Follow Slashdot stories on Twitter


Forgot your password?
Silicon Graphics

SGI's Linux Server 103

More details of SGI's upcoming Linux server have emerged. According to the article, SGI is already shipping a 4 processor machine, and plans to ship 8-way and 2-way machines at a later time. Both Linux and NT are available pre-installed. The servers look like they're not your mother's typical x86-based server, and come with some interesting hardware features that I didn't even know Linux supported (hot swap drives). Am I not well-informed, or is it possible SGI has some patches (that they're hopefully itching to give us)? Regardless, I'm glad to see a big Unix vendor shipping Linux, and touting it so highly.
This discussion has been archived. No new comments can be posted.

SGI's Linux Server

Comments Filter:
  • by crow ( 16139 ) on Thursday July 29, 1999 @04:10AM (#1777671) Homepage Journal
    [Note that SMP is very different from clustering, so forget all you've heard about Beowulf, Mosix, and such for this discussion.]

    Yes, Linux supports SMP. If you have a bunch of user processes that are doing mostly user-space computation, then you should get a mostly linear speedup even with a 2.0 kernel. But if the processes depend on a bunch of kernel services, watch the kernel version carefully.

    You see, with SMP the kernel has to be sure that two processors don't try to modify the same data structure at the same time so as to avoid confusion (e.g, deadlock or a crash). The 2.0 kernels supported SMP by putting one big lock on the kernel, so only one process can be doing something in the kernel at a time. One of the major features of 2.2 is that this one big lock was broken up, allowing Linux to scale far better. Unfortunately, there are still many places where the locks need to be broken up further, and work is ongoing.

    I believe that it was such a locking issue that caused Linux problems in the Mindcraft benchmarks when multiple network cards were in use.

    So the performance of SMP will vary depending on your application and your system. (I've heard, for example, that building the kernel on a 4-way box gave a 3.7 times speedup in one case, which is pretty good.)

    Note that many of the commercial Unixes scale better than Linux (i.e., have finer grained locking). I know Solaris has a very good reputation for large numbers of processors. I would suspect that Irix must be good, based on the systems that SGI sells. Linux is getting there. Upto 4 CPUs should be fine for most tasks. I haven't heard reports of Linux with more than 4 SMP CPUs.
  • Not to burst anyone's bubble, but this system is just a re-skinned version of Intel's SC450NX server platform. The machine is made by Intel for system integrators. You can find the Intel information at n/servers/SC450NX/ []. The company I run sells systems based on these, and the placement of the status LEDs, the lock, and other things that I can see in the pictures in the PDF that's found in a comment above are exactly the same.

  • Actually, the fbcon X server has been working on the VisWS for quite a while now.

    The new 1400L seems to be quite different than the VisWS, enough that I wouldn't call it a VisWS biled as a server - This one doesn't have the Cobalt graphis (of course), it has integrated SCSI, an emergency management port (monitoring stuff in some way), three hot swappable power supplies, and of course it installs RH as "server" by default :)

    It's probably built with the same core UMA (Unified Memory Access) chips as the VisWS, but if you call it a rebranded VisWS, then I guess Compaq et al does the same thing, since their servers aren't that different from their high-end workstations - they use the same intel 440 GX chipset and such.
    /El Niño
  • >Has always been available in linux...with scsi or >ide.

    That doesn't mean it ever worked reliably.
  • The fellers down at SGI gave a talk at my school's LUG (Waashington State University) a while back and we had to telnet to it via serial port because there's no USB support in Linux, and this box has USB everything. So does anyone know if SGI is going to release all their patches to the public? I'm assuming they wouldn't release a inux box that you can't use a keyboard or mouse with.
  • Will they use the Cobalt Chipset?


    Will they use the same Motherboards?


    Will this simply be "adding a drive sled bay" to a visual workstation?

    No. It's not based on the VW architecture at all.

    Will they be cutting back on the Video and Audio abilities?

    Why does the price mentioned seem higer than the Visual Workstation (if you just adding a sled, but taking out all the video and audio stuff?)?

    Well, the 2+1 redunant power supplies, for one thing. The hot-plug drive bays, for another. The 11 fans, for another. From what I've seen, it's a fairly well-engineered machine.

  • As much as I'd like to believe this, I think that contract provisions such as his would have come up in US vs. Microsoft.

    That no representative of SGI has been called as a witness for the DOJ has no bearing on it then?
  • I just hot swapped the other day when a disk failed in my RAID 1 (mirroring) setup. I just had to take out the disk that was making lots of noise and plop another one in. Unfortunately the md and raid1 driver didn't automatically populate the new drive, so I had to take my box down to single user mode and cat /dev/hda>/dev/hdc *grin*.


