Submitted by Sun’s CEO Scott McNealy has published an open letter to HP, proposing a merger of their respective UNIX platforms. In the past, Sun’s top execs have been highly critical of HP-UX’s future. Sun’s president Jonathan Schwartz has repeatedly referred to the ‘demise’ and its troubled commitment to Intel’s Itanium chip. Now, it seems, they’ve hit on a new idea. “We propose an alternative,” writes McNealy in the letter, “that Sun and HP commit to converge HP-UX with Sun’s flagship volume UNIX, Solaris 10.’
Can you really merge two operating systems? Even if they are both Unices, it still seems like a hard task. I’m clueless about OS development but it doesn’t seem like very smart idea.
No. You really can’t.
Good point.
The kernel internals alone represent a nearly insurmountable obstacle to actually mixing code.
Userland is less problematic, I suppose, but still presents a plethora of obstacles.
Sun definitely has posturing down to a ‘T’. But having open-sourced Solaris, they haved gained some credibility.
Not to sound like too much of a Linux fan (even though I am 🙂 ) I feel that Linux is forcing the various Unix camps to cooperate. After all, HP needs to do *something* with HP/UX. And “legacy cash cow” is not all that prestigious a label.
Maybe they can agree on some interoperability standards and actually stick to them this time?
Time will tell, I suppose. 🙂
Edited 2006-03-04 20:02
Sounds good since HP was always more closely tied to the customer. In the past HP OS was based off of DOS I think so now maybe they should base themselves off of Solaris like IBM is with Linux. This is more options unlike everything being based off of closed DOS/UNIX in the past.
The cool thing about GPL is it’s a contract and more transparent then one strict OS.
Edited 2006-03-04 18:08
uh… wtf?
Sun has some serious issues with their direction. First they badmouth Linux, then they support it, now they tell you how many ways Solaris is better than Linux. Solaris is a great product and has some great features, but in many areas, Linux still beats it. The exact same can be said for HP-UX.
Some really nice HP-UX features: VPARS (Virtualization), EAL4 Role Based Access Control, Stack based buffer overflow detection, and more. Additionally, the HP performance tools like the glance plus pack and perfview are far better than anything I’ve seen when working on Solaris systems. Sure Solaris has Dtrace and thats a great capability, but glance is faster and more intuitive than dtrace.
Sun keeps talking about how HP-UX is dead touting OpenSolaris. The funny thing is that HP continues adding new features and updating HP-UX they just don’t market it so agressively.The truth is in the numbers and IDC reports that HP holds the # 1 position for high-end UNIX revenue and units on a worldwide basis for the calendar year 2005 and Q405.[1]
The funny thing is that we are still moving many of our older PA-RISC HP-UX servers over to quad opteron Linux boxes in the near future. Java runs so much faster on Linux it is sick.
[1] IDC’s Worldwide Quarterly Server Tracker, February 2006.
Edited 2006-03-04 18:19
Why is it whenever there is an article that involves UNIX we always have to hear about how someone has done Linux wrong lately. Linux hasn’t a damn thing to do with this article or discussion. If you want to run Java on Linux, knock yourself out. Nobody really cares.
“Some really nice HP-UX features: VPARS (Virtualization), EAL4 Role Based Access Control, Stack based buffer overflow detection, and more. Additionally, the HP performance tools like the glance plus pack and perfview are far better than anything I’ve seen when working on Solaris systems. Sure Solaris has Dtrace and thats a great capability, but glance is faster and more intuitive than dtrace.”
VPARS may be nice, but limited to HP’s hardware, if you want to look at hardware limited choices, take a look at domains, availible on Sun’s larger boxes. Zones on the other hand are all done in software and will run on both ultrasparc and x86, with work on ppc speeding along. Solaris has RBACs and will soon be releasing a Trusted expansion add-on that meets the EAL4 and other security specs, this will be accross the board, not just on Sun’s own hardware.
Glance at first glance looks very limited, something that a modified version of vmstat, iostat and prstat could be made to emulate. Dtrace on the other hand is alot more powerful and flexible.
If HP is so opensource/GPL friendly why haven’t they offered to port any of this to linux? Oh wait HP is only friendly when it comes to printer support( well not really, they still haven’t released drivers to all there printers), oh yeah they are only opensource friendly at opensource conventions, well they say they are anyway.
Sun keeps talking about how HP-UX is dead touting OpenSolaris. The funny thing is that HP continues adding new features and updating HP-UX they just don’t market it so agressively.The truth is in the numbers and IDC reports that HP holds the # 1 position for high-end UNIX revenue and units on a worldwide basis for the calendar year 2005 and Q405.[1]
Well i guess, if you charge for your UNIX, you make more money selling it. Perhaps HP is a fraid that if they make alot of noise, about its new features the Linux croud will duplicate them or perhaps they are worried that crackers will go fishing and find that they are full of holes, when i did my google search on “HP glance” two thrids of the links were about exploits and security holes.
The funny thing is that we are still moving many of our older PA-RISC HP-UX servers over to quad opteron Linux boxes in the near future. Java runs so much faster on Linux it is sick.
Just a note, Java runs faster on Solaris then it does in Linux.
When is Xen being ported to HPUX? I guess they won’t be running any Linux apps in HPUX anytime soon. Does Linux even run on pa-risc? or any of HP’s other custom chips?
“Does Linux even run on pa-risc?”
Yes, it was ported years ago by Alan Cox. Now, are there a significant number of people running Linux on PA-RISC machines? I doubt it, the same can be said for Linux on SPARC.
Right now x86(-64) has all but defeated all other architectures, and people don’t buy SPARC or PA-RISC iron to run Linux, they buy it to run Solaris and HP-UX, because they have apps that only run there.
So, the issue is mute. Linux/x86(-64) fights Solaris/SPARC and HP-UX/PA-RISC on new installations to run new services, and Linux/Itanium fights Solaris/SPARC and HP-UX/PA-RISC on big-iron.
Sun fights Linux with Solaris/x86 and is watching its SPARC big-iron market going smaller. On the other hand, HP embraces Linux/x86 and has HP-UX on the Itanium as its the big-iron offer (not that Linux isn’t good there, but it doesn’t have the mindshare yet).
Guess who is better positioned to survive the Linux onslaught… McNealy is just trying to pull HP into the Solaris/x86 camp with this bogus proposals. But I don’t see HP going for it.
Besides, this is just ludicrous. HP had people working on merging stuff from Tru64 into HP-UX, and ended up axing these efforts. It just didn’t made any sense… The proprietary unixes are mainly in the business of supporting legacy stuff, and are being quickly eroded by Linux and Windows because of this. HP knows this, and that’s the reason why they don’t do aggressive marketing on HP-UX (it keeps improving, but just enough to avoid becoming stale, and that’s the kind of budget HP is willing to assign to it).
