How to Incorporate an Origin 300 In Homelab?
#1
How to Incorporate an Origin 300 In Homelab?
Hello, I hope all is well.

I recently acquired an SGI Origin 300 this past weekend (drove from Maryland to Connecticut), in addition to an SGI Altix NL4R router, and an Apple Xserve with a single Xeon.  I have no idea how to run the router (I probably will never be able to, as it uses DC power, and requires a couple Altix units I believe).  The Xserve is no issue, I just need to install OS X Server.  I already purchased a 73GB IBM-rebranded Seagate SCSI drive (80-pin, Ultra160) for the Origin 300, and I think I have some idea on how to install IRIX.  I'll cross that bridge when I get there.

My main question for now is the historical use-case of the Origin 300.  How did people run calculations remotely on the Origin 300, or even use an Origin 300 remotely (over network/ethernet, and not serial)?  How were Origin 300s usually used?  My home lab is relatively small, I have a R730xd as my main server that I use for webhosting and VM provisioning.  I have an IBM X3620 M3 for backups, and the rest of my stuff is pretty old.  SPARCserver 5, Sun Blade 2000, IBM Netfinity 5000.  I have an idea of the use cases for those older ones (e.g. small intranet sites, COBOL environments, FTP), but I want this Origin 300 to be special.  I want to interface it with my personal website, or perhaps an intranet site hosted on my SPARCserver or Blade 2000, and somehow be able to run calculations remotely.

I could really use some ideas or resources I can use, especially for requisite software.  I haven't used IRIX yet, so I have no idea what that will be like.

Thank you in advance!

   
jander31
O2

Trade Count: (0)
Posts: 12
Threads: 2
Joined: May 2025
Location: United States
Find Reply
05-01-2025, 01:26 AM
#2
RE: How to Incorporate an Origin 300 In Homelab?
Reading about it, it looks like it could combine multiple sgi origin 300 as a single machine with with NUMA Link cables. That way it would probably behave as a single bigger machine and you could just run a multithreaded application or multiple compute cores on what looked like a much bigger machine.

It would likely be connected to a big storage array for input/output data and the network would be used for control. IPC accross nodes would happen over the much faster NUMA Link.

You would likely use something like MPI to write the distributed compute tasks but there were different frameworks for scheduling and controlling HPC jobs across multiple nodes over the network.
(This post was last modified: 05-01-2025, 10:29 AM by mosca.)
mosca
O2

Trade Count: (0)
Posts: 18
Threads: 2
Joined: Apr 2025
Location: Norway
Find Reply
05-01-2025, 10:23 AM
#3
RE: How to Incorporate an Origin 300 In Homelab?
Iam a owner of a o300 and o3800, o2000 and countless o200 Smile But havent them powered up since 10Y and nekochan died.

Its a normal server.... and its a headless server. If a GUI is needed than you can export your screen via X-Server.

Yes you can add a GFX Pipe to certain models.

Based on your LCD its the first modell wich is good. Yes you can expand the system by connecting multiple o300 via CrayLink together.

Regards,
Joerg
Origin3k
O2

Trade Count: (0)
Posts: 8
Threads: 1
Joined: Sep 2015
Find Reply
05-03-2025, 09:03 AM
#4
RE: How to Incorporate an Origin 300 In Homelab?
I've been busy all week so I've not had the inclination to respond with a quality response.

In absence of getting multiple origins together, your best applications for it are gonna be as follows:

Retro styled intranet site. It comes with sgi_apache or else I would honestly recommend a more modern web like thttpd.

NFS server. IRIX has a very good and easy to use NFSv3 implementation (it has no nfsv4) and you could run it as a server for that.

Beyond that there is also the option of running things like naemon/nagios or whatever but you will more than likely have to build that stuff because the stuff that is out there is pretty ancient.

I'm the system admin of this site. Private security technician, licensed locksmith, hack of a c developer and vintage computer enthusiast. 

https://contrib.irixnet.org/raion/ -- contributions and pieces that I'm working on currently. 

https://codeberg.org/SolusRaion -- Code repos I control

Technical problems should be sent my way.
Raion
Chief IRIX Officer

Trade Count: (9)
Posts: 4,240
Threads: 533
Joined: Nov 2017
Location: Eastern Virginia
Website Find Reply
05-03-2025, 02:31 PM
#5
RE: How to Incorporate an Origin 300 In Homelab?
I got my Origin 300 to boot to L1 (bought an appropriate console cable, DB9 to USB). A few L1 commands are unavailable, but I cannot seem to get past L1. FAN 0 (left) died on me, and I'm not sure if this is causing the "power up" command to not really work. Does anyone know what my options are for replacement fans (if this is indeed the issue)? I bought a 73GB SCSI (U160) for the server, and it starts up alright. I'm trying to do a network install of IRIX 6.5.22 via LOVE, and I believe I need to get past L1 to even begin this process.