  • And you'll get one hell of an intel system for $14,000 from SGI, I reckon. Their pricing seems to be about inline with the other server vendors - Dell, IBM, Compaq - - - There is a premium involved, but as everyone points out, that premium equates to added bandwidth
  • I'll bet that they've been doing some tinkering with the Linux kernel. I've noticed quite a few posts on the Linux kernel mailing list lately from IRIX programmers inside SGI.. I thought it was kinda interesting... SGI is a great company, and having them getting involved in Linux is like having a major league ballplayer join your softball team. :)
  • Unless you're 100% I/O bound on DMA controllers, I guarantee that NT and IRIX both will kick Linux in the *** on 4+ CPUs, since both have 100% reentrant multithreaded kernels, while the Linux kernel is not reentrant. This is changing, but for now we have to live with it on big boxes. (see the post Re: David Miller; I know he's done a lot of testing on 8+ CPUs and the scaling is quite horrible; Solaris SMP really sucks and it outperforms Linux about 2:1 on 14 CPUs)

  • Also, why the hell dont they make a move to the
    EV6 with the k-7's... that will surely rock the house!

    Read some pages on their site. They're migrating toward IA-64 and ditching MIPS within the next couple of years. This is one step in their migration. More than likely, they'll release one last round of Onyx2's and Origin 2000's based on the R12k CPU, and then we'll start seeing all new BIG machines. IA-32 is just their way of testing the waters.
  • by DdJ ( 10790 )
    Don't forget that Linux boxes with PCMCIA support have been able to hot-swap a variety of devices, including hard disks, for quite a while. Even if the hot swapping *isn't* handled by the RAID controller, it's not as suprising as it might sound.
  • The best thing about this is SGI "giving back" to the Open Source community. SGI has written some great software over the years. If we get just a piece of that, we should be grateful. It wouldn't take that much work to integrate SGI technology into something like the GIMP, or other great Open Source software.

    What's especially great about this is that SGI seems willing to participate in Open Source advancement. It will be a great boon to see such a powerful name behind Open Source.
  • I spoke to an SGI insider about this a while back. His comment (anonymously) was that the issues that hold back an SGI "workstation" (ie., X with accelleration for the Visual Workstation) are mostly legal; apparently, SGI had to sign some agreement with Microsoft in order to get the information they needed to get NT working on the box.

    According to this person, (Surprise!) Microsoft insisted on language to the effect that SGI wouldn't support a competing OS environment on the same hardware. SGI's strategy for getting around this is supposed to be to focus their Linux efforts on things that can be branded "servers" until the agreement runs out.

    I don't know how true any of this is (I don't have enough contact with the source to establish how accurate their statements are), but it may well be that the technical issues in getting X working right on the Visual Workstation aren't the only problems that SGI faces.
  • Really eh? I noticed that too, and it was coming from some sort of IS research pundit to boot. Maybe she's pining for the days of restrictive AT&T source licenses and $1995.00 binary licenses?

    I agree, SGI probably picked Linux because it made the best combo of economic and technical sense.


  • Yup. I loaded Linux on an HP Netserver LHII with dual PPro and an HP NetRAID controller, which is really just an AMI Megaraid resold by HP as OEM equipment. RAID5 & hotswap. At the time there were no system utilities for monitoring the status of the drives under Linux, but I think AMI did release some a little while back.

    It was also a repurposed NT server. It was a good feeling when I wiped NT off the hard drive. I haven't missed it once.
  • Is that similar to the old familiar "lp0 is on fire!" message?

  • >> SGI has written some great software
    >> over the years.

    If they could port imgworks over, I'd be set.
    I really miss that program... I use Gimp
    all the time, but imgworks just made some
    things soooo simple, like transparencies in

    Also, I'd love to see the SGI software
    manager ported over to use RPM images. That
    was such a masterpiece of software upgrade
    and installation efficiency.

    -Mike (who likes having a Linux box and an Indy at home)
  • It could be just supply and demand...

    Just becouse the managers ordering the computers know what Linux is dosn't mean they know what will happen on a Linux box with 8 Xenon IIIs.
    My guess is SGI found a way to make it work but don't count out the "They'll buy it anyway" factor. It's not SGIs style to market to the idiot quotent but you never know.
  • The framebuffer Xserver is extremely slow since it uses no hardware acceleration at this time. (Project like KGI/GGI [] will help by combining the simplicity and reliability of framebuffers with robust acceleration.) Since one of the main features of the Visual Workstation was excellent graphical performance (including hardware OpenGL), using an Xserver more suited to a microVAX's framebuffer seems somewhat subpar.