Why do you think VPARS are hardware limited…. You may be thinking of Npars…
Albiet I am not an HPUX expert…
Here is the difference between nPars and vPars:
http://poeticgeek.net/archive/2005/02/02/hp-ux-partitioning-continu…
Wow, this shows how much you know what you are talking about:
http://h21007.www2.hp.com/dspp/tech/tech_TechSoftwareDetailPage_IDX…
The GlancePlus Pack has ran on linux for awhile. We run glance on our SLES Linux servers in production.
Just a note, with the Jrockit JVM on Linux and using the gencon garbage collector, we get better performance than with the Sun JVM on Solaris. I work on some of our Weblogic boxes.
Edited 2006-03-04 23:19
HP might be adding features to HP-UX, but if you are a PA-RISC user that doesn’t matter unless you intend to migrate to Itanium. HP is pretty arrogant thinking all of their Enterprise customers are going to go happily to Itanium and eat all of the costs associated with it. And if I was one of those customers, I’d be pretty pissed off, especially if I owned a SuperDome that could not be converted to Itanium.
In terms of the features comparison, have you actually used Solaris or is this based on what you have read?
While Sun’s PR move takes balls, it does allow for HP to sell x86 hardware and potentially make a sale that they otherwise might not be able to make at all.
The funny thing is that we are still moving many of our older PA-RISC HP-UX servers over to quad opteron Linux boxes in the near future. Java runs so much faster on Linux it is sick.
Usual Linux FUD in my opinion. Java runs faster on Opteron, OS probably contributes +/- 10% at best. Even if it was a PA-8900, even Windows on Opteron would be faster per core.
Always amused to hear the old ‘We moved from a 75 MHz SPARCstation 5 to Linux and Xeon and it’s 10 times faster, Sun sucks!’
Just a note, with the Jrockit JVM on Linux and using the gencon garbage collector, we get better performance than with the Sun JVM on Solaris. I work on some of our Weblogic boxes.
Hardware?
Edited 2006-03-04 23:37
Well I tested it on an HP DL360 G4 (x86) because thats the only machine I had laying around to throw opensolaris on. Because that is the only hardware I have that runs Linux + Solaris. Granted… Jrockit is available for Solaris, but it seems to run better on Linux.
Well I tested it on an HP DL360 G4 (x86) because thats the only machine I had laying around to throw opensolaris on. Because that is the only hardware I have that runs Linux + Solaris. Granted… Jrockit is available for Solaris, but it seems to run better on Linux.
Just curious, was it actual OpenSolaris or the official Sun Solaris 10 build? Interested because I haven’t heard of anyone trying to use an OpenSolaris distro for production so far. Not that I know if it is suitable or not–have not had the time or inclination to try it, so far. But it does seem sort of ‘beta’.
Of course, at the ASP I work at, one of the things that constantly annoys me is customers using Fedora for production–for the same ‘beta’ reason, I just don’t get it. If I had/wanted to run Linux for free, CentOS or another RHEL build seems to make much more sense.
Thats a very valid point on Solaris vs OpenSolaris.
My company phased out the last of their Solaris boxes long before I started working for them an I just used it as a proof of concept for the management. According to our testing, Linux is the best platform for our use case.
>In the past HP OS was based off of DOS
WTF are you smoking…. you could not be more wrong!
Jesus H Farking Christ. There is no DOS in UNIX?
You need a serious education on OS history/reality.
This is the most amazing comment I have ever read in all the years I have been coming here.
While I agree, that Unix needs more combined firepower these days, this seems to be nothing but a publicity stunt.
Firstly, a merger of the two unices would probably not be dealt with in the press.
Secondly, the letter also seems a bit patronizing in the way that Scott McNealy touts the momentum of the Solaris platform, in comparison to HP-UX. Its probably true, but it sends a message that HP-UX should be happy to be assimilated, rather than merged with Solaris. Not a great way to start a relationship.
Anyway, I don´t think this is gonna happen. Its just another of those ‘hot-air balloons’ that Scott McNealy sends up so often.
Will the real Sun strategist please stand up….
As a current HP employee (not much longer though). I seriously doubt HP is taking this even seriously. Of course with hurd in charge maybe the C levels are considering this. Internally I am seeing a greater push for LinuxWindows and I keep hearing about OS devs (tru64,vms,HP-UX) getting replaced by indian coders. Of course what better way to cut costs then by outsourcing the whole OS. It wouldn’t surprise me to find out if we had ports of HP-UX for x86. Maybe merge the two on x86 to fight off linux, however HP is very linux friendly and have close ties to RH and Suse…
“HP has yet to respond, but at least one potential hurdle is that Solaris is open-source while HP-UX is not.
However, the ongoing SCO dispute gives hints that this might easily be overcome. Both companies were found by SCO not to have misappropriated Unix IP, even though Sun made Solaris open-source. Perhaps SCO would look equally kindly on HP should it do the same.”
It’s important to learn the lesson of keeping your mouth shut, especially if you’re engaged in a lawsuit. Cause anything and everything you say can come back to bite you on the ass.
As for merging, I think HP would gain little and SUN would gain a lot. But it would be an opprotunity for HP to make a ton of brownie points with the community, being GPL friendly as HP is, and pressured SUN into at least dual licensing Solaris CCDL and GPL as condition of the merger.
again shall we?
For those that don’t know, Sun had Openstep natively ported, but the politics between executives buried the partnership.
This albeit a different beast is an interesting idea. Consolidate to grow.
Sun fights Linux with Solaris/x86 and is watching its SPARC big-iron market going smaller.
The big iron market is shrinking across the board in terms of units.
However the SPARC market is growing in terms of volume. Sun is shipping more SPARC units than ever before with big increases in the 1-8way space. HP & IBM compete just don’t compete in this market properly preferring instead to push with Linux/x86.
This is actually what I like about Sun & Solaris. I can scale from 1 to oodles of CPUs just by adding 0’s to the cheque. No hassles with different OS architectures for differing workloads. HP instead says Linux for low end & HP-UX for highend.
Sun is already working on their “Linux” environment where it will run native Linux applications on top of Solaris. Conceptually, Sun and HP could create a similar system that will allow legacy HP-UX SOURCE code be compiled and run on to a SPARC/x86 Solaris box and still run against HP-UX libriaries and even, at least to some extent, kernel calls.
Add in a /usr/hpux/bin for a userland, and a ported SAM for administration, and you might well be able to have a decent transitory environment for HP SA’s, and for HP specific products to port to.
