The collected works of rosmaniac

Once again, thanks to Voralyan's howto, here's the second Altix box on Debian 6. This one is a tad larger; there are 13 compute nodes, two of which have PCI slots and IOC10's (SATA drives instead of SCSI), and it's running with a single-plane router. The hwinfo and dmesg follow.

And pics.....

Update:

After working for a while on the additional two Altix 350 compute nodes from this system that had issues, I've now populated this system to 15 compute nodes, making 30 CPUs, and 54GB of RAM.

I had updated the PROMs on the 13 working nodes, but not yet on the two non-working ones. A PROM version mix, as has been noted elsewhere, will cause the system to enter POD, before ever loading EFI. And on compute nodes without disks or an IO9 or IO10 (there are three in this system with IO10's, but neither of the repaired nodes have any IO or disks) booting a node outside the system and updating it individually is not an option.

However, POD has the tools to fix this. So, if the system drops you to POD due to a PROM version mismatch, there is hope. In POD, you can flash the downrev nodes by nasid; you can get the nasid in POD by (note that I entered POD from the EFI shell for this example):

Suppose that brick 001c26 was at PROM 4.43, however. In my output above, I find that the nasid is 22 for 001c26; I can program its PROM from POD by:


Then I can exit POD and reboot. As a note, if anyone knows where a POD reference document can be found, that would be cool.....

Also as a note: the Altix 350 router brick has the same embedded L2 capability as the 3700BX2 does; all that's required is to plug in a 'USB to Ethernet' adapter into the L1 USB port on the router. Finding what kind of 'USB to Ethernet' adapter was difficult, but I finally found and just tried a 3Com 3C460B that worked right away. Simply plugging in the proper USB to Ethernet adapter into the L1 USB causes the L1 on the r-brick to load the L2 emulator, access to which is Ethernet only. You can set the IP address from the L1; the instructions for doing this I found, ironically, in the 3700BX2 documentation and not the Altix 350 documentation. It appears that the same NL4 router is used for Altix 330 systems, and maybe even for the same vintage Origin systems; if your NL4 router has a host type A USB port labeled 'L1 USB' it may have this capability as well. Altix 330 systems appear to have this capability on the compute node, too, so the separate L2 controller isn't required for these systems to get L2 functionality.

A reboot_l1 seems to be a good thing to do after plugging the adapter in; the L1 serial port stays an L1 serial port and can't access either the L2 prompt or the system console (have to use one of the other compute node's serial ports for console, but have to use the router's L1 to power up the whole system, since the router is the only brick with full L1 connectivity in a routed, non-ring, configuration. In a ring config, the L1 at any node can see all nodes, but not in routed mode, where node visibilities from the compute bricks' L1's are, to say it bluntly, weird).

But the emulated L2 works great once a supported USB to Ethernet adapter is found. Again, I found that a 3Com 3C460B works fine; I haven't tried any other USB to Ethernet adapters. Would be interesting to see what others work.

The box is still running Debian 6, but as soon as I can get CentOS 5 to rebuild it's gong CentOS, since the ProPack and Foundation tools will work there, and I can rebuild system updates from source (can't do that with SLES) can keep the OS updated (with ia32el, and i386 binary capability) at least until 2017, the EOS year for CentOS 5 (and all RHEL5 rebuilds).

Hmm, the BYTE UnixBench is still around and updated? Cool; I remember running a much older version on my Tandy 6000 running Xenix System III back in the '80's (that box was a 68000 CPU at 8MHz and 1MB of RAM with two 70MB MFM hard drives and a pair of 8 inch floppies; but it was real Unix!). I most recently ran the 4.10 version a while back; didn't know it was a google code project now.

Right now I am waiting for a second 64 process run of the HPCC Challenge benchmark to finish; the hpcc package is distributed in the Debian repos, so it's an easy install and a pretty simple configure. It is fun to watch all 30 CPUs hit and stay at 100% for half an hour. (Wall time: it is really strange to see something like:

as output of 'time' though!)

The HPCC benchmarks make a good stress test; I know this box isn't going to break any records, but since I'm going to be using it for planetarium graphics rendering I need to stress it pretty hard before putting it in production. It is also somewhat disquieting to see the load average parked at 64 and the system still in a responsive and usable state.... A really large Altix (say a 512 processor 3700) would be really interesting to see, but I wouldn't want the power bill. This Altix 350 is one compute node (and one router) shy of a full dual plane max config, so is a good indication of how powerful such a maxxed out box would be.