  • I can't offer "significant proof" yet, as the SGI 1400 doesn't appear on the external SGI web page. But it is not a Visual Workstation 320/540 without graphics - the 1400 is a new hardware configuration. As a previous poster mentioned, it has the standard Intel-based server options - integrated SCSI, redundant power supplies, etc. How do I know - I work there. :) There's a couple 1400s in my lab 100 feet away.
  • I can't offer "significant proof" yet, as the SGI 1400 doesn't appear on the external SGI web page. But it is not a Visual Workstation 320/540 without graphics - the 1400 is a new hardware configuration. As a previous poster mentioned, it has the standard Intel-based server options - integrated SCSI, redundant power supplies, etc.

    How do I know this - well, I work for SGI. :) There's a couple 1400s in my lab 100 feet away.

  • So, I guess, after review, the real questions are:

    • Will they use the Cobalt Chipset?
    • Will they use the same Motherboards?
    • Will this simply be "adding a drive sled bay" to a visual workstation?
    • Will they be cutting back on the Video and Audio abilities?
    • Why does the price mentioned seem higer than the Visual Workstation (if you just adding a sled, but taking out all the video and audio stuff?)?

    What are you at liberty to say?

  • Actually, they don't seem much different in price from the Origin 200 servers running Irix. Why buy Intel when you can get purebred SGI?


  • The biggest question around here is "what do we do
    if the patches aren't accepted?". That and the
    usual big company run-arounds about "procedure"...
  • As much as I'd like to believe this, I think that contract provisions such as his would have come up in US vs. Microsoft.

    They have, with other manufacturers. It's apparently a standard part of the Windows licensing agreement.

  • Ahh, good.. I was waiting for this. I had the good luck of being able to sit in on an SGI non-disclosure agreement demo back in December of last year when this was being discussed.. Glad to see they finally came through on what they were promising. :)

  • To quote the last paragraph from the article:

    "SGI officials said the company plans to give back to the Open Source community by making some SGI software code available to the Linux community"

    So maybe they will give the hot-swap patches...?

  • by Anonymous Coward
    Expect more super-duper Linux announcements from big companies as we get nearer to showtime. Watch with glee as they try to out-do each other in Linux support. It's gonna be an awesome show! Wheeee!
  • Wonder what they're doing to make 4-proc efficient? 8-proc? Good question. The file system and networking improvements are supposed to allow Linux to scale into that range, but that's only a best guess. I've never seen or heard of anyone demonstrating resonably good >4-way scalability on an intel platform with any operating system. (NUMA machines don't count) Last time I checked, NT couldn't scale past 4procs worth a damn either. What's SGI smoking. Whatever it is, Compaq and a few others must be smoking it too since they're planning 8way xeons as well. 8way K-6 would be a much better value proposition since the EV-6 architecture can scale to 14procs with reasonable performance.
  • Hot swap per se isn't really an OS issue, or even a software issue. It's a matter of being able to remove and insert the hardware without damaging the drive or the machine.

    Of course, for hot swap to be useful the system needs to be able to deal intelligently with failed drives, generally via RAID. I'm running Linux on a hot-swappable Compaq Proliant 1600 using software RAID-1, which works just fine. I've done the pull-a-running-drive test and watched the RAID driver rebuild the RAID system after adding the drive back in. It's pretty cool.

  • I believe you mean the new K7, not the K6.
  • "FOOD FIGHT!!" (obscure kernel panic from way back when)
  • I have to admit they've done better with the case on this one compared to the VW320/540. I especially dig the perforated metal on the front.
  • Typo, sorry. I'll just nip off and shoot myself now.
  • ... are a little hard to come by. There is a PDF File with some stuff on it []. But is appears only to be linked from the search engine.

    There is no mention of any nice Linux extras, just that it uses RH6.0.

  • by tgd ( 2822 ) on Thursday July 29, 1999 @02:38AM (#1777727)
    The SGI machines offering hot swap isn't that big of a deal. Any many linux systems can support it and any that support a RAID controller that supports it, will support it. (Wow, that was a lousy grammatical construction!)

    I meant that there are a bunch of RAID controllers that handle the hot-swapping themselves, including the automagic rebuilding of failed drives in a RAID-5 configuration, for example. I think the Mylex controllers are an example of that.