I think what is interesting here is how Sun is positioning Solaris across the spectrum, from low end white box hardware up to huge Fujitsu mega processors.
Meanwhile, HP is push HP-UX to the high end (where I suppse the margins are), while Linux to the low end.
Sun’s already tried that kind of segragation and didn’t have much luck it, perhaps HP will do better.
But by having an HP-UX compatable system on the edge blade servers as well as in the back end data centers you would like to think is interesting to some markets.
It’s basically an issue of whether the passion for HP-UX is going to hold over to the Itanium servers or not (or, whether the Itaniums will attact folks to HP-UX).
No self resepcting HP-UX SA uses SAM…
It screws up LVM for one…
OpenSourcing HP-UX and GPL RTL for the PA-RISC. Since it is not happening I suspect that SUN is trying to push yet another paltform, Solaris on Itanium. Good for marketing purposes but they already have the amazing Sparcs. It seems that they could implement some of the kernel space APIs of HPUX in Solaris and port it on Itanium. This way HP will sell their hardware with foreign software (they have done it with Win$$$ and Linux). It shows the inability for HP to keep a Unix and push a hardware platform. Good for SUN, good for HP CEOs but bad for HP as a company. Just another story of lazy braindead company. On the contrary SUN is not so lazy.
I think people forget is this; There has already been an Itanium port in the past; if HP is hell bent on using Itanium, and quite frankly, isn’t too bad of a chip, then why not standardise on the most popular commercial UNIX out there?
The simple fact is, part of the process to encourage porting of applications to your platform is not only to offer incentives, but once on the platform, make it as easy as possible – considering how wide spread Solaris is, many times it should simply be a recompile away and voila, HP would have access to same software that SUN/SPARC/x86-64 has.
What I think people fail to realise is this; the greatest threat to HP and SUN isn’t Linux, but IBM – unlike SUN and HP, IBM can do *VERY* deep discounts, to the point of making an isance loss off a server sale, then make that loss up through services they sell – HP and SUN on the other hand don’t have that same luxary of having a large services group to fall back on.
Atleast if they work together on the one platform, it’ll cut costs over all and allow them to compete effectively not only with IBM, but also with Microsoft and probably to a lesser degree, Linux; personally, if it were me and in HP/SUN position, I owuld also buy out SCO, and opensource their whole stack, take the interesting parts out of their products and merge them.
Basically what you would have is a commercial opensource UNIX in one corner supported by SUN and HP, and AIX being supported by IBM.
Actually HP has a very large services brach. It’sjust that we aren’t as expensive as IBM. I have literally seen millions of dollars of equipment given away in deals too
Well i have a proliant ML570 and solaris 10 wont work (last time i tried) HP may be better of doing the IBM thing and moving to linux which is even faster development than Solaris. How would the money get spread around? Looks like sun is trying realy hard to get other companies onboard to help with development. Once iscsi target is intergrated into opensolaris will foster more storage OEM’s.
Solaris hasnt even got USB down pat.
Edited 2006-03-05 00:54
Which generation ML570 is it? Here is the matrix from HP on Solaris support for their x86 hardware:
http://h10018.www1.hp.com/wwsolutions/solaris/index-all.html
I would also be careful about the drivers on their site, I ended up downloading drivers for the bcme interfaces on the 4 DL360’s I was working on from Broadcom. The HP drivers did not have the script that removed the bge to bcme mapping and the end result was using the HP drivers the network interfaces would not work at all.
though this seems to just be marketing ploy for attention i wouldnt mind seeing Solaris oficialy suported by sun on Itanium systems. but it wouldbe a little counter productive being that Itanuim(like POWER) is a direct compeditor of SPRAC. but merging the 2 or what would mostly likely happen is solaris takes the technology it wants out of HP’s os and then they rebrnd the combination “solaris” (and jsut whipe out HP-UX) and then SUNonly has to compete with AIX and Linux on the high end sever world. oh ya and windows…..slight chukle
What is clearly HP-UX’s greatest strength?
What is clearly Solaris’s greatest strength?
I would like to know. They must be different beasts.
>What is clearly HP-UX’s greatest strength?
I think there are two main ones that HPUX has always spanked Solaris:
1) LVM – HPUX had the best Logical Volume Manager of all. In fact, when it was time to build LVM for Linux the overwheling consensus was to pretty much copy command-by-command, function-by-function the LVM in HPUX. I clearly remember the discussion on USENET, LKM about it. HP’s LVM is soo good that even though at one point HP started including Veritas Volume Manager for free with HPUX, NOBODY USED IT, becase the HPUX LVM is so much better and easy to use.
2) HA Clustering. HP’s MC/ServiceGuard is the BEST HA clustering failover software there is. I have experience with them all, SUN Cluster, IBM HACMP, Veritas(on sun), and MC/SG. MC/SG is ROCK SOLID and easy to work with.
These 2 products are clearly the jewels left over from when HP actually had some seriously smart engineers working in the garage out back. But, alas those days(and it appears those people) are long gone at HP.
There are other cool things that HP’s people did in their implmentation of UNIX such as the rc.config.d system which takes all the configuration data out of the RC scripts and makes it VERY logical to configure things like networking, nfs, nis etc.
While Solaris is always said to be the “MOST POPULAR” UNIX, that is because of all the workstations they sell. In my experience, HPUX is more popluar for pure servers in the business world doing things like Oracle. HPUX is a multi-year uptime OS thats fore sure, it has proven it too me over the years with seeing many heavly loaded servers that chug away for 2-3 years without a single minute of downtime.
Traditionally HP built their hardware like battleships too, they used to put like 2 layers of steel in their systems. Picture an outer steel sheel and all the “guts” were in a whole inner steel shell that fit inside that outer steel shell.
Sun on the other had, didn’t even discover what rack mounting was until the late 90’s when the Enterprise series came out. They were a desktop pizza boxes, or deskside type boxes only company for a long time that aways made me consider them to be not really “business class” for a long time. But SUN’s new hardware is quite nice.. I love the V880s and our new V890’s.
>What is clearly Solaris’s greatest strength?
I LOVE Solaris too, but the one thing that always sort of bugged me was the fact that Sun never seemd to “PUSH” Solaris. Until Solaris 10, it felt pretty much like it was STUCK at Solaris 2.3 circa 1995.
It got faster and more reliable but the “features” and implementation stayed pretty much generic UNIX and capable for a long time…not much innovation in Solaris until just recently. I found it very boring, and somewhat lacking in any having any orginal thoughts put into it myself. But, I have no problem depending on it when things are on the line.
