Operating System - OpenVMS
1748169 Members
3530 Online
108758 Solutions
New Discussion юеВ

Re: VMS vs. U*ix - any horror stories?

 
SOLVED
Go to solution
Cass Witkowski
Trusted Contributor

Re: VMS vs. U*ix - any horror stories?

You can always show your boss John Wisniewski's recent security briefing

http://www.mindiq.com/resources/webcasts/JohnwEncompasswebcast031804.ppt

John will be missed!
Paul Jerrom
Valued Contributor

Re: VMS vs. U*ix - any horror stories?

Martin: Thanks, I am familiar with the success stories, my company is one of them. Unfortunately it is hard to tell new VMS users from old ones in most of these (actually, it appears all have been using the OS for yonks.)
Mike, Willem, John&Jan: Great ammo, thanks.
Andreas: Thanks, I have been in touch with my friend Sue. Unfortunately most of the success stories, including the VA one, are for existing VMS customers who have been using the OS for years.
Brad: I have requested this info from my local Ambassador often. Seems little 'ol HPNZ has other focusses...
Antoniov: thanks for the story, but unfortunately you fit into the category of pre-existing VMS customer.
Cass: No, I hadn't seen that presentation before, thanks for the pointer. And yes, John will be sadly missed.

From what I can fathom, the attempt will be made to show that VMS resources are scarce and limited to old wrinklies like me, who attract a reasonable hourly rate, rather than the kids straight from school who will cut code/patch OS's for minimum wage...I will, of course, be fighting from the availability/security/scaleability/clustering/support/good-over-evil angle.

As I said, I can fight the fud, but just want some ammo to put out any last lingering hot spots that might flare up!!

Let me now if you hear of new VMS customers or other horror stories.

Thanks for your help, folks, I will let you know how the good fight goes (or ask you for job vacancies!!)

Keep the faith!!
Have fun,

Peejay
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
If it can't be done with a VT220, who needs it?
Willem Grooters
Honored Contributor

Re: VMS vs. U*ix - any horror stories?

Paul


From what I can fathom, the attempt will be made to show that VMS resources are scarce and limited to old wrinklies like me, who attract a reasonable hourly rate,...


Indeed a problem but do not despair: about two years ago, I have been involved in educating and training 10 YOUNG VMS system administrators. (VMS system administration is core business of our company).
Somewhat longer ago, when I set up a VMS system, a Windows-oriented collegue was sent to a VMS user course and he came back very enthousiastic. He is now a convinced VMS admin.

...

... rather than the kids straight from school who will cut code/patch OS's for minimum wage...


... but often it's forgotton what the cost will be in case there is something wrong.

IMHO, most of these "programmers" simply don't have a clue of real programming, let alone good programming practices. They just learned to do the trick.
Never heard of change control, version management, co-existence....All things YOU would care about.

Your task would be to coach these youngsters and learn them the real thing. That's an investment that will pay back. Regardless the OS.
Willem Grooters
OpenVMS Developer & System Manager
Jan van den Ende
Honored Contributor

Re: VMS vs. U*ix - any horror stories?

Paul,


... rather than the kids straight from school who will cut code/patch OS's for minimum wage...


I coundn't agree more with Willem.

I even have one other 'point of professionalism'.
From some of my other postings you can read that in my present environment (Amsterdam Police) there is also a big political pressure to "standardise (!) " to N*IX (btw, NIX & NIKS pronounce the same, and in Dutch means: Nothing) for backoffice & M$ for frontend. Pressure mostly at the overlaying countrywide coordination & application development.

EVERY new development, when reaching us for pilot runs EACH TIME AGAIN proves they have NOT got the slightest idea of scaling ean concurrency. YES, the app HAS been 'tested' for multi-user use. Usually meaning that several users CAN access the app simultaniously. And if each of them locks 'something vital' for about 5-10% of the time, it runs just fine with 5 users. And if that uses 5 % CPU, you extrapolate 100 users need 10 times the test-ssytems CPU power, and 'everything will run fine'. So after a disastrous pilot with 25 users I drastic redesign considering locking has to be done. And then all tests are done on a db with 1 - 10 K records. And for some reason nobody had imagined that this kind of work doesn't scale well.... which is found out after (a subset of) the production data, say 1 or 10 M records is loaded. Only then someone starts extrapolation to 100+ M records.
Well, it IS good for my own job security:
10 years ago I was hired for 6 months, with 3 more of option) "to help VMS keep alive in the final days while the apps get ported to Unix." The options were "a bit" extended, and now "we are in the final days of VMS, everything is re-develloped for Unix or Windooze". And the horizon has remained steady at "1 year, two at the most".

And if, for legal reasons, something new MUST me done, then chanches are that one of the older develloppers 'quickly' uses some old skill, and a new VMS app is born.

fwiw,


Jan
Don't rust yours pelled jacker to fine doll missed aches.
Anton van Ruitenbeek
Trusted Contributor

Re: VMS vs. U*ix - any horror stories?

Paul,

...

... rather than the kids straight from school who will cut code/patch OS's for minimum wage...