    Two years ago I managed to get Linux running on one of HP's high end (at the time) multiproc machines (it had two Pentium 200's I think). For the life of me I can't remember what drive controller was in it -- it was a repurposed NT server, not something purchased new -- but you could pull drives out just fine...

    I'd guess SGI's just using a hardware RAID solution, not a software one, which is the only thing that would need real Linux support.

    Wonder what they're doing to make 4-proc efficient? 8-proc?
  • I don't mean to be redundant, several people have pointed out that Linux already supports hot-swappable drives. This is true, to an extent. In a RAID configuration, the RAID controller is responsible for handling drives that are swapped in and out. In our Compaq ProLiant 3000 server, we have 3 X 9 GIG SCSI hard drives in a RAID 5 array. This server could run almost ANY OS (ours runs NetWare), providing that support for the RAID controller is available. Of course, in order to handle hot swapping it also requires that the hardware like drive bays, drives, and cables adhere to the standard used by the RAID controller. The Compaq Proliant has drives with nice lights on the front and tabs that allow the drives to be removed without tools, and a modular cage to hold them. As far as I know these are primarily proprietary solutions specific to each of many server vendors (HP/Compaq/IBM and others).
    In order for hot swap ability, a company like SGI has to offer hardware designed to work together (which they already do on their other server lines). They would then have to write or modify a driver to allow Linux to use the RAID controller, it doesn't require any great change in the OS. While its great news for Linux that SGI is building servers with Linux pre-installed, they haven't contributed any significant amount to the OSS. However, I hope that they will contribute to, and perhaps speed development on making Linux more SMP capable.

  • Not true. I have pulled a processor from a running system once in a while, and the system kept running without problem.

    I'm told there are some systems that you can upgrade the memory at any time too, but I don't normally work with those.

    For hot swap there are two issues, the hardware needs to be able to do this without damage electricly, and support what software cannot do. That is easy (relativly). The software needs to recignise the swap and deal with it. For riad the only hard part is detecting a new disk, but pocessors gets tricky.

  • I think this is planned for 2.4 (or somewhere in the future). Except it would probally be something like "umount /dev/cpu3" where you have to tell linux that you are killing the cpu. Maybe there could be someway that it could detect a failing cpu and shut it down automatically, but i think yanking a running cpu would do very bad things to the computer.
  • "Becknell is not convinced it is the best bet.
    There are many other Unixes they could adopt that are already scalable, high-end, high-available, designed for performance. Linux is not even fully SMP [define.gif] yet."

    Err.. many other Unixes like Irix for instance?! Even if we put aside the the yeas and nays of the SMP debate - how much imagination does it take to realise that when a company that has a Unix of _of_its_very_own_ looks at another unix they're looking at a bigger picture. Even if we take the most cynical position and assume SGI are just hedging their bets this is a win for linux. SGI have recognised the value of open source - not just its conciderable technical merit (potential) but also its "friendliness" to corporations that want to be part of it.
  • Hot swap doesn't even require RAID. It involves plugging the device in and making the SCSI bus aware of the change, via the tickling of some /proc/ values. Also, the documentation suggests that it is possible to hot-swap with software RAID as well as hardware RAID.

    Bottom line: it has little/nothing to do with the question of RAID or not; it has to do with your SCSI setup.
  • Thought I'd mention, with SCSI drives its also very easy to get Linux to rescan the bus. I have a shell script on my system at home that does that, because I'm a knob and forget to turn on my CDR drive more often than not. Flip it on, run the script and voila! No reboot.

    Of course, kiddies, don't try this at home. SCSI isn't really supposed to support powering on a device after the system is powered up, unless it specifically says it can. YMMV.

    I'm interested in seeing where FireWire goes from that standpoint. I like the fact that you can disconnect them at will. (Well assuming your OS isn't going to complain!)
  • Somebody enlighten me as to why I'd spend $8000 for An entry-level system with one CPU and minimal RAM... Or A more typical configuration, with two CPUs and more RAM, starts at $14,000.

    I love SGI machines, but Linux is free so what am I paying for? I could put one hell of an dual Intel system together for $14,000.
  • Quick summery of comments [pays to read before posting]
    It's a hardware thing. Accually thats a good thing [tm] software solutions cost in CPU time.

    Remember the origianl Amigas? Everything done in coprocesors. Let the Amiga do things much more powerful PC couldn't.