Edited 2006-03-05 02:04
I LOVE Solaris too, but the one thing that always sort of bugged me was the fact that Sun never seemd to “PUSH” Solaris. Until Solaris 10, it felt pretty much like it was STUCK at Solaris 2.3 circa 1995.
It got faster and more reliable but the “features” and implementation stayed pretty much generic UNIX and capable for a long time…not much innovation in Solaris until just recently. I found it very boring, and somewhat lacking in any having any orginal thoughts put into it myself. But, I have no problem depending on it when things are on the line.
That has more to do with the managements view that the crown jewels of SUN was the hardware, not the software; when they were hit by the dot-com bust followed by the rise of Linux, they realised that what made their hardware tick was the software, the operating system itself.
From around version 9 onwards, there has been a greater focus on evangelising Solaris and its features as hardware these days, minus the few remaining niche areas, has pretty much been made a generic thing where there is really hardly anything that differentiates the various platforms out there, hence the massive price cuts of UltraSPARC hardware to bring the prices down to something that actually ressembles the market reality of today.
What SUN need to do is this; they want to push the concept of SUN Ray appliances as the successor to PC’s, and I’m sure with volume, they could get the price down to $200 per unit, but it won’t be achieved until they turn Solaris into something that resembles a well thought out desktop, right now, JDS, quite frankly, is shithouse, the problem is made worse, when people like me point out the failins in Solaris, and the best SUN engineers and management can do is send abusive emails claiming I’m the one with the problem.
Thank you.
If HP had half a brain, they should have considered the offer seriously. To be quite honest HP-UX is in a quite pathetic position at the moment — the OS smells really old and is hopelessly behind Solaris 10. I can say because I’m managing a bunch of low end PA-RISC machines and quite honestly I’m really itching to get those boxes booted out and replaced with Solaris/SPARC at the upcoming refresh. It is not that I hate HP-UX without any particular reason, but rather because Solaris 10 has so much more to offer. HP-UX has absolutely nothing interesting in the pipeline for years and the forced migration to Itanic just freaks me out. HP could on the other hand kill a few birds with one stone just taking OpenSolaris as a foundation for HP-UX 12 and instantly improve their OS. Simply the fact that HP-UX would run on x86 would be worth it. If HP does not diversity HP-UX from Itanic, their entire enterprise strategy will go down alone with the Itanic. Itanic has been pretty much written off as a chip that has any future whatsoever, so HP holding on to it like it is the only way forward is just stupid.
If HP had half a brain, they should have considered the offer seriously.
And when they are at it, they should probably press “self destruct” button too? Just as smart move as your comment, when you take Sun PR/OS moves in question. I’m specificaly reffering to their deeds and comments about Linux.
It is not that I hate HP-UX without any particular reason, but rather because Solaris 10 has so much more to offer.
Every OS has something special to offer. Personaly, I value Linux higher for some specific reasons tied to my bussines.
Simply the fact that HP-UX would run on x86 would be worth it.
And by gaining poor selection of x86 hardware as Solaris supports? They would gain what exactly? I can tell you what they would loose.
Sorry to burst your bubble, but even my Opteron is not fully supported under Solaris10. Few friends of mine simply just returned to Linux without questions asked for the same reason.
/*my personal opinion*/
This is nothing but another PR scam Sun produces. Linux love/hate mood days are so common in the last years. But, hey Linux got a friend here, it’s HP-UXs turn now.
What Sun is interested here is probably just to be able to produce something like brandz (in linux case, in my case I’m just as interested in brandz as in running Win apps on wine, zilch, zero, not) to do the same for HP-UX. After that, we all we can expect are “Solaris is better HP-UX than HP-UX” claims.
And what does Sun’s comments about Linux have to do with this topic? This isn’t so much about Solaris being better than anything else, but an opportunity for HP to preserve some of its customer base by joining with Sun. Now whether they do so or not is up to them and their customers, and I am willing to bet the customers will be driving this train.
Your response to the comment by 0xbadbeef is ridiculous, he is talking about HP porting HP-UX to x86 which isn’t a bad idea. Why would HP short change themselves in terms of hardware support? Or is this yet another “Solaris sucks because it doesn’t support my white-box hardware” slam?
If you want Solaris x86 to work on your hardware, start by reading the HCL! It has always worked for me and I have been using Solaris x86 since Solaris 7!
>And what does Sun’s comments about Linux have to do with
>this topic?
Other than the fact that Sun’s double talk about Linux over the last few years has pissed off a large portion of the Systems Admin community who are the ones responsible for makinga and influencing these sort of decisions.
I guess Sun doesn’t realize that their waffling back and forth between being cozy and backstabbing the Linux community is actually insulting to their own core customers and decision makers.
It only seems to piss off people who based on what I have read here would probably not use Solaris at all anyhow (or haven’t). I’m sorry but I am sick of the “Sun is evil” comments because they don’t embrace Linux. It’s their choice and one that I don’t have a problem with. And the place I work at has spent plenty of money on Sun gear and plans to continue that trend, regardless of Sun’s official or unofficial position on Linux.
I see it as yet another stick the Linux zealots use to beat Sun and Solaris users over the head with every chance they get. And you think IBM and HP are embracing Linux because they want to help the F/OSS community?
It only seems to piss off people who based on what I have read here would probably not use Solaris at all anyhow (or haven’t). I’m sorry but I am sick of the “Sun is evil” comments because they don’t embrace Linux.
Ok, here you told everything about your self. And now I finally see what is wrong with your logic.
People aren’t angry because Sun doesn’t embrace Linux. Quite versa. They would probably be happier if Sun would finaly leave it alone in peace.
McNeally/Schwartz are posting love/hate change of mind in a daily fashion.
Just few pointers:
1. McNeally dressed as penguin shouting “We love Linux”
2. Not even a month later, linux sucks
3. Sun Linux Desktop. It was fiasco from the start. Who would preffer Java desktop? Most Linux users I know don’t even install java or flash on their computers. I know I don’t. And I don’t know anyone who would. But, instead of acknowledging this, they simply said “Linux sucks, Solaris rulz”
4. “Solaris is better Linux than Linux”??? Ok, this one was to much for my personal taste. In fact it was so much that I stoped Solaris part of my originaly Linux/Solaris service. Why would I do Solaris if they can run everything? No need for me to waste my time.
It’s their choice and one that I don’t have a problem with. And the place I work at has spent plenty of money on Sun gear and plans to continue that trend, regardless of Sun’s official or unofficial position on Linux.