This is what happens when I do "power up":

05/05/25 20:40:33 power up (COMMAND)
05/05/25 20:40:38 reset again MIPS
05/05/25 20:40:56 FAN 0 warning limit reached @ 0 RPM.
05/05/25 20:40:56 Environmental redundancy lost.
05/05/25 20:41:28 SMP-R: UART:UART_NO_CONNECTION

And here's "env":

001c01-L1>env
Environmental monitoring is enabled and running.

Description State Warning Limits Fault Limits Current
-------------- ---------- ----------------- ----------------- -------
12V IO Enabled 10% 10.80/ 13.20 20% 9.60/ 14.40 12.25
12V DIG Enabled 10% 10.80/ 13.20 20% 9.60/ 14.40 12.12
5V Enabled 10% 4.50/ 5.50 20% 4.00/ 6.00 4.99
3.3V Enabled 10% 2.97/ 3.63 20% 2.64/ 3.96 3.34
5V aux Enabled 10% 4.50/ 5.50 20% 4.00/ 6.00 4.97
3.3V aux Enabled 10% 2.97/ 3.63 20% 2.64/ 3.96 3.41
2.5V Enabled 10% 2.25/ 2.75 20% 2.00/ 3.00 2.50
Speedo2 CPU Enabled 10% 1.44/ 1.76 20% 1.28/ 1.92 1.61
1.5V Enabled 10% 1.35/ 1.65 20% 1.20/ 1.80 1.49

Description State Warning RPM Current RPM
-------------- ---------- ----------- -----------
FAN 0 LEFT Fault 2160 0
FAN 1 CENTER Enabled 2160 4155
FAN 2 RIGHT Enabled 2160 4207
FAN 3 PS Enabled 2160 3146
FAN 4 PS' Enabled 2160 4549

Advisory Critical Fault Current
Description State Temp Temp Temp Temp
-------------- ---------- --------- --------- --------- ---------
NODE 0 Enabled 30C/ 86F 35C/ 95F 40C/104F 24C/ 75F
NODE 1 Enabled 30C/ 86F 35C/ 95F 40C/104F 22C/ 71F

001c01-L1>cpu
command not supported on this brick type.
001c01-L1>
jander31
O2

Trade Count: (0)
Posts: 12
Threads: 2
Joined: May 2025
Location: United States
Find Reply
05-06-2025, 02:48 AM
#6
RE: How to Incorporate an Origin 300 In Homelab?
OK so this is very important, you have a last generation machine platform. This is the same basic architecture used inside Tezro and Fuel workstations, they are simply single node systems.

All of these systems have a series of monitoring chips in them collectively called environmental monitoring. These are an underlying system run by the L1 that ensures that the hardware is properly cooling, has the correct voltage within incorrect tolerance, and also performs hardware inventory checks and bootstraps the PROM. The PROM is what performs bootup diagnostics and is considered what you would call the BIOS in personal computer parlance.

There are three reasons you can have this message:

1. Your fan is actually dead, this is very easy to check, is your fan not working?
2. Your fan is actually running fine but the system can't tell it's running at all, this is a failure of one of the environmental monitoring chips if this is the case, the chip would need to be replaced, not impossible job I do it all the time, but not easy for someone who's not familiar with professional soldering.
3. Your fan is not running, the system knows it's not running, but when you plug in a new fan it still won't run, this would be a failure of the MOSFET at the area of the fan connector that literally gives the fan 12V power. Only reason this should occur is if someone plugged in a replacement fan not knowing that the pin out has changed and they destroyed the boss, requiring board repair.

Most of these systems have several stages of start up. First stage is L1 start up, you could also consider this idle standby power. When you plug in a modern SGI it immediately starts its L1, just buy pulling in wall AC power. Normally using the 5V standby power on the motherboard to trigger this process. The motherboard is not technically active during this time but the base level environmental monitoring firmware and inventory program is active. This allows you to use a terminal signal to communicate with the L1 to start up the system and also control how the system starts up, diagnostic switches, and such.