    Anyway it's allwase wise to make shure there is sorce code to ask for. As with this case there is no code and then there is the BeOs/Linux hybred that apparently dosn't even use Linux.
  • the hardware must handle it, I mean, you must be able to add/remove a drive to a IDE bus/SCSI chain without everything going mad. Most hardware I met doesn't care if you remove a (umount'ed) drive.
    Well, if you want to do this reliably, you want a connector that's designed to deal with this properly. The standard SCSI SCA connector is a good example: It's not possible to misalign, there are no pins to bend, and the connectors mate in the appropriate order (ground first, then data, then power). This also requires some software support, since you really want to be doing this on a quiescent bus.

    If you want really good performance/reliability, go for RAID-5. Hardware support is not required anymore.
    Well, it's not required if you don't want really good performance and reliability with your RAID-5. :-) To do it properly, you need an external unit (such as a Baydel []) that has a separate SCSI bus for each drive, so that a single drive can't lock the bus. Then you need two controllers on the Baydell, one for each of the two controllers on your host. Now you can lose any of the following and still stay up:
    • a drive
    • a RAID controller
    • a SCSI cable
    • a SCSI controller on the host
    Of course, on many machines you're still prone to motherboard failure at this point, which is why you're probably hooking this up to a Sun E-series box. Make sure the two SCSI controllers that the RAID box is hooked up to are on different I/O boards, so that if one I/O board dies, the other is still there. And of course make sure you have at least two separate CPU/Memory boards, so if one of those dies, the other is available.


  • Will they use the Cobalt Chipset?

    On a server? Are you kidding? What on earth for?

    Will they be cutting back on the Video and Audio abilities?

    If SGI follows past practice, then I would say yes, but only as applicable to making the server a functional workstation: SGI servers have almost always been headless. Video out on a Challenge (and most of the Orgin series) was a port for a dumb terminal.

    But for media serving, I would imagine it would be more than up to the task.
  • If you have any familiarity with SGI, it's half dead. People are leaving--best people.

    If you have any familiatity with SGI, you'd realise they posted $158 million *NET* profit for last quarter.

    Strange behaviour for a corporation which is supposed to be half dead. Perhaps someone forgot to tell them this?

    Heck, SGI can't even get their Oracle database that holds customer data and all the servicing information and all to work with a reasonable latency (say, give a response consistantly under one minute).

    So who the hell wants to run an Oracle database on an SGI? Oracle databases are boring grunt work fit for Suns, not high-performance SGIs.

    Seriously, I suggest you acquaint yourself with MediaBase, an app that finds itself more than suitable runnging on an SGI server.

    Another poster suggests that they are not running *BSD because of driver & applications issues. I do not think it is relevant. The decision to ship Linux boxes was not made by a technical person

    How doe we know this? Granted, FreeBSD is open source, but I think it's far too likely that most of the work had been done already--on the engineers' own time with Linux. This is mighty fast turnaround time for a server-class computer for SGI.

    I figure time has the most to do with it. Eventually, I'd expect FreeBSD to be up and running on one of these servers.

    Face it: They're not shipping Linux because Linux is good. They're not shipping Linux because of any strategic alliances (they have none at the moment, the industry bails out, the place where I work throws away Challenge XLs--have you ever seen a Challenge XL?).

    *horrified gasp* Are you kidding! Are they insane? Oh, the *humanity*!

    Save me one! I'll pay for shipping! Those are fabulous machines (consume more power than godknowswhat, but *still!*)

    In closing, however: I find your comments largely flawed and irrelevant to SGI, a company which has performed miracles both technical and visual (and now, coporate). People have been insisting that SGI is dead for nearly two years now.

    I think it's time for an acceptance of reality: SGI thrives, and you are quite simply wrong, and have been wrong all along.

  • Yes, well, but if one needs binary comaptibility with x86, it's easier to have a box built around same.
  • I think one of the main reasons behind SGI's recent embrace of x86 is it's compatibility with existing software. It's expensive to maintain a hardware and software division (just ask Apple) so in going with x86 they can put on existing software with basically no porting. Porting IRIX to x86 would cost them alot of money and probably wouldnt pay off in the long run, but using linux and NT, operating systems more people are familiar with, they stand to make a much larger profit from these boxes, because their investment is almost entirely in hardware, with a few linux patches to get their machines running well. Supporting NT means that 3,000 dollar program suite you bought for your old systems can be used on your brand new SGI machines, supporting linux means you basically have a support base of thousands of developers and gurus. You dont have either of those with IRIX.
  • too bad oracle hasn't released oracle for linux on alpha!

    I stand ready to put my foot in my mouth if this has changed, however!
  • Different computer.