As I said. Me and all Sun haters as you call us, don’t have a problem with it either. They should just stay on their turf and stop posting one-sided competitive bull. Problem here is that Sun tries to compete with Linux, not other way around. When was the last time you were reading Linux-sided competitive article against Solaris (be honest and exclude answers to the Sun posting the original bull). I don’t even remember one.
There are reasons to love Solaris, and there are reasons to love Linux. But postings like “brandz is better than Linux”, well it is just a non-sense. Just as it si non-sense to post runing on wine has the same quality as native on windows.
I see it as yet another stick the Linux zealots use to beat Sun and Solaris users over the head with every chance they get. And you think IBM and HP are embracing Linux because they want to help the F/OSS community?
Stick, no. Problem is that you Solaris zealots don’t get it where Sun is doing wrong.
No, IBM and HP are doing it for money, simple as cake.
Edited 2006-03-06 19:34
Here is the pertinent section of the Solaris HCL that deals with Adaptec controllers:
http://www.sun.com/bigadmin/hcl/data/sol/components/views/disk_cont…
Notice there is only one entry for a motherboard integrated controller. If your motherboard has the controller logic on the board it would be best to take it up with the board manufacturer.
Notice there is only one entry for a motherboard integrated controller. If your motherboard has the controller logic on the board it would be best to take it up with the board manufacturer.
Nope, it is a card (yes, I looked at the same page before buying machine for Solaris). I avoid using on-board controllers. Cards are simpler to move or replace.
And from their page, as I later found out
Tier Level: Reported to Work. Not Tested or 100%, reported to work.
According to my experiences with Reported like that, I just become depressed. As soon as you found out someday how vendors certify their hardware, you’ll get depressed too.
This more or less usualy means some user or 3rd party saw disk on controller, succesfuly booted and reported it works for him. I can boot too. But there is a noticeable lag in how disks perform. 3-4 times slower than anywhere else and constant errors in console. I solved this by reinstalling on IDE and avoiding SCSI. While network problem is not solved yet.
btw. how was this question linked to my refferenced post? This was a topic is not even nearly connected to my responce on your “Sun is evil” comment
And what does Sun’s comments about Linux have to do with this topic?
Did you actualy read my post or are you just acting blind in your daily zealoting manner?
In my last sentence I posted what would Sun comments be after this would happen’. At least history of Sun embrace-extend says that.
This isn’t so much about Solaris being better than anything else, but an opportunity for HP to preserve some of its customer base by joining with Sun. Now whether they do so or not is up to them and their customers
No, actualy, if they would backstab HP-UX as they do in a daily fashion to Linux community. HP-UX would only loose customers, not preserve.
And face it, most of the Sun cooperation or their embrace-extend are done by backstabbing techniques.
And I bet you live in Neverland where the happy rabbits are sharing lunch with wolves without need to be scared. Either that or you’re naive.
and I am willing to bet the customers will be driving this train.
Customers like that (the ones demanding Solaris port) would be probably better to just move on Solaris, smarter move than waiting few years to get what they wanted after who knows how many bugs. Either that or they would be stupid.
Your response to the comment by 0xbadbeef is ridiculous, he is talking about HP porting HP-UX to x86 which isn’t a bad idea. Why would HP short change themselves in terms of hardware support? Or is this yet another “Solaris sucks because it doesn’t support my white-box hardware” slam?
No, actualy the one ridiculous here are you.
My, point was simple (even simple minded should understand it). There are 99.999% of other OSs with much more x86 support. Doing a port on system that supports almost zero x86 hardware would be waste of money. First porting costs, and then after all the drivers costs and for the cake, there is a Sun hardware which competes now for the same customers. Stupid.
Yes, porting on x86 is not a bad idea. But, porting like Sun proposed is. Or, it can be called marketing suicide.
If you want Solaris x86 to work on your hardware, start by reading the HCL! It has always worked for me and I have been using Solaris x86 since Solaris 7!
Great, all I need now is you (always being so preachy) to come at my home and persuade my Adaptec controller to stop displaying errors and slowdowns (and no, it is not faulty. Works like a charm under any other OS). Yes, all my hardware is on HCL. This is the first thing I check before I install any OS. Network card just hangs sometimes for a minute and comes back (perfectly everywhere else), again it is on HCL.
Off course, I should just tell my Adaptec that errors displayed are illogical and can’t be. In the end it is listed in HCL and how could he know.
And please, face it. Solaris x86 support is slugish and mediocre at best. I know you use SPARC. Solaris was made for SPARC. No argue there. But on x86, it sucks.
Here, let me refresh your memory:
If HP had half a brain, they should have considered the offer seriously.
And when they are at it, they should probably press “self destruct” button too? Just as smart move as your comment, when you take Sun PR/OS moves in question. I’m specificaly reffering to their deeds and comments about Linux.
Simply the fact that HP-UX would run on x86 would be worth it.
And by gaining poor selection of x86 hardware as Solaris supports? They would gain what exactly? I can tell you what they would loose.
I can read just fine, obviously you can’t and by the way this is the second time I have pointed this out to you. My response to your statement is correct, you slammed 0xbadbeef’s statement about HP porting HP-UX to x86 with the typical Linux zealot nonsense about hardware support. Answer the question, why would HP short change themselves in terms of hardware support? If you are going to counter a statement, at least get it right!
If you have a problem with an Adaptec controller, have you submitted a bug to Sun and Adaptec? I would, instead of bitching about it.
Opeteron is not supported by Solaris 10, are you absolutely sure about that? That’s funny, I had Solaris 9 and 10 running on a V20z (dual Opteron) with no problems. Maybe the motherboards you and your friends are using are not supported in one way or another, I can’t say because I am not sitting at the console of your systems.
For a long time Sun has ignored x86, it is no secret. But if my memory serves correctly, Linux had a serious upward climb for hardware support as well.
Answer the question, why would HP short change themselves in terms of hardware support? If you are going to counter a statement, at least get it right!
Why HP maintains large market as it does, even though I agree that his OS is probably the most abandoned version of Unix?
1. For now, they are keeping their customeers in their lock-down. And they can keep this going quite a while more. Some apps and solutions are HP-UX only based. Agreed?
2. One can’t buy SPARC or x86 and load HP-UX to run those HP-UX specific apps. Agreed?
3. Meaning until apps don’t move they have no competitor.
Now introduce what 0xbadbeef and you proposed.
1. Since Solaris and HP-UX merged, suddenly all apps run on Solaris too. They got competitor in OS field. Agreed?
2. One could now buy x86 or SPARC to replace HP machine. They got compeitor in hardware field. Agreed?
3. Apps just moved, there is competitor. Agreed?
4. And points 1,2,3 all together put a nice perspective on HP making marketing suicide.
The only way HP could possibly go with this is if they would decide to go with SPARC. Which is NOT GONNA HAPPEN’
Now let’s get realistic. For a minute forget that you would like this to happen’ and think what HP is.