Stage 1 start up is the L1 attempting to power on the system, by starting the 12V power & 12V fans and checking environmental monitoring, Stage 2 would be the power on beginning of the bootstrap of the PROM if the environmental monitoring fully passes. L1 does not run on or take much use of the main CPU, it's a separate sub system and processor. The bootup of the PROM will test your CPU cash and all that junk.

This is showing you are reaching stage 1 and halting on error, and the system is detecting a hardware fault preventing you from going any further because going further could damage the board. This allows you to affect proper repair and continue on to see if there's any more issues with later on diagnostics to the board or any of its components.

All the fans in the systems were commercial fans available at the time the systems were made, however be very much aware that they're pinout were often non-standard. That means that if you get the exact same fan from a normal supplier that the pins are very likely going to be in the wrong order and if you plug them in may result in damage to the SGI or the fan, a simple re-pinout is often required, where you take the pins out of the connector and rearrange them to the correct order for normal connectors. Some platforms have a giant unified connector where you then replaced the fan and solder the old wires to the new fan instead of the opposite.

If you gently peel back the fan label you should be able to read what the wires/pads are and when you get the new fan make sure the correct lines go to the correct pads. Some fans manufacturers, over time, have swapped polarity, so don't just pay attention to position, make sure that the actual stencil marking on the fan PCB still matches as well!

Some of the fan connectors may be integrated into larger connectors, this makes the choice easier because you'll be reusing the original lines and desoldering at the fan body.

Assuming an actual dead fan and the replacement fan starts up just fine you will see if all errors are cleared. There is no command you need to use to clear old ENV errors, the environmental monitoring system automatically refreshes.

Technically there is a way to ignore the environmental monitoring system and proceed with PROM, no one here will advise you to do that. The only advice we give you in this scenario, if you were to tell us that you don't want to spend any money until you know for example that the PROM GUI can come up at least (BIOS SCREEN, NO OS). We would then give you information to bypass the environmental monitoring system just to see if you can get to the PROM screen. But after that fact we would strongly urge to fix the underlying problem and reenable environmental monitoring as it's the only thing protecting you from your old power supply going weird and destroying things or causing severe overheating and other damage to your vintage machine. We would also say that once you reach the PROM GUI you would be shortly to almost immediately shutting down the system to prevent overheating or other problems and then fix that fan issue.

In the case that you put a new fan in and prove it should be working and it doesn't work at all then you'll need to fix the underlying electrical circuit that powers the fan, which shouldn't be a huge deal but we don't do that every day so we'd have to have pictures of the board around the fan connector to help you troubleshoot. Normally that shouldn't be a problem unless the previous owner plugged in a random fan and actually destroyed the circuit. As I said often on these platforms the fan headers look standard but they're wired differently. If you pay attention to the old original fan and simply rewire to match it you should be fine as long as you buy the same fan.

The L1 ENV's job is designed to stop you from going any further because you have a hardware fault that could lead to serious damage if you go around it. Yes there is a way around it that we can tell you, but we won't likely want to give that information until you understand what doing that would mean and that you'd only do that to make sure you don't have a much more serious problem and you can at least boot to PROM, before investing the time and money in repairing that fan error, by getting another duplicate fan and wiring it in incorrectly.

Please consider what I've said and let us know what you decide. This is not bad news by any stretch. A fan is going to go out I need replacement at some point. That's not a huge deal. Now the fact that SGI is dead and gone means you can't get the fan from them you'll need to read the fan and order a replacement, possibly new old stock. When it comes you'll need to open the old fans label and check that they take the same signals and what the order of those signals are. You'll then take a picture and gently de-solder the old wires out of the old fan body and solder them into the new body at the correct signal positions. Then they should start up on stage one start up when the L1 decides the hardware is free of errors and attempts and auto power on, unless it's been told not to auto power on and you'll need to order a power on. If the NVRAM has been defaulted it should try to AUTO power on (pwr).

Also SGI's were designed to run with their cases/skins on. You could run them for a limited amount of time without it but as you see most of them use ductwork or other systematic airflow designs to properly cool chips with just heat sinks on them. Do not run an SGI for more than a few minutes without its case on! Troubleshooting obviously becomes a problem so just make sure that once you get the error your documenting it and taking pictures and then immediately turning the system off to give it a cool down. Most SGI's are not prompt overheating, I'm just saying they're designed to properly cool with their case fully intact. Not what they're lit off being poked at by a technician.