    Notice the lack of the cobalt chipset, for starters... it is possible for a company to ship two systems based on the same CPU and have radically different machines in the end.
  • Disclaimer: I know little about the Linux kernel. This stuff might all be nonsense.

    What if any new functionality has been implemented by SGI as modules, loaded on demand by kerneld? They can be shipped as binaries only without violating the GPL. Since they're not techically part of the kernel, they don't have to be covered by the same license.

  • Correct me if I'm wrong, but if you write let's say a kernel module or an X driver that doesn't use any GPL code, then you do NOT have to give the source away....

    There is such a module (or patch I don't remember well) for the diskonchip device.


  • Well, it really depends on how it all was implemented. If they modified existing GPL code, they'd have to give it away. On the other hand, if they wrote a bunch of new stuff that plugs in to the existing framework, they wouldn't necessarily. If it is enough separated to qualify as a separate program, they could release it under any license they wanted to.
  • by larien ( 5608 )
    I guess this partially explains why SGI is giving XFS to linux; the kind of need it themselves to allow linux to have some of those features as part of their corporate strategy.

    Not that I'm complaining; any improvement to linux is a good thing. With the resources they have, and some of the development work they have under IRIX, linux should come on in leaps and bounds.

  • I guess there shouldn't be a post asking if this runs Linux. :)

    But I suppose someone will metion an ancient mythical story that was redone from the moster's point of view.....
  • The probably wouldn't waste the time releasing drivers as binary.

    Fortunatly binary modules often break with new kernel releases making binary modules a royal pain for any company stupid enougth to use them. Thus they are likely to release it as source.

    Then again, the only thing that can be expected form large companies are stupid decisions. Anything better is just a bonus :)
  • i noticed that the "consultant" they interviewed seemed to think that linux didn't fully support SMP. I thought it did. Was I wrong?
  • by skaya ( 71294 ) on Thursday July 29, 1999 @03:51AM (#1777756) Homepage
    Chances are, SGI servers running Linux offer hotswap capabilities thru dedicated hardware. But it's also possible to use software only solutions. There are a few requirements :
    - the hardware must handle it, I mean, you must be able to add/remove a drive to a IDE bus/SCSI chain without everything going mad. Most hardware I met doesn't care if you remove a (umount'ed) drive.
    - the software must handle it. With IDE drive, there are ways to force Linux to redetect hard disk geometry (with 2.2 kernels, use modules, with 2.0 kernels, there are unofficial kernel patches to do that. check for an ugly patch allowing to do hotswap with your secondary IDE channel with 2.0.36 kernels)

    If you want really good performance/reliability, go for RAID-5. Hardware support is not required anymore. Linux supports software RAID-5 since quite a long time (with the appropriate raidtools), but on-the-fly reconstruction is a recent update. And if you want things like LVM ("oh dear, my 80 gigs pool is full, nah, just add another 18 gigs scsi drive, and poof! I have 18 gigs more free"), you will have to play with latest 2.3.* kernels. It's a domain where Linux hasn't reached (yet) the level of others like HP-UX, but it's improving (it's a big work, because there are filesystems consideration underneath - how do I resize an ext2 filesystem, etc)

    A last note about SMP Linux boxen : according to Alan Cox if I remember well, Linux scales very well to 2 CPUs, poorly to 4, and not at all to 8. To solve that, give 'em linux coders octo-xeon servers to play with, I promise they will do their best :-)

  • Can you say I/O bandwidth!!!

    You can be sure that SGI has some more than a
    100 Mhz bus connecting the CPU's! Aint something
    that Intel is good at!!
    Also, why the hell dont they make a move to the
    EV6 with the k-7's... that will surely rock the house!
  • Let's settle this silly SMP debate once and for all: get one of these machines to run NT and Linux (or swap out hard drives, I don't care how it's done as long as it's fair...) and test SMP. Preferably with the same applications, but whatever measures of SMP performance that can be done on both NT and Linux... (of course Povray, maybe web serving or whatever apps are sexy this week...)

    I know that Linux doesn't have some of the cool features that IRIX does, but it should soundly whip NT, since NT 5 *might* have some of the essential features that UNIX has had forever... (note to dissenters: sure, I'll give you examples if you like...)
  • I can't find any pictures with the story. Is there anything on SGI's site about this?

  • According to this person, (Surprise!) Microsoft insisted on language to the effect that SGI wouldn't support a competing OS environment on the same hardware. SGI's strategy for getting around this is supposed to be to focus their Linux efforts on things that can be branded "servers" until the agreement runs out.