1. HP is a company and it is out for money.
2. It is not interested in common good as much as in bigger proffit
3. If they would merge with Solaris they would practicaly go against point 1 and 2, because they would introduce competitor (where there now is no competitor) by their own will.
p.s. I also remind you on your post where you said why IBM and HP support Linux. Well, here is the same. To make money. But as long as they can keep their HP-UX customers happy…
But on the other hand, why IBM and HP keep supporting Linux and why not Solaris? Because, Linux is not a competitor while Solaris (and with this fact Sun) is.
Ok, but lets be funny now. Now, you explain it to me how introducing direct competitor for their own customers and hardware would be positive for HP. I’m probably to stupid or too realistic to see this.
Opeteron is not supported by Solaris 10, are you absolutely sure about that?
What I meant my Opteron PC (you just have to learn that some people in PC land use opteron, pentium, xeon instead of PC term). Where both Adaptec and network card are on HCL.
But as far as my friends go, they usualy buy cheap HW.
For a long time Sun has ignored x86, it is no secret. But if my memory serves correctly, Linux had a serious upward climb for hardware support as well.
Except that I’m not jung anymore, which means I’m not prepared to wait for something that is gonna be in 7 years or so, I’m solely interested in what is now. We had that talk already, remember my fruitless question about which SPARC to buy.
In regards to porting HP-UX apps and data to Solaris (either x86 or SPARC) depends on the application. In many cases the application already exists for both platforms, it is simply a matter of porting the data from one to the other. Applications written in-house or custom are an entirely different matter.
I haven’t used HP-UX in several years, last time was 2002 but I remember not having too much trouble with either the OS or the hardware, with the lone exception of CD mounting. For HP-UX 10.20, 10.10 and 9.x it required the use of a third-party utility called pfs_mount in order to mount a CD. I thought that was pretty lame considering Solaris had vold since 2.5.1.
IBM and HP support Solaris, not only in hardware but software (OpenView, Tivoli, etc.). They have no choice because that is what the customers want, and they want to sell both hardware and software. Even to Solaris users.
In regards to your difficulties with Solaris x86, try using this CD on any x86 machine you want to install Solaris x86 on (hopefully this will be updated):
http://www.sun.com/bigadmin/hcl/hcts/install_check.html
Another thing, on your x86 box is ACPI and Plug-and-Play shut off? In some cases turning off these might help.
Applications written in-house or custom are an entirely different matter.
And here is where you hit the nail. Those are most common case of lockdown on some platform.
last time was 2002
it’s more 90’s for me. HP-UX and IRIX
Another thing, on your x86 box is ACPI and Plug-and-Play shut off? In some cases turning off these might help.
Yep, ACPI and PnP mode is off (although some BIOS settings don’t allow you to disable ACPI, most you can do is set Sx state on those boards). This is default correct mode for 99% of PC systems.
will check install_check, but I suspect it will just say what I already know. 100% hcl compatible. My problem are probably sub-standard drivers here.
> What is clearly HP-UX’s greatest strength?
> What is clearly Solaris’s greatest strength?
There is not a single area where Solaris 10 would be behind HP-UX — Solaris is clear winner pretty much across the board (performance, scalability, security, virtualization, resource management, supported platforms, etc.). HP-UX fans can probably only claim having small and generally not very significant tools like SD-UX, SAM, and glanceplus, that’s it.
> HP’s LVM is soo good that even though at one point HP started including Veritas Volume Manager for free with HPUX, NOBODY USED IT, becase the HPUX LVM is so much better and easy to use.
Hahaha, you don’t know much about HP-UX, do you? HP LVM was always based on VxVM and HP was quite open about it. So, if you like something about LVM in HP-UX, you should be praising Veritas and not HP. This is also probably part of the reason HP-UX costs an arm and a leg — HP has to pay Veritas with every unit of HP-UX shipped.
> HA Clustering. HP’s MC/ServiceGuard is the BEST HA clustering failover software there is. I have experience with them all, SUN Cluster, IBM HACMP, Veritas(on sun), and MC/SG. MC/SG is ROCK SOLID and easy to work with.
Pfft, please. ServiceGuard the best HP solution? Sounds like that is probably the only solution you touched. To my best knowledge ServiceGuard is also based on Veritas Cluster which sits above the OS, so it can not see a lot of things that might be happenning at the OS/hardware level. Sometimes it takes ServiceGuard up to an hour to make failover happen. Sun Cluster on the other hand (just as TruCluster on Tru64) is tightly integrated with the kernel, so fault detection and failover a lot cleaner/faster. BTW, all the future versions of ServiceGuard will be just carbon copy Veritas Cluster because HP gave up porting TruCluster to HP-UX and desided to use Veritas brains instead of their own.
Hahaha, you don’t know much about HP-UX, do you? HP LVM was always based on VxVM and HP was quite open about it. So, if you like something about LVM in HP-UX, you should be praising Veritas and not HP. This is also probably part of the reason HP-UX costs an arm and a leg — HP has to pay Veritas with every unit of HP-UX shipped.
True, but VxFS/LVM is an integrated part of HP-UX and in my book a feature in its favor, even if it did not originate with HP. It would be interesting to know what the royalties do to the price though. Cut the man some slack, at least until ZFS is in production Solaris, I think HP has the filesystem upper hand.
But anyway, I mostly have to agree with your earlier comments. I like HP-UX just fine, but I don’t think I could justify deploying it in a new install. I don’t see anything but maintainence in it’s ongoing development, and heck, even IBM puts AIX prices on its website now.
Now, if they had done what they said they would do with Tru64, things might be different. OpenTru64 would be awesome, but I think that AdvFS also has Veritas code or at least IP, and so I don’t think that would be possible.
> HPUX is more popluar for pure servers in the business world doing things like Oracle.
Hah? I guess that explain why 2 out of every 3 Oracle deployments take place on Solaris. Also it’s kinda interesting that Oracle chose Solaris as their primary development platform for Oracle DB and not HP-UX.
Not trying to troll here, and will admit that my understanding of the unix enterprise market is limited.
But, last I saw HP was the leader in UNIX server revenue and linux server revenue (and units shipped). I’ve read the posts above about Solaris vs. HPUX, but seriously, if HP is in the top spot is there a gain for them here?
I applaud Sun for embracing open source. I’m not sure that anything pivotal has changed in the industry because of it, but these things take time. That aside, Sun just lately strikes me as being that annoying kid always jumping up and down saying “look at me! look at me!”. I can’t tell if this is more marketing spin or if there’s something substantial here.