Sounds like you have a promising system on your hands, best of luck and keep us updated on the forums.
weblacky
I play an SGI Doctor, on daytime TV.

Trade Count: (10)
Posts: 1,716
Threads: 88
Joined: Jan 2019
Location: Seattle, WA
Find Reply
05-06-2025, 04:14 AM
#7
RE: How to Incorporate an Origin 300 In Homelab?
(05-06-2025, 04:14 AM)weblacky Wrote:  There are three reasons you can have this message:

1. Your fan is actually dead, this is very easy to check, is your fan not working?
2. Your fan is actually running fine but the system can't tell it's running at all, this is a failure of one of the environmental monitoring chips if this is the case, the chip would need to be replaced, not impossible job I do it all the time, but not easy for someone who's not familiar with professional soldering.
3. Your fan is not running, the system knows it's not running, but when you plug in a new fan it still won't run, this would be a failure of the MOSFET at the area of the fan connector that literally gives the fan 12V power. Only reason this should occur is if someone plugged in a replacement fan not knowing that the pin out has changed and they destroyed the boss, requiring board repair.

Most of these systems have several stages of start up. First stage is L1 start up, you could also consider this idle standby power. When you plug in a modern SGI it immediately starts its L1, just buy pulling in wall AC power.  Normally using the 5V standby power on the motherboard to trigger this process.  The motherboard is not technically active during this time but the base level environmental monitoring firmware and inventory program is active. This allows you to use a terminal signal to communicate with the L1 to start up the system and also control how the system starts up, diagnostic switches, and such.

I really appreciate the detailed response!  I started up the system briefly to see the physical location of the problem fan, and sure enough, I saw the left fan (not PSU) rattling a bit when it spun, then it stopped spinning after about 20 seconds.  I ruled out #3, as I swapped plugs with the center fan and the left fan, and the center fan ran with no issues while plugged into the left fan's spot.

I know you don't recommend it, but I am curious to see if PROM will load before I begin replacing the fan.  What is the command to ignore the environmental monitoring system?  It wouldn't be running for long, as I definitely want to avoid damaging this system.  I'd like to know if PROM comes up with any issues, that way I can purchase replacement parts in one pass.  After running the ignore command, I assume this command is not retained by memory, and the next startup will NOT ignore the monitoring system?

I found a replacement fan (3110KL-04W-B79), already bought it because it was cheap, but the only issue now is swapping the new fan's plugs with the broken fan so it'll fit on the motherboard.  I did find some old forums, including some on archived Nekochan forums, regarding fan pinout and whatnot, so I'll have to closely read those once my fan arrives, that way I don't fry my board.

How will I know if my system is in good condition?  Once PROM successfully loads and whatnot?

Thank you again for your response, I really appreciate it.  I feel reassured that my system is in good shape for now.  One last question (unrelated) though, do you know if this SGI fiber optic card (P/N 9210169) will work in my Origin 300?  I really want to experiment with fiber optics on this server.  The manual for that fiber optic card mentioned the Origin2000/200 and Onyx2, but it is PCI-X (I believe it is PCI-X at least, looks like it).
jander31
O2

Trade Count: (0)
Posts: 12
Threads: 2
Joined: May 2025
Location: United States
Find Reply
05-06-2025, 05:43 AM
#8
RE: How to Incorporate an Origin 300 In Homelab?
Sounds like you understand the risk involved. Here's the information you asked for:

This is the general L1 "public" guide for the entire platform: https://bukosek.si/hardware/collection/s...-guide.pdf
There is slightly more known by the community but there is no published source for it. This is pretty much all you're going to find openly.

From INSIDE the L1 terminal, command to ignore ALL env should be: env off
https://web.archive.org/web/202208140340...206/1.html

From INSIDE the L1 the commands to ignore JUST THE TEMPS is: env temp off
https://pixelbart.net/SGI/IRIX/i/65/19/d.../ch03.html

From inside Irix you'd prepend l1cmd to do that...but you aren't in Irix. I am unsure if shutting off the temp monitoring also ignores fans or not. It's been too long and I've never actually done it myself. So you may have to shut off the entire ENV, but you might try the temp subset first in case that will still allow you to maintain voltage monitoring.