    As much as I'd like to believe this, I think that contract provisions such as his would have come up in US vs. Microsoft.
  • Uh, it's an SGI box, with Intel CPU's... The only one I know of that they now manufacture is the "Visual Workstation."

    I would need to see some significant proof to believe that this hardware is substantually diffrent than the existing Visual Workstation they sell running Windows NT.

    No one has mentioned that they have built a totally new hardware configuration, based on Intel CPU's. Therefore, it's only reasonable to conclude that they are mearly installing Linux on the Visual Workstations, and calling it a "server."

    The differance between a server and a workstation is in the eyes of the beholder. If you think just calling it "server" makes it diffrent, I would have to point to the countless users who run httpd, ftpd, nfsd, etc on thier "workstations." But, to me, it's neither, it's just a piece of hardware. What they do with that hardware doesn't change the fact that I believe that the hardware is the same.

    So, even if it's the same as a Visual Workstation (which is just a "product name," and doesn't automagically make the product incapable of acting as a server), but even if they took that box, and yanked the video card out, it's still the same hardware, and Linux is know to run on it, and X doesn't run on it yet, so, I just don't see anything new here.

  • Hot swapping drives is more of a matter of hardware than software. All you need is an sca drive and a good back plane. Now what would be useful would be if they were seen as removable media. Otherwise you need a SAFTE backplane and a RAID controller to get much use out of it.

    Of course you could what I do when I want to remove an external jaz drive with out powering down. (Note don't try this if you have anything else attached the scsi card.) It works well for hot swapable drives, but I've not found much use for it in real life.

    from scsi.c
    * Usage: echo "scsi remove-single-device 0 1 2 3" >/proc/scsi/scsi
    * with "0 1 2 3" replaced by your "Host Channel Id Lun".

    * Usage: echo "scsi add-single-device 0 1 2 3" >/proc/scsi/scsi
    * with "0 1 2 3" replaced by your "Host Channel Id Lun".

    Note: You really shouldn't be doing this so don't blame me if it goes wrong. Not all removable drives are hot swapable, and not all hot swap drives/carriers are created equal.
  • About Cobalt Chipset (from SGI):

    The bus connecting the memory controller and RAM moves data to and from main memory at an astounding 3.2GB per second-six times faster than an AGP 2X graphics bus.-- This would be very good for a server as well.

    Combining the most advanced graphics engine available for Windows NT, unheard-of memory bandwidth, and a multiprocessor interface, the memory controller ensures that your most critical data is on the shortest possible path.The tightly coupled Cobalt graphics engine performs lightning- fast 3D geometry, sophisticated shading, lighting, hardware-accelerated texturing, and pixel fill.-- Clearly, they are selling the advantages of thier chipset as "better video preformance." But, when you increas memory I/O rates, and general communication bandwidth in a system, that's going to help out a lot for servers too. Intel Pentium III Xeon Processors Up to four Intel Pentium III Xeon microprocessors provide the pinnacle of Intel processor performance. Your Silicon Graphics 540 workstation features the fastest Intel Pentium III Xeon processors available, with performance-enhancing features such as a dual independent (cache and system) bus TM architecture, dynamic execution, Intel MMX multimedia technology, and a closely coupled Level 2 cache bus running at the full speed of the processor, with cache capacities up to 2MB. -- Clearly, they have a 4 CPU Intel Xeon system, called the "Visual Workstation" that comes very clsoe to thier new Linux/Intel Server claims.

    This is all why I suspect that it's the same. Now, there are some features that the Cobalt chipset gives them that will help a server, and it's a lot of work to develop a chipset for a system at this level, so I doubt they will be using something diffrent for the new "server." But at the same time, there are a lot of video and audio subsystems in the Visual Workstation that they could probably remove. And, the focus could be shifted to network capability and away from Video.

    So, I guess, after review, the real questions are:

    • Will they use the Cobalt Chipset?
    • Will they use the same Motherboards?
    • Will this simply be "adding a drive sled bay" to a visual workstation?
    • Will they be cutting back on the Video and Audio abilities?
    • Why does the price mentioned seem higer than the Visual Workstation (if you just adding a sled, but taking out all the video and audio stuff?)?

    So, coming back at this, I have to think your right, it's a diffrent box, because if they claim "hot swap" they probably have sleds for it. (SGI sleds are WAY to expensive BTW). And, given that one significant differance will probably be in the case, it makes me wonder what the other differances will be... because there may well be many, this might be totally new box, and I sure wish I knew what the specs really were.