Sadly, confused.
Only a pony-tailed hippy moron like Schwartz goes around writing idiotic open letters as a bad PR move. This guy is almost as clueless as all the linux fascists out there.
(1) Open source Solaris in a spastic attempt to get 12 year old source forge kids to contribute and belong to a commmunnniiiittty.
(2)Don’t open source Java though – which is about the only chance that disaster of a language has of not making it in to COBOL status in the next 5 years
(3)Threaten to release Solaris under GPL v3 as if the Linux/GPL nazis will suddenly come to the rescue.
(4) Write open letters that a 3rd grader would be embarrassed to do.
(5) Don’t profit!- but hope that you can find another lawsuit to shakedown Microsoft with and get another billion to prop up the already dead Sun
still a f–ktard barkley
>Hahaha, you don’t know much about HP-UX, do you? HP LVM
>was always based on VxVM and HP was quite open about it.
Actaually, I have more than 15 years of Unix Admin experience, including continuously being employed as an admin of HPUX since 9.05 and Solaris since 2.1.
HP’s “LVM” is NOT BASED on Veritas, its HP’s own brilliant invention… HPUX DOES include the Veritas FILESYSTEM (vxfs) though as its base filesystem type which is what your obviously confusing with everything else you said.
>Sounds like that is probably the only solution you
>touched. To my best knowledge ServiceGuard is also
>based on Veritas Cluster
Again, your confused and probably out of lack of experience. HP’s MC/SG is a unique HP developed product and is ALSO NOT based on anything from Veritas. I HAVE used Veritas cluster as well as Sun Cluster and IBM HACMP. In fact we have all of those and MC/SG in use today in our shop.
>Sometimes it takes ServiceGuard up to an hour to make
>failover happen.
HUH? Our multi-TB oracle databases take only a few seconds to failover…basically just the time it takes to shutdown/unmount-mount/startup the database. And don’t get met started on Sun Cluster. I have been using it since it REALLY sucked. Before version 3 it used a SERIAL console server for its quorum LOCK…very, very lame and unreliable.
Not to make too much offense, but my knoweldge of nix-stuff is most likely more than a few levels then yours. I would suggest not sounding so authoratative when your real world experience is obviously lacking.
Edited 2006-03-05 03:51
> HP’s “LVM” is NOT BASED on Veritas, its HP’s own brilliant invention… HPUX DOES include the Veritas FILESYSTEM (vxfs) though as its base filesystem type which is what your obviously confusing with everything else you said.
Hmm, gee I don’t have 15 year os experience with HP-UX, but as sure as hell I have known for a long time that LVM is just a rebranded version of VxVM integrated into the OS. Here is an excerpt from Wikipedia on HP-UX that can vouch for that:
“HP-UX was also among the first Unix systems to include a built-in logical volume manager, a derivative of the Veritas Volume Manager (VVM). HP has had a long partnership with Veritas, and they use VxFS as their primary file system.”
> Again, your confused and probably out of lack of experience. HP’s MC/SG is a unique HP developed product and is ALSO NOT based on anything from Veritas
May be you should talk to HP engineers about that. Also think about it for a minute, there is a good reason why the next version of MC/SG will be based on the current version of VCS — it is the only way to provide continuity for the MC/SG because they have common origins.
> And don’t get met started on Sun Cluster. I have been using it since it REALLY sucked. Before version 3 it used a SERIAL console server for its quorum LOCK…very, very lame and unreliable.
You’re talking about Sun Cluster 2? How many years back are we talking about? BTW, the quorum server concept still being in use in the *current* version of MC/SG is not much different from using serial in Sun Cluster 2 and is just as lame. Sun Cluster 3 and above are a big departure from that and are much more refined products.
Edited 2006-03-05 04:32
>”HP-UX was also among the first Unix systems to include
>a built-in logical volume manager, a derivative of the
>Veritas Volume Manager (VVM). HP has had a long
>partnership with Veritas, and they use VxFS as their
>primary file system.”
Doesn’t say that now because it was incorrect. HP like IBM for AIX developed their own LVM and it does not have any shared history with Veitas’s LVM. Veitas VM is available as an add-on to HPUX but HP’s OWN LVM is the default.
Believe me I had these LVM history discussions over lunch with developers on the LVM team at HP many years ago.
> Doesn’t say that now because it was incorrect. HP like IBM for AIX developed their own LVM and it does not have any shared history with Veitas’s LVM. Veitas VM is available as an add-on to HPUX but HP’s OWN LVM is the default.
Oh, man! HP’s LVM *is* based on some antiquated version of VxVM (probably 3.x at the very most or even earlier). This is the reason you can also install VxVM from Veritas — if you want the most up-to-date features of VxVM, Veritas have you an option to install the latest version of VxVM. As simple as that, but it doesn’t change the fact that LVM *is* based on VxVM. Raise a support call to HP for HP-UX and ask them that very question, you should get the same answer.
>This is the reason you can also install VxVM from
>Veritas — if you want the most up-to-date features of
>VxVM,
So… Solstice Disksuite is also based on Veritas because you can install if on Solaris?
Seriously, you really don’t know what your talking about.
Unix History School is now in session.
HP’s LVM came from the OSF LVM which was IBM contribted to the open software foundtation by donating a copy of the LVM from AIX3 in 1990. AIX had its LVM for a few years before that.
In 1990, AT&T and SUN made a “pact” which made several other UNIX vendors feel threatend by the SUN’s buddy buddy relationship with AT&T. They formed the OSF, The OSF released its version of UNIX called OSF/1 based on BSD+AIX3 in 1991. DEC, HP, Appolo, and IBM were all members of the OSF. They all had licenses to the code in OSF/1. However, by 1993 most of the commercial vendors of Unix had changed their commercial variants of Unix to be based on SysVr4, so they all pretty much stopped their plans to use OSF/1. Digital being the exception and they continued with it and it eventually became Tru64.
However, HP took the LVM code from OSF/1 that they licensed and made some enhancements and put it into HPUX 10.0 in 1995, although it was available to customers to run on the 800 build of HPUX as early as 9.05 around 1993(about the time I first started working with HPUX).
So in summary:
AIX —> OSF/1 —> HPUX
>Oh, man! HP’s LVM *is* based on some antiquated version
>of VxVM (probably 3.x at the very most or even earlier).
UNIX history class is now in session.
Nope.. HP’s LVM is based on the LVM from OSF/1 which was donated by IBM from the LVM in AIX 3.0. AIX had it for a few years before IBM dontated to the Open Software Foundation.
Around 1989 or so, SUN made an a rather exclusive agreement with AT&T over access to UNIX SysV.
