Sun Bashes Linux on (IBM) Mainframes

dagbrown writes: "An article linked from Sun's front page, entitled "Linux on the mainframe: Not a good idea" by Shahin Khan, Sun's chief competitive officer, has the interesting theory that Linux on mainframes makes no sense because, among other things, the VM/Linux combo isn't a very good match. What do the folks on Slashdot think?"

It makes sense...

[Anonymous Coward]

Like Solaris on pc

Really?

[TheReverand]

Actually, I thought Sun would be saying lots of nice things about their arch-rival, and complimenting them on their excellent choice of operating system.

Re:Of Course not!

[JWW]

Yeah I love their talk about utilization. Oh no i f you buy one of these IBM boxes you have to plan out what your peak utilization is!!!

And if you buy some sun machines to do the same thing you have to .... ummmm ... plan out what your peak utilization is!!

The add that the server can't dynamically create more utilization capacity (extra hardware) dynamically. If anyone out there were selling a box that could do that, I'd buy it in a heartbeat. First we'd need some good nanobots, or maybe a replicator.....

Re:New title?

[Andrewkov]

Yeah, his job is to sling FUD at the competition.

What is it with these Sun guys?

[lhand]

They keep shooting themselves in the foot wrt the Open Source crowd. Now the've reloaded and started shooting again. You'd think they would have run out of ammo by now.

One of the beauties of Linux is that it can be ported to so many different platforms easily. Sun uses it and then goes on to say IBM shouldn't? wtf? There are valid reasons to run Linux in multiple virtual machines. I even do it here on my PC.

Note to self: Must drink less coffee....

Hey - me too!

[SubtleNuance]

And it was a seriously easy. There were never a problem with the virtual machine suddenly giving up the ghost from underneath us, and we'd never loose a Samba process for any reason whatsoever. We had load averages that were very normal, and it was like we were always scrambling to find new reasons to use it, as opposed to setting it up and just having forgetting about it. We used to tell the students that the machine was seriously hot, and it would never fall-over. We were even thinking of doing up artwork -- AND STARTING A RELIGION!


All those impressive demos where they have 32 hojillion instances of linux running on a mainframe are so meaningfull. Sure, you can do without it, but it will do everything. If you try actually working with the setup, you never have to reboot your machine -- instead of doing it 10 times a day, and those other take forever to freakin' reboot.

But who are you going to believe, ummm, me or ummmmm, that other guy.

Sheesh.

Re:CCO?

[pnatural]

Next he'll be telling us that Sun(tm) Java(tm) is a fully Object-Oriented Systems Language, Sun(tm) J2EE(tm) is a Standard, and that they, Sun(tm), do not ignore the Sun(tm) Java Community Process(tm).

Re:I actually worked on Linux on a Mainframe...

[darkonc]

We used to tell the students that the machine had caught on fire and had (literally) fallen over.

I gotta hunt down and re-digitize the video that a set of students did back in 1992. At that time the UBC Computer Science GraFic lab had a stack of IBM RS-6000 computers running AIX (which I sometimes pronounced 'aches') which tended to crash far too often.

For their animation lab one group's video was an RS6000 shaking, smoking and melting down into a grave with (computerized) flowers sticking out of it and a gravestone blinking 888 (which the 3-digit LED display would do after a general system panic)

With later versions of AIX and some coddling from your's truely, I was able to make the IBMs a good bit more stable -- but never quite as rock-solid as the SGIs running IRIX -- and they never really outlived their reputation.

Then, of course, there was the obligatory IBM PC... Let's just say it was really good for running Castle Wolfenstein (the original).

Reading between the lines...

[Pig Hogger]

There is a lot of activity around Linux in the marketplace today. Strategies differ from vendor to vendor, and there are a variety of ways companies are experimenting with, evaluating, and deploying Linux in their IT infrastructures. Just like any technology decision, if and where to deploy Linux is a substantial decision which involves matching the right technology to the right business problem.

READ: Of course, with an infinite amount of money, every business problem is solvable at the highest profit for the problem-solver...