In any case, UnixBench doesn't seem to like >16 CPU's..... I'll see what I can do to fix, perhaps. I'm also running a comparison on an 8 CPU Xeon system; two quad core 2.0GHz 64-bit Xeons, running CentOS 6.3 with 16GB of RAM (Dell Precision 690 workstation).

I will report back.

Ok, some results. First, the 30 CPU Altix 350. Since UnixBench is a system benchmark, the slow SATA drives do somewhat lower the total score (and I've snipped some repetitive sections out):


Now, the Dell Precision 690, two quad core Xeons (E5335's) at 2.0GHz:



There are some areas where the Altix shines, especially considering the age difference and the much faster drives loaded in the 690. But that gives an idea of system speed differences.

I'm working on getting the 20 CPU Altix 3700 running the benchmark, but a script is tossing an error, so it hasn't happened yet.

And, for grins and giggles, a dual-processor 2.4GHz 32-bit Xeon (Dell PowerEdge 1600SC):



The Altix fares quite well against the NetBurst core Xeon.

The Altix systems are both at least as responsive as the Dell 690 8 core, even for remote ssh-tunneled X programs. The 690 is going to be used for virtualization, but was available for a quick benchmarking before getting KVM guests going....

I'm overall quite pleased at the responsiveness, even under heavy load, of both Altix boxes. The old NetBurst box is definitely less responsive under load.

And the Altix boxen are destined to do MPI-friendly rendering tasks, so they will do quite well indeed.

[edit] Also, I have a Sun Enterprise 6500 here that I might can run the benchmarks on at some point, but it will be a while. The E6500 has 18 CPUs and 22GB of RAM. Will be an interesting test.

I just went through a release update on the maintenance stream from 6.5.19 to 6.5.22 on my been-a-long-time-since-last-boot-because-I-was-busy-playing-with-the-Altix-350-beast Indigo2 R10K. Inst completed successfully, and I rebooted appropriately. However, uname -R returns:


Going back through inst, it seems like the overlays installed properly (they show as S instead of U in the list inside inst, and they showed as U the first run through inst). The install was from 6.5.22 tarballs downloaded from supportfolio while 6.5.22 was still available.......

My install media is the 6.5.19; should the uname -R show me 6.5.22, or am I missing something?

5086? Yes.

Ok, thanks to user hamei I got pointed to the correct direction.

For whatever reason, /unix was left in place and was the 6.5.19 version, with the 6.5.22 version as /unix.install. Doing a cp /unix /unix.old and then a cp /unix.install /unix seems to have done the trick, and the system now reports 6.5.22m to a uname -R.

It might have had something to do with the Phobos E100 drivers that I had started loading years ago (yes, I do have the E100 in the machine, but it wasn't being recognized as fe0 until I pulled /unix.install over correctly).

Ok, thanks. I am a bit speechless as to how long ago I had originally started the install...... I bought the 6.5.19 in February of 2005, and the I2 a week or so later ($111.48 plus shipping from eBay seller sunnking....), and then installed IRIX in March 2005. Last boot was, yep, you guessed it, 2005. Yow. So I picked up the install a bit less than eleven years after I started it and pulled the IRIX 6.5.22 overlays down from supportfolio.....

And the beast just started up like it was supposed to, at least with a serial console (my 13W3 to VGA adapter and my KVM/monitor don't like each other, but I have an SGI monitor I can hook up.....). Hinv to the appropriate place.....

Thanks all!

Whee. Eleven years after starting to work on this machine..... finally:





Machine was purchased to do Audio DAT downconversion....... just never got around to it until now. Thus, the Archive Python DAT drive.....

Can you say 'PhantoM?'

Thanks to everyone who has mirrored things.

robespierre wrote: I think it's an aftermarket CDROM drive, they shipped with Toshiba XM-3501B units.

It is; the Plextor 12X drives are settable to 512 byte sectors, and I happened to have fifty or so at the time.... I'm not sure how many I have right now.

It did come from the seller with a Toshiba drive, but it was broken, and I had the Plextor drives......

The Archive CTD-8000 DDS2 were stock.
Let the forum know if you have success with audio DAT transfers. I found that the DATplayer software really liked to hang.

Thanks for the notice. I have the DATgoodies tarball, but haven't built them yet (I did grab the tardist of DATgoodies from somewhere else, but I've always been one to build things myself if possible). And while there is software for Linux and other u*ix variants out there, it is rumored to be less stable than the SGI stuff was, so for now it will be DATman. I also have an O2 that I'm getting up to 6.5.22 also, but it will have to be an external.

I will definitely follow up on the DAT transfers once I get some more round toits to get things set up properly.....

