11 vi2 and 11 vi3

[khilari]

Hi guys, i wanted to ask that if i am running on 11 vi why would i want to switch to 11 vi2 or 11vi3......

[Johnson Punniyalingam]

EOS -> End of Support

[Patrick Wallek]

Much more functionality! 11iv3 has a significantly better i/o stack. Native multipathing, persistent device names, better performance.

Read the release notes for each release to make your own determination.

However, this presumes that your application(s) are supported on the newer releases.

Also be aware the v2 and v3 may not necessarily run on your hardware. Be sure to check that before contemplating an upgrade.

[Michael Steele_2]

Hi

11.11 will be supported til 2013.

11.23 is transitional the way 10.30 was transitional.

11.31 has many differences

PA-risc is supported in a minor way. Many applications are forced to use a Pa-Risc emmulator called Aries.

Kernel Parm.s are very different

Native Multipathing is used instead of PV links

SD-UX goes away and SWA is now the patching method.

IA64 virtual machines are very different from vpars and npars.

Because you have virtual machines crash dumps are different.

You have the ability to upgrade some servers from Pa-Risc to Ia64. I.e, rp8420 to rx8620.

Note: I've been pretty steady into 11.31 for 6 months now and still making discoverys.

[Michael Steele_2]

Hi again

diagnostics are very different in 11.31, somewhat different in 11.23.

[Michael Steele_2]

Hello again

Here are the Sept 2009 release notes for 11.31, refer to the "What's New" sections.

[Steven Schweda]

> PA-risc is supported in a minor way.

Huh? Minor? What's minor about it?

> Many applications are forced to use a
> Pa-Risc emmulator called Aries.

Well, yeah, _if_ you replace your PA-RISC
system with an IA64 system. Are you saying
that that's required?

[Wim Rombauts]

Maybe, let's revert the question.
Why stay on 11i v1 ?
-> If your server and the applications on it are running fine, and there are no new implementations at hand, there is no real reason to upgrade.
Yes, you will have newer and better technology in your hands when you upgrade, but if everything is running fine now, you don't really need it. The year 2013 (EOS for 11i v1) is still a lot of year away.

But, if you are planning to upgrade the hardware, if you are planning a major upgrade of the software, if you need (or want)to implement new things that may be easier to do on HP-UX 11i v3, or not possible on 1&i v1, yes, then upgrade. It will be worth it.

[Dennis Handly]

> Michael: 11.23 is transitional the way 10.30 was transitional.

Not sure what you mean by that?
10.30 was for special customers. 11.23 is for general release.

[Michael Steele_2]

Dennis,

What was the same between 10.20 and 10.30? They were both transitional.

What was the difference? Probably the extra telephone network stuff like X.25.

When at ATT, we used 10.30 on maybe 3,000 servers.

Steven,

Sigh. You really need to work on your tedious attitude. Maybe you don't realize it, but its a personal problem that you were recently called on by another forum user. So I take that back, you are aware of it but do it anyway, for here you are again doing the same thing in what, only 7 days?

11.31 was orginally marked exclusive to Itanium. But like many things in 11.31, that changed. Hyperthreading for example. First advertised as supported in 2008, later advertised as not supported in 2009. Probably a lot of Pa-Risc people objected. I only read the release notes.

[Dennis Handly]

>Michael: What was the same between 10.20 and 10.30? They were both transitional.

No, only 10.30 was. (If you use the English definition of transitional, every OS version is. I assume you meant the music one, brief.)

>What was the difference?

10.30 had a limited distribution. Same as 10.10, 11.20 and 11.22.

>When at ATT, we used 10.30 on maybe 3,000 servers.

Then you were special.

>11.31 was originally marked exclusive to Integrity.

11.23 was originally for Integrity with PA support added a year later.

[Steven Schweda]

> [...] You really need to work on your
> tedious attitude. [...]

And some people should work on the accuracy
of the information they supply.

> [...] its a personal problem [...]

"it's". Post less nonsense, and my "problem"
will disappear, as if by magic.

> 11.31 was orginally [...]

So, what's true now? What you said, or
something different?

[Michael Steele_2]

Dennis

Obviously you weren't around for HP-UX 8 or 9 when HP-UX was Berkeley based. Then begin the transition into ATT based 10.00 which was complete until 11.00.

Steven

You are so predictable.

[Michael Steele_2]

"....The HP C/ANSI C compiler installation package provides the capability
to create and remove transition links from HP-UX release 9.x locations to
HP-UX 10.x locations...."

There are several example of left over BSD 9.0 versions of HP-UX in 10.20, and software still be written to make the move into ATT HP-UX. Page 14

http://docs.hp.com/en/5967-0005/5967-0005.pdf

BTW Dennis, you are the first HP-UX admin I've ever run into who didn't state 10.20 was transitional.

[Olivier Masse]

> BTW Dennis, you are the first HP-UX admin
> I've ever run into who didn't state 10.20
> was transitional.

Then I'll be the second one. It was not. But that depends on your definition of "transitional". If you're strictly speaking about a transition from BSD to SysV, then you're right, but 10.0 was even more major in that respect.

It is my opinion that BSD/SysV are details that will be irrelevant to a high-level architect who is just looking into deploying an infrastructure where millions in expenses are involved. That's the guy HP markets HP-UX to, and under that perspective, I think that HP's definition of "transitional" is one where the OS has been released to cater a specific technology need and will not be supported for long.

For example, AFAIK, 10.30 and many of the initial 11.1x Itanium versions were released as "transitional", under that definition. There were also special versions for the V class which I don't remember.

Olivier

[Michael Steele_2]

"...10.30 and many of the initial 11.1x Itanium versions were released as "transitional", under that definition. ..."

Geez, you contradict yourself just jump into a pissing match.

This is real bore and typical of techies who refuse to back down and fight to the death rather than acknowledge a peer.

[Emil Velez]

IF your applications are running ok and you can get patches for your application and OS everything is ok.

If your application is no longer supported at 11.11 and you need a newer version of your applications that will require you to move to 11iv2 or 11iv3 then upgrade.


IF you need to replace the application running on 11.11 with a new application and that application runs on 11.31 or 11.23 then you need to upgrade.

You are probably running a system with limited amount of memory so 11.23 will require more memory.

This sounds like a redneck joke. You know you need to upgrade when .... (I live in georgia)

[Hakki Aydin Ucar]

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1311383

[Dennis Handly]

>Michael: Obviously you weren't around for HP-UX 8 or 9 when HP-UX was Berkeley based. Then begin the transition into ATT based 10.00 which was complete until 11.00.

I was around but then HP-UX was an appliance to me.

>BTW Dennis, you are the first HP-UX admin I've ever run into who didn't state 10.20 was transitional.

You are very confused. I never said I was an admin, I just play one on TV. I'm probably an admin's worst nightmare, I'm R&D. :-)

[Michael Steele_2]

Geez, you only confirm how bad HP support and service have become since the Compaq merger.

[melvyn burnard]

Gentlemen. You all need to take a step back, a deep breath, and consider how this thread may reflect to other users.
There are a number of statements made by various people that erroneous/misguided/uninformed.
I have been with HP-UX since the first PA RISC server (9000/840) back in 1985/86, as a hardware engineer, a software engineer, a Response Centre engineer, and now a Technical Consultant and have seen a VAST number of changes.
I am not going to comment on any particular point, other than to again suggest you reconsider how you respond to threads.
We are all human, and we can all "get " something to appear different to someone else.
I also think that I will edit this thread to remove various comments once it is closed, unless anyone objects.

[Michael Steele_2]

No objections.

I've already written a complaint into the itrc about it.

The complaint is about Steven Sweda's constant nasty, snied remarks, and how Dennis has objected to my objections to Steven poor attitude. In my opinion Dennes failed to do proper research into my objections.

No matter.

This is not a mental health blog for Steven Sweda to dump upon.

In my complaint I've asked that Steven be limited to technical citations and references and procedures only, in order to have yet another one of these, the 3rd in 10 days.

[Patrick Wallek]

Thank you Melvyn....I think you said what a LOT of people were thinking!

[Kevin_Paul]

I would like to remind everyone that while we encourage open dialog/conversations here in the ITRC forums, we *do* still have guidelines we ask everyone to abide by to when participating in this community.

Please take a look at the FAQ entry "Are there rules of conduct I must adhere to?" @ http://forums13.itrc.hp.com/service/forums/helptips.do?admit=109447627+1258555346710+28353475#9 within our forums overview.

We'd prefer not to have edit/delete posts, but will if necessary - please refrain from activity that others may construe as abusive or a personal attack.

Thank you for your consideration.

[Steven Schweda]

> [...] please refrain from activity that
> others may construe as abusive or a
> personal attack.

This can be difficult when a simple question
of fact may be construed as nasty and/or
"snied". It's easier to control one's own
postings than it is their interpretation by
others.