Just as it is important to understand when technology can provide you with an advantage, it is equally valuable to know when a technology is not suited to a particular task regardless of how 'hot' that technology might be. Sun does see a place for Linux in the IT infrastructure, as evidenced by our recent announcement to ship Linux-based servers. Sun introduced low-priced, horizontal Linux servers as an alternative to proprietary systems. Suited for Web delivery at the edge of the network, Linux servers offer an alternative to proprietary and closed environments such as Microsoft Windows.

READ: But Sun is of the opinion that it's proprietary servers, running with proprietary hardware (such as the special disk drives with the "magic" secret partition table, and the memory chips with the notch moved 1/10 inch) and, of course, running proprietary Solaris Operating System is the best solution overall, especially if you have an infinite amount of money to throw at YOUR problem...

Recently, IBM announced a new 'Linux-only' mainframe, the z800, which IBM is promoting as a way to consolidate multiple Linux and Unix[r] servers(1). Running Linux on a mainframe doesn't change the fact that you must still maintain an expensive, proprietary system, defeating the whole purpose of introducing open standards like Linux. Although it's technically possible to configure such a system, the question remains, "How well-suited is the system to the task?"

READ: Definitely, here, we have a blatant attempt at fitting an hexagonal peg into a pentagonal hole. Linux was conceived to run on discarded low-end hardware in order to satisfy teenage-geek impulses, which is quite a different thing than to run on the Big Iron dinosaurs IBM is well-known for.

Linux on the mainframe just doesn't compute. Here's why:

READ: You just can't run JCL and CICS and MVS and CMS and Assembler on Linux. Cobol (even GNU-Cobol) will make the kernel break into hysterics.

Linux on the mainframe is actually hosted by another proprietary operating system, z/VM. The optimized operating system for IBM mainframes is z/OS, formerly called MVS(2). Compared to z/OS, z/VM is a niche operating system with virtual machine (VM) support for new hardware features added late or often not at all(3). And Linux isn't designed to run in a virtual machine; implementation decisions that make sense on PC hardware don't fit well in a virtual machine(4). This is Linux. It's designed for Intel. It's not tuned for the mainframe hardware in which it's running.

READ: On the other hand, Solaris is designed to run (well) solely on Sparc architecture that made Sun famous.

Linux on the mainframe is complicated; this isn't Linux running on a two-way Intel server. Despite IBM's claims of easy management(5), customers still need a special machine room and specially trained staff for both z/VM and Linux. Finding mainframe staffing is an obstacle in many organizations(6); combining mainframe and Linux staffing further complicates the matter. Running multiple Linux images still requires administration that needs to grow with the number of images being run.

READ: Of course, you don't have any of those virtual machine nonsense with Sun products: you simply plug in as many boxes as you need of machines. No more software headaches for your operators!!!

Linux on the mainframe can't respond to the workload demands of Web serving with high utilization--something IBM touted at the time of its z800 announcement. Horizontally scaled Linux farms are designed to handle unpredictable demand with above average peak loads. As demand rises, a load balancer distributes the traffic evenly across servers, which increases utilization. Because design capacity needs to handle peak demand, server farms often have a low utilization.

READ: Why do in software what can be done far more profitably (and piracy-immune!) in juicy, expensi^h^h^h^h^h^h^h profitable hardware?

Given the relatively low cost of hardware, some organizations find this trade-off acceptable to ensure appropriate service levels. Contrary to what many believe, consolidating a Linux farm into multiple images on a mainframe would not change the demand pattern. Although z/VM can start and stop Linux images, it cannot dynamically add resources to match demand. As a result, a mainframe would need to size for peak demand just as the Linux farm would; high utilization is a myth.

READ: Of course, you can dynamically add and remove ressources from a Sun cluster with the use of specially trained monkeys and servoids, which is a better proposition than software hocus-pocusery within the deep, dark bowels of an IBM mainframe.