Ok, I thought I'd share this with everyone, since it made things much faster for me.

First, note that many modern Linux distributions do not build or bundle the efs filesystem anymore. I am running CentOS 7 on my main box, and followed the instructions at https://wiki.centos.org/HowTos/I_need_the_Kernel_Source and then https://wiki.centos.org/HowTos/BuildingKernelModules to get a working efs.ko, which I put into place the normal way. Setting EXTRAVERSIONS in the kernel Makefile was required for my build; YMMV. I am an old hack at building RPM's, but I haven't built a kmod-efs as yet.

Ok, so once a properly versioned efs.ko was in the correct place for the installed kernel to find it, the following commands, as root, will load it:


Now, I had previously imaged my IRIX distribution disks, and had all of them in one directory, named with a '.image' extension. I then created a mount point, mnt:

Here's the one-liner that, in a little over two minutes, exploded all the files out of the 22 disk images I have:

(yeah, I know --progress is just eye candy..... but this is an IRIX forum, no? ) and here's the resulting directories with their sizes:


I hope that helps someone.

Now, to set up a DINA VM on my CentOS 7 box using KVM......

Additional information:

I did a fresh install of 6.5.19 on one of my O2's, and upgraded to 6.5.22 using the same procedure. The same thing happened as with the Indigo2. But this time I was expecting it, and knew what to look for.

On the O2, I used the GUI Software Manager, and let it do an automatic install of the 6.5.22 overlays.

robespierre wrote: Archive.Org seems to skate under the Fair Use defense for everything that they do. Mirroring web pages without prior permission is not sanctioned by the copyright laws either.
They have a project to upload and share every CD-ROM ever published.

It would be nice if they would get 6.5.30 up.... I'm missing a couple, and one is scratched (my youngest son used to think CD's were playthings to throw, and while I did let him sail some 8 inch floppies that had nothing useful on them, he decided that Daddy's box with those 8 inch floppies was just full of toys..... ). I have a 6.5.24 overlay set on the way from an eBay seller, but was hoping to get up to 6.5.30; and the eBay sellers who want $425 or $700 for a set..... well, I can't swing that. Oh well, it is what it is.

While over the years there have been several threads here about the best version of IRIX for different platforms, in the reality of the new post-IRIX supportfolio and the complete non-availability of patches and even the 6.5.22 maintenance release from SGI, what would the brain trust here consider the 'best' version of IRIX for these machines to be?

If I were a new SGI hobbyist and I had just scored an O2 on eBay or at the flea market or on craigslist or wherever, and there was either no hard disk or a blank hard disk, what would my best IRIX option be, assuming several were available? Is it 6.5.22 since that is easily found, and thanks to the internet archive the otherwise difficult to find foundation CD's are available for new installs? Or should someone try to find something later, perhaps even 6.5.30 (even though the CD sets are expensive through eBay and probably not even available through SGI any more)? Are there advantages to 6.5.22 over 6.5.30? (I've heard a few already, but I'd like to see the knowledge pooled given the current IRIX situation, and links to SGI's site are lilkely to not be valid soon enough.......the http://www.sgi.com/products/software/irix/releases/ link, for instance, is no longer useful.

Having said that, if one wanted to get into development, has anyone set up a development environment without using the IRIX Dev Foundation and the MIPSPro compilers using any version of GCC? Specifically, in my case, writing automated tools for pulling audio from audio DATs and stuffing them into properly chunked and named WAV files? I have the compilers that came with 6.5.19 AWE but something more modern would be nice, and the GCC suite is well-tested, if not as optimized.

And I'm well aware that likely the 'best' OS option for the hardware today , if I want to do general things and not 'IRIX-y' things like pulling audio from DAT, is the latest OpenBSD, but that's not IRIX.....

foetz wrote: [ setting up a dev suite without the IDF ] won't work. even for gcc or whatever other compiler you need the basic dev stuff for the os which would be the dev cds except for the actual mipspro.

First, I appreciate the response, and all that you have done for the community here over the years.

My starting point is 6.5.19 AWE, so I do have the IDF and associated items (MIPSPro C and F77 7.2 along with the 7.2.1.3m maintenance release. But I see here that building Nekoware packages at least needs 7.4? (not that I'm planning yet on anything like that; but I did build PostgreSQL RPMs for Red Hat-ish distributions for 5 years long ago)...)...

the GCC suite is well-tested, if not as optimized.

it's neither as recent tests from some guys here including myself have shown. if you have a mipspro stick to that.