In terms of what's actually good, the prom performs a fairly quick diagnostic of most things, including your graphics memory, before you should see the graphical interface. There is a much more rigorous PROM diagnostic called factory diagnostic that you can enable with virtual switches but that's usually not really required. I would personally say that if you get to the PROM graphical interface and go to the terminal and type: hinv

And get a good hardware listing and things look like what they're supposed to be that I would assume you're good until you hit a problem. The only problem I would expect you to ever hit is either that you have a flaky memory module or that your power supply is so old that it's nearing failure and you'll wind up with very odd VRM dips and other power related alerts and warnings. Outside of those two things I wouldn't expect anything odd. I would however say that when you load Irix it naturally resets a lot of your boot params, so that tends to clear out any weird stuff that people do. But overall I would trust the PROM diagnostics 98% of the time.

So I have some personal experience with fiber channel arbitrated loop drive raise, and I even have some controllers for SGI but I've never actually got that far on them. I can tell you that later SGI's all use the same PROM and have a note in the final revision about updating to allow the fastest 4Gb fiber channel controller to support the PROM. So I do assume that SGI branded or clone-style PCI-X cards could actually boot the system from external FC-AL drives. I can't make that guarantee.

I wanted to try this for a long time because it would allow most my machines to remain without any hard drive in them at all and that way they would not only be much quieter but I could simply use an emulated fiber channel link on something like SCST for Linux, to create boot volumes for them to natively boot from prom. I know that late model SGI's like yours should be able to do this, I do not believe that systems like the O2 and the Indy and indigo to and such can do this. I'm pretty certain the octane can do it with the XIO Fibre card version.

When it comes to compatible clones I really don't know. You'll have to try to find an old information of which compatible clone equals what SGI card. I'm not even sure that every SGI fiber card was actually boot compatible. All of the compatible cards will obviously read secondary volumes once Irix starts. But if you're looking for a fiber HBA to actually boot the system from an external SAN I would say I'm not entirely sure but I know one or two of the cards should be able to do it but you may have to get an SGI specific card and not the chipset compatible clone from like Qlogic or whatever.

It's actually been so long you may be the one to tell us what seems to work and what doesn't. Like I said there is very limited support for bootable HBA's in all of SGI's. Even the officially supported SAS and SATA controllers are not PROM bootable! But I'm 100% sure there is one or more cards for your platform that is bootable, assuming you have the last PROM revision loaded. Because that is part of the 1.48 prom changes list. Patch 7148 I think has the last SGI PROM...hopefully someone else can verify.

basic starter info:
https://archive.irixnet.org/apocrypha/ne...376/1.html
https://web.archive.org/web/201603251532...er_Updates
https://forums.irixnet.org/thread-4032-p...l#pid29516

user Jwhat has a nice archive of this stuff, I should have a copy somewhere, of all the PROM versions and all the flashers that you would need as each ProM bin file only works in a certain flasher version, it's a problem. So if you're L1 and PROM are very old, fix it first and then once you get Irix running you'll need to upgrade whatever's left. You have to chain load the bin files to the right flasher versions as you go. But you can do so within the same Irix version as long if it's a later version like 6.5.30. I've done this once. On your platform the serial number will sometimes disappear during flashing. Make sure to write down your serial number before flashing, it's usually easy to add it back in.

This should all have what you need for now. Good luck and let us know.

NOTE: Oh, so when you turn off the ENV you will be turning it off! You turn it back on by doing the opposite command and use the word "on" instead of "off". So assume all commands are "sticky" unless told otherwise.
(This post was last modified: 05-06-2025, 06:20 AM by weblacky.)
weblacky
I play an SGI Doctor, on daytime TV.

Trade Count: (10)
Posts: 1,716
Threads: 88
Joined: Jan 2019
Location: Seattle, WA
Find Reply
05-06-2025, 06:17 AM
#9
RE: How to Incorporate an Origin 300 In Homelab?
(05-06-2025, 06:17 AM)weblacky Wrote:  This should all have what you need for now. Good luck and let us know.

NOTE: Oh, so when you turn off the ENV you will be turning it off!  You turn it back on by doing the opposite command and use the word "on" instead of "off".  So assume all commands are "sticky" unless told otherwise.

Will be attempting this first thing after my morning classes, I will definitely keep you posted.  Gotta get some sