service guard whats that?
Wow. I never knew that the Vax was that good. I only used it for a little while, on the client side for writing programs,but now PCC is getting rid of it also.
and will be released 2nd Half of this year.
OpenVMS is not dead yet (even though HPAQ has no idea what they are doing).
The Alpha was killed by Intel when Compaq sold them the IP.
The world’s record for uptime (over 20 years) is still held by a VAX. I don’t think that record will be challenged for sometime
ok, ok, ok that was a bit harsh :s . reading the other comments I will have to concede it cant be that bad – especially if its still around and ported everywhere.
“VMS’s superiour stability, clustering, and security…”
We hear this from time to time in here, and yet VMS remains after all these years an obscure OS limited to only a couple of platforms. Why? There must be a catch to it, no?
I work in a company that over the years has delivered several hundred VAX/VMS and Alpha/OpenVMS systems for SCADA at utilities and DOT’s.
But times have changed. The company has now switched to x86/RH as platform for the productline. In doing that a large portion of the product software was ported – to get the product on its feet on the new platform in a timely manner. Most of it is in C/C++ and porting was a breeze and routine work once the base services had been ported.
New software is written in Java and the plan is to eventually replace all the ported C/C++ software with new written in Java for a product entirely based on x86/RH/Java.
The former platforms served everybody well. No complaints. But now after having taken the plunge and gotten used to the new platform the company hasn’t had second thoughts about the platform switch.
Windows was considered but lost out – mostly because of price, cost of porting/rewrite and the risk of lock-in.
I’m the one that made the original comment about the hard-wired memory management on the Alpha.
I was working on one of our embedded devices all day and got mixed up when I got home (damn four 10-hour work days instead of five 8-hour days). I was wrong! I guess a I am a poor deluded chap…..
sorry
Now I haven’t used VMS, but why is it so awful? And how can using VAX hardware lead to a lost career?
From what I hear, those VAX computers are (were?) very reliable, and to me, that sounds very positive. My noname PC has had about every piece broken (PSU, harddisk, memory, keyboard, 2x monitor (Belinea, Philips), speakers, CD-ROM) so I wouldn’t mind sacrificing a little of usability for a machine that doesn’t break every six months. Yet apparently you would, so what’s so terrible about VAX and VMS?
Yet apparently you would, so what’s so terrible about VAX and VMS?
VAX/VMS is a dinasour, like already mentioned, and like all dinasour we wish it to do what all dinasours have done a long, long time ago: Die out.
I don’t know if the VAX/VMS community will welcome my comparison, but i will do it none the less.
As VAX/VMS are dinasours I’d like to compare it to IBM’s mainframes, which were dead-said by the day microcomputers aorse. But the mainframe still exists. Why would a joungish mind now ask? Why not replace it with PCs? With workstations? With a server-client architecture based on *NIX?
Because it’s been running for decades now. And because it runs particulary highly customized software written in COBOL, PL/I, ASSEMBLER. For some of these software the sourcecode vanished a long time ago, so programmers had to manipulate their binaries to apply the coming changes.
Porting this software to an alternative system, aswell as it’s establishment, would cost millions, the sudden lack of reliability and throughput — some IBM mainframe architecture can sustain 32k simultanous queries — would again cost millions.
Banks and insurance companies try to earn millions not waste them on IT — and that is the exact reason why the mainframe will never die.
Of course there are efforts to supply a better user interface.. Front-Ends in Java for Workstations or in Perl for the Web, are already reality, but the programmer interface is still a Pain in the ass.
I’ve been programming 370 ASSEMBLER and PL/I on such a Mainframe in school — and believe me compared to an IBM 370z running MVS [without CMS!] VAX/VMS, which is a multitasking OS that even can run X, VAX’ VMS is heaven.
I don’t understand HP’s move. But this is probably just because I don’t like the ix86 architecture, and the Itanium is a child of this architecture…