It's neither fish nor fowl. Linux on the "mainframe" is not an open system, and there is little incremental RAS benefit. Although IBM claims "zSeries servers inherit the legendary IBM S/390 strengths in the areas of fault avoidance and tolerance, recovery from failures, and concurrent maintenance and repair for "always-on" availability"(7). We don't believe this to be true for zSeries servers running Linux. The "legendary" IBM S/390 strengths IBM references are the result of decades of development work on IBM's flagship mainframe operating system, known today as z/OS. The fault recovery features of z/OS are not found in Linux. z/VM does have some fault recovery features, but it is not nearly as resilient as z/OS. For example, z/VM cannot take advantage of Parallel Sysplex clustering, and VM hypervisor is an added single point of failure(8).

READ: Of course, Sun cannot take advantage of Parallel Sysplex clustering either, but that's more alphabet soup to muddy the waters and instill doubts into potential buyers.

Applications that run on Linux for Intel need to be recompiled and recertified for each new platform; thus the application portfolio to run Linux on a mainframe is small(9). Consolidation without application availability just can't happen--and if the applications don't run on your platform, or if there are costly ports and changes to be made, cost savings can't be realized. Often the difference in Intel versus mainframe applications makes porting difficult(10). Additionally, different applications are ported to different distributions of Linux (for example, Red Hat, SuSE, and Turbolinux). Getting applications to run on the mainframe might require supporting multiple distributions of the 'same' Linux operating system.

READ: Again, Linux fragmentation is a terrible tragedy that will never happen to Solaris.

The economics just don't work. IBM claims it is financially justifiable to consolidate as few as 20 Linux servers on a z800(11). With an estimated starting price of $400,000 for a z800(12) with a single CPU engine enabled, that claim seems exaggerated compared to Linux servers that hover in the $1,000 to $2,000 range. Sun's rack-optimized 1U form factor servers start at list prices under $1,000. When customers realize Linux on mainframe utilization will be low, and administration costs have still not been factored in, you can begin to see how the costs will add up. And let's not forget the support costs that will need to be purchased, either from the distributor or IBM Global Services. One example of a distributor's cost on an IBM Multiprise ran in the tens of thousands for the initial services and thousands a month for ongoing service(13).

READ: Even if " nobody ever got fired for buying IBM ", we whish the same could be said about Sun.

Thus, when considering consolidation projects, why consider putting workloads onto a completely different and more expensive architecture? Why put an open operating system such as Linux on a closed proprietary mainframe? Why consolidate on a system with limited application support and one that demands a rare combination of skills?

READ: IBM is double-plus uncool amongst geeks. SUN is the hip thing to use in IT!!!

Re:New title?

[2Bits]

Yeah, this is for people whom you don't know what to do with. You give them a nice title with no responsibility, until they figure out that they are useless and they leave. But some of them will never do.

In our company, we have some many "chiefs", and not enough Indians. I'm working on a project, and I'm the only architect, system designer, programmer, QA, integration engineer. In another word, a one-man show. But I have to report weekly status to 5 chiefs, each one wants it in a different format.

The 5 are:
- Chief Technology Officer (boy, this is a high visibility project, even the CTO has to know the weekly status. This guy wants the status in MS Project format).
- Chief Product Strategist (What the heck, the guy is responsible of another product, but I don't know why he wants the weekly status of my project. But he makes sure that the CEO knows that he MUST get the status every week, so I can't get the CEO (of the division, that is) off my back. This guy wants it in Excel format, as he has a lot of product matrices in Excel).
- Chief Competitive Office (well, I never figured out his role, and will never figure out why he wants the status either. Ah, but he wants it in PowerPoint. He doesn't want to waste his time to recreate the information when he has to do presentation, does he?).
- Chief Project Management Officer (he can't even keep up with other real projects, but he still run after me every week for this stupid pilot project/prototype. Another MS Project format guy.).
- Chief Customer Service Officer (what the fuck, the project I'm working on is just a prototype, for god sake... This guy wants the status in his searchable database, so I just give him all files that I gave to the other 4 chiefs.)

Life sucks sometimes.

Computers literally falling over

[billstewart]

There's a story from the '89 quake that somebody placed a service call to Tandem after the quake. (Tandem, later bought by Compaq, made highly reliable fault-tolerant machines with N processors, disks, backplane busses, etc. backing each other up.) The machine had fallen over on its face. It was still running just fine, but they wanted Tandem to come put it rightside up just in case anything went wrong in the process... (Also, they were big heavy machines....)