  • Gotta remember to preview... hate it when I forget to put the bold off tag in.. only meant for the first line to be bold :P
  • by jd ( 1658 ) <> on Thursday July 29, 1999 @03:57AM (#1777768) Homepage Journal
    I found the critisism at the end of the article odd. Yes, there -are- plenty of other Unixes. Of those, most are expensive, without source, and impossible for SGI to customise to their hardware.

    In order to sell at even close to the same price as the Windows NT version of their server, they would need to use either a -VERY- cheap Unix (and most of those aren't scalable, high-end, or offer high availability), OR one of the *BSD's, which, for all their merits (which are considerable), don't compete with NT in terms of driver support or software availability.

    IMHO, SGI's powers-that-be are no idiots. I believe they picked Linux as the optimal OS to sell alongside NT. That doesn't mean "best" at X, Y or Z, it means optimal, when considering ALL the factors SGI would need to take into consideration.

  • I've worked with Compaq/HP/Dell servers in the past, and I can tell you that when you're doing a RAID-5 on the majority of these machines, the "control" of the array is handled by BIOS/HW systems, not the OS. You run a system partition utility to create the RAID-5 stripe sets, and controll it from there. The OS doesn't know about how the HW is set up, it only sees one contiguous partition, as provided by the HW controller. This may sound a little wierd, but the speed is (IMHO) usually fater by running through the HW rather than the OS.
  • First of all, hot swap, isn't nescessairly new. Lots of mylex controllers support this and you can buy boxes from VALinux that have this feature. I believe that the VARServer 3500 is one of them that comes with a raid 5.

    As for SMP, it baffles me the comment about linux not having full SMP support. I run Oracle on 4x Xeon at work (with the parallel options installed of course) and it works wonderful. Giving queries the /*+ parallel */ hint makes them about 3 times faster or so. Starting multiple SQL*Loader sessions imports the data about 3 times as fast also. I'd imagine that its waiting on the disk in these cases as we unfortunately only have one disk controller in this system. In any case, odds are we'll have to upgrade out of linux sometime in the future if it doesn't get support for 8 or more processors and files larger than 2gig. Which according to Linus probably ain't gonna happen soon.
  • > you misspelled 'fortunately'. hope this helps.

    So! People just don't get that spelling isn't really all that important, as long as it is readable. I don't have a spellchecker for posting, so there may be a few mistakes. Don't waste time pointing it out!! I want peoples opinions, not my spelling mistakes.
  • by BadlandZ ( 1725 ) on Thursday July 29, 1999 @04:02AM (#1777772) Journal
    Well, Linux runs on an SGI Intel Box.. Woo. Forgive me for not doing cartwheels.

    This looks like a nice public relations press piece that SGI was happy to get out. Server? Thier Visual Workstation billed as a server? Sure sounds to me like they don't have X working yet.

    We knew it would make a good Linux box last year when we saw the memory I/O, and chipset, and all the fancy hardware. Nothing new there, that happened last year. We knew Linux would run on it when kernel 2.2.0 came out and had Visual Workstation patches in it. Nothing new there, that happened last winter.

    So, now, we get "news" that SGI has an Intel Linux Server... If that means they are just now getting it installed right, that's sort of slow progress, considering how long ago the first reports of it running were.

    Maybe I am just in a bad mood this morning, but, what this story says to me is "Well, we still haven't got X to work, we promised OpenGL, and we don't even have X working yet... We can't get enough people working on it, it's taking way too long, what can we do? I know, let's get some press on it running Linux again, call it a 'server,' and maybe we'll look like we did something good."

    Don't get me wrong, I like that SGI is supporting the Linux community. I just don't see any big reason to get excited by this specific story.

  • First if you unmount a scsi drive with an SCA connector you can pull it out hot, raid or not. This is not something new.

    Now my main point I want to make is that the RS/6000s running linux is a much bigger deal. (Yes linux is running on RS/6000s) How does 32 processors sound? Hot swappable processor/memory/drives/cards? RS/6000s will autodetect bad processors and turn them off. When you come into work and your computer tells you processor 5 has been turned off and you replace it, add some ram, and put a new video card in all without ever rebooting your computer that's impressive.

    Reasons to reboot a RS/6000:

    1. You are moving it to another state
    2. All of your processors failed at the same time
    3. Your UPS just caught on fire
    4. Oh wait, there are no more reasons

"I have not the slightest confidence in 'spiritual manifestations.'" -- Robert G. Ingersoll