After DEFCON 9 was ended we mentioned this at some students who where here during there training period working at a client of mine. At that time we also needed to get rid of some old VAXes we had phased out. We add VMS (and latest patches) on it and these student went home with those machines. The most of these students had implemented these machines as common file server in there student appartments and where wild about the operating system.

So there is hope .......

I hope also in the long strugle against NIX.

Anton.
NL: Meten is weten, maar je moet weten hoe te meten! - UK: Measuremets is knowledge, but you need to know how to measure !
Brad McCusker
Respected Contributor
Solution

Re: VMS vs. U*ix - any horror stories?

Paul,


From what I can fathom, the attempt will be made to show that VMS resources are scarce and limited to old wrinklies like me, who attract a reasonable hourly rate, rather than the kids straight from school who will cut code/patch OS's for minimum wage...


We're really working hard to address this.

The UNIX Portability effort has as its primary goal to get OpenVMS to look and feel like *NIX from a _programmers_ perspective. A natural side benefit will be similar look and feel for the system managers and operators. In other words, we'll be making as much of the UNIX shell and utilities available on VMS as we can.

Our UNIX Portability effort is detailed here: http://h71000.www7.hp.com/portability/index.html

GNV includes a bash shell on VMS (currently mainatined ans supported by VMS engineers as part of the Open Source community): http://gnv.sourceforge.net/

At the fall bootcamp, I had a hands on session that allowed the users to use GNV and some of the recent enhancements to VMS to port some OpenSource applications. Some of the students were VMS newbies (UNIX pros), and they were all suprised at just how well the GNV tools worked on VMS. Yes, it is not perfect, yet, but, it was very manageable for them. It might be worth your while to fire up GNV and show them just how easily the UNIX types can work on VMS.

While our initial focus is on programmers, not managers, the further we advance with the programming interfaces, the easier it is to get the management tools running too.

Lastly, as engineering winds down the IA64 development, its getting very exciting for me to see the activity starting on VMS 8.next. Lots of work on UNIX Portability related features is getting underway.

I'll be more than happy to discuss this further offline with you (or anyone else).

Regards,

Brad McCusker
OpenVMS UNIX Portability Project Leader
OpenVMS Engineering



At the fall bootcamp I had a hands on session that attempted to show off
Brad McCusker
Software Concepts International
Willem Grooters
Honored Contributor

Re: VMS vs. U*ix - any horror stories?

Hi Brad,

I'll be more than happy to discuss this further offline with you (or anyone else).


Count me one of the rest!

Willem
Willem Grooters
OpenVMS Developer & System Manager
John Eerenberg
Valued Contributor

Re: VMS vs. U*ix - any horror stories?

Paul,

Good to hear you are working hard to help the decision makers to understand what is at risk when short cuts are taken (i.e., lower cost).
I too have been using VMS for 20+ years and worked for Digital in pre-sales/consulting.
Sounds like you are the lone voice. My rhetorical question is "where is HP in all this?" I really don't want a response since I would have to respond and do a "$ Set Flame/On." ;-)

> "From what I can fathom, the attempt will be made to show that VMS resources are scarce and limited to old wrinklies like me, who attract a reasonable hourly rate, rather than the kids straight from school who will cut code/patch OS's for minimum wage...I will, of course"

IMHO you shouldn't back down from the wage cost argument since a lot of VMS people now have other OS experience but that doesn't mean a number of them would not go back to VMS.

Which is better? A) Training "kids straight from school" to become professionals. or B) Employing professionals to start with.

I think if management wants a system that is cheap and fast but not "mission critical" in orientation (i.e., clusters, security, blah, blah, . . .), then generic unix is fine and I would't argue the point very much. If management's goal is to have a fast and solid system, there is no cheap for any OS and they need to get a grip.

In this response I am dealing with reality. The reality is that taking an OS, any OS, from shrink-wrap to mission critical (or a level approaching), is costly; better to start with VMS in that case. If approaching a mission critical level of operation is not needed . . . you always have the stability, security, etc. points in your favor. :-)

John
It is better to STQ then LDQ
Martin P.J. Zinser
Honored Contributor

Re: VMS vs. U*ix - any horror stories?

Hello Paul,

unfortunatly I do not have this in a publishable form, but anyhow for your entertainment:

Recently our company was evaluated by one of the big consultancy firms to determine how to increase efficency. One of their results was that our department doing all the VMS stuff (plus some other work) gets x% more salary than average, but also provides x% more productivity per employee ;-)

Still I do concurr (and have tried to point this out several times to hp already) that getting VMS back into Edu is an important thing and should not be measured by immediate profit made by HW sales!

Greetings, Martin
Ian Miller.
Honored Contributor

Re: VMS vs. U*ix - any horror stories?

The unix portability program is vital especailly if hp then use it to get more apps ported to VMS and shorten the time delay for VMS versions of oracle etc. However it is also important to show new people the 'VMS way' of doing this. VMS is properly designed from the start and this makes a huge different when programming and using VMS - the consistance of API's, the simplicity and consistancy of DCL, the insistance on backward compatability (does that VMS V1 program still run on the latest VMS), etc.
____________________
Purely Personal Opinion