Is that just GCC for IRIX, or GCC in general? I'll look for the threads....

if I want to do general things and not 'IRIX-y' things ....

no point in running something on an sgi that runs way better on x86. that aside with anything but irix you can run nothing that was made for irix which is the very point of an sgi ... you get the point, without irix an sgi is just an old box.

Exactly.

as for your actual question, with an o2 you can only choose between 6.3 and 6.5.x so unless there's something special you have that requires 6.3 the answer is quite obvious

I was meaning something in 6.5.x, and I should have mentioned that in my post..... my bad. It seems the consensus of previous threads was that 6.5.21 was the one to have, with .22 a close second.

Thanks for the information!

Nice. I've used MESS/MAME for a while to do TRS-80 Model II development.... yeah, really old. However, it is very processor-intensive to emulate all the hardware in a TRS-80 Model II; it would be interesting to see if my O2 would emulate a 4MHz Z80 system with all the support chips like a Model II.....

I see a 0.149 MAME on IRIX contributed by axatax, but I fear getting the latest 0.171 might be a real feat. I build it on CentOS 7 with the help of the Software Collections' DevToolSet-3 compiler chain. It would be a real treat to run IRIX that way, but that's getting pretty complicated, especially given the internal architecture of MAME.

Linux GUI support is pretty broken on SGI, from what I remember from my research a while back. Linux is basically completely broken on R10K on O2 and not especially well-supported at all, at least the last I looked. OpenBSD is a better non-IRIX choice on that hardware (and I am primarily a CentOS Linux user, with a grand total of one BSD machine around.....), although O2 is better supported on the GUI.

But it has been quite a while since I looked; YMMV and all that, plus people do change things and ports come and they go. The core of the Linux MIPS stuff has been for non-SGI MIPS stuff, especially routers. But there is some good info, if you're into Linux on MIPS or want to be into Linux on an I2, at https://www.linux-mips.org/wiki/IP28

Part of the fun of retrocomputing (and SGI computing is at the point of retro, now) is getting the hardware to do things it was never originally designed to do, or in my case of getting to play with hardware on the cheap that used to cost kilobucks (kindof like getting to play with a VAXstation 4000/96; at one time that box was the bomb). So if you're able to get a usable GUI (or CLI if that's your thing) Linux of BSD working like you want, kudos.

dexter1 wrote: ...
If you get an Indigo2 like Foetz suggested, get a fairly decent system with at least an R10K and Solid impact GFX. That doesn't have texture support, but should be good for programming and modelling. But if you want to run Debian and Gentoo i am not sure if there is graphic driver support for Impact systems, ...

The Octane port supports Impact; the I2 port may or may not at this point (the online docs are pretty old in this area). This is for console framebuffer; X support is (or at least was) either experimental or nonexistant.

I had looked into running a Linux on some Teal Indigo2 Extreme's here, but then I found the stash of IRIX 5.3 CD's that came with the machines, and that might be a better choice for them; there are better Linux machines out there, and IRIX is a far better choice on R4K hardware, IMO. My R10K Indigo2 with SolidIMPACT is running IRIX 6.5.22m, and performances is pretty good (I have 512MB of RAM in it, and a relatively fast hard drive).

electrithm wrote: Wait so you're saying I wouldn't be able to run Gentoo with the MATE gui on the Indigo?

To the best of my knowledge, and I always reserve the right to be wrong, no, you would not be able to do this. Linux is not NetBSD, which really does run on everything (command line, at least). Gentoo's MIPS hardware requirements page is at https://wiki.gentoo.org/wiki/MIPS/Hardware_Requirements Indigo2 R10K IMPACT is listed as very experimental (italics is on their page.....). O2 R10K is unsupported entirely. Graphics support is minimal at best, inoperative at worst. Serial console recommended for anything other than Newport (XL) framebuffers.

That sucks because I was gonna use it for playing around with SDL. I'm not gonna do very heavy modelling on IRIX, just some small little animations and making 3d models so I probably wouldn't need a very heavy duty SGI.

Use IRIX if you're doing SGI. It was built for these systems, and it runs well on them. SGI didn't open up any of the documentation on how to program the various 3D graphics pipes (there is documentation on Indy, and that's pretty much it), and so without the hardware acceleration, yes, things aren't going to be fast. With IRX, things are usable.

FWIW, I decided to stick with 6.5.22, which overlays I have from my Supportfolio download long ago, to go with the 6.5.19 CD set (AWE set) I also bought, long ago.

It would be interesting to see if someone was successful in doing a scratch install with only what is up on archive.org. I may try that myself on my I2/R10K, once I find another drive that will work in it.