This really pissed off HP,DEC,IBM, and Apollo so they got together and formed the OSF. The OSF released its version of UNIX called OSF/1 in 1990. OSF/1 was mainly build on the current BSD that everyone was using plus a number of things from AIX 3.0 code that IBM donated to the project.
By 1992, AT&T had sold their UNIX stuff and SUN no longer had its buddy-buddy agreement to use and everyone else was able to get easy access to the SysV UNIX. Almost all the vendors migrated their OS’s to using SysV instead of BSD, so most of the partners who formed OSF/1 ended up not using it. Only Digital stayed with it because they really didn’t have anything else. They keep using it for many years and it eventually became Tru/64. BUT, HP did take the LVM out of the copy of OSF/1 that they were entitled to and made some improvements to it and added it to HPUX 10 in 1995. Although, it was availabe to be used in the 9.05 version only on HP800 sytems in 1993(back when I first started working with HPUX).
So…. The history of HP’s LVM is basically this:
AIX 3.0 —> OSF/1
……………………..
……………………… HPUX 10.0
There is no veritas volume manager in the HPUX or AIX LVMs, or for that matter Tru64’s either.
Also, the original poster makes it sound as if VxVM 3.x is old. It’s not, only very recently has Veritas released VxVM 4.0.
Sit and swivel on this middle digit.
Seriously, McNealy better be careful, because the way the impression he’s giving his customers and his rivals is one of complete and utter desperation. Someone would think there is something wrong with Solaris’ business or Sun’s future as a company.
You know, having read your (segedunum’s) comments to SUNW-related threads on OSNews for a while, I thought you’d recognise the theme.
Both McNealy (and Jonathan Schwartz) like SUNW to be the current topic of conversation.
What have we here? Oh look, a current topic of conversation. Reading through this thread it seems that there are quite a few people talking about SUNW, and Solaris, and how HPUX ain’t all that great any more. On a brighter note, McNealy is offering an olive branch of sorts for the customers — “we could do this together, and all our customers win.”
Remember that quote from (iirc) Oscar Wilde? “The only thing worse than being talked about is not being talked about.”
“Someone would think there is something wrong with Solaris’ business or Sun’s future as a company.”
Yes, if they keep this up people might start thinking Sun is going the way of SGI.
I suppose the same “desperation” is being displayed by IBM with its “Solaris to Linux migration” and HP’s Martin Fink on why you should be using Linux ads. Of all of them I would say that HP is the most “desperate”, their “brain trust” killed off most of their Enterprise product line and expect companies who have spent millions of dollars on HP PA-RISC hardware and software to just migrate to Integrity servers, or ProLiants running Linux or Windows and spend even more money. Yeah right! What is that saying “Once burned, twice shy”?
What you call an act of desperation is what smart businessmen call an opportunity, and I don’t blame Sun for taking advantage of it. And since all existing HP PA-RISC customers are going to have to migrate away from PA-RISC at some point in the future, why not appeal to those customers. And if Sun actually helps HP, that is a lot more than what I have seen from HP. And you don’t think IBM is looking at opportunities to woo HP customers, think again. It is just that IBM isn’t so blatant as Sun is about it.
> Nope.. HP’s LVM is based on the LVM from OSF/1 which was donated by IBM from the LVM in AIX 3.0. AIX had it for a few years before IBM dontated to the Open Software Foundation.
So now you’re starting to sing a different tune — it is now IBM that created the LVM? Well, I’ve heard this theory many times as well and quite frankly the jury is still out on that one — since there is no official publicly available documentation on this matter, a lot of people see OSF/1 LVM and automatically assume that it is the same thing as HP-UX LVM. To me the Veritas theory makes more sense simply because 1) the semantics in HP-UX LVM are lot closer to the old VxVM than AIX LVM and 2) there must be a pretty good reason why HP-UX LVM relies on JFS (rebranded VxFS). As I said before the jury is still out on the real ancestry of HP-UX LVM, but either way you slice it your claim that LVM was the product of HP engineering is false — LVM was not developed by HP, it was just handed down to them.
> Sorry to burst your bubble, but even my Opteron is not fully supported under Solaris10.
Hah? What the hell are you smoking over there? Solaris 10 runs better on Opteron than any other OS out there. Solaris crushes both Linux and Windows on most benchmarks on the same Opteron hardware.
>HP-UX LVM relies on JFS
Worry..if you don’t know the difference bettween a LVM and a FILESYTEM then your talking about stuff WAY out of your league.
AND since you obviously are lacking any real-world experience, fyi the commands in AIX LVM are “pretty close” to HPUX LVM. The ,names are slighty differet, but the process and the steps you do are the same.
They are both WAY different than Veritas..no encapsolation stuff, HPUX and AIX have no concept of the “private region” etc…they work way different than Veritas.
Some how I have the feeling that, I was using AIX’s LVM back in the early 90’s, probably before you could walk and talk at the same time.
> Other than the fact that Sun’s double talk about Linux over the last few years has pissed off a large portion of the Systems Admin community who are the ones responsible for makinga and influencing these sort of decisions.
What double talking are you talking about? Sun’s message in relation to Linux has been very consistent over the years — Linux is a good thing on the low end, but Solaris makes more sense accross the board from low to high end. And I absolutely agree with that. It is just Linux zealots were mighty pissed off just because Sun refused to kiss the penguins ass while saying the truth and standing up for their products. I really respect Sun for not budging to the latest fad of the day and actually standing up for themselves and their products (Solaris in particular).
Edited 2006-03-05 23:26
> no encapsolation stuff, HPUX and AIX have no concept of the “private region” etc…they work way different than Veritas.
Actually that is a good point and I have to agree with that — the private region and encapsulation concepts are not present in HP-UX and AIX LVM.
> if you don’t know the difference bettween a LVM and a FILESYTEM then your talking about stuff WAY out of your league.
You know that I know what I’m talking about — there are many rather intriguing connections to Veritas scattered around in HP-UX I/O subsystems, but you don’t have to be such a d*ck about it.
Edited 2006-03-06 01:37
Really??? Can 1 company flip flop more? They’re worse than some political parties…
“We love Linux, and I hope there isn’t any doubt about it,” McNealy told financial analysts at its annual meeting here, dressed as Tux, the seemingly innocuous penguin mascot chosen years ago by Linux founder Linus Torvalds. http://news.com.com/2100-1001-832463.html
Better read up on the latest FUD that “Solaris is a Better Linux than Linux” – not sure how that’s even possible, but leave it to the SolarFUD machine to suggest it is….
http://www.linux-watch.com/news/NS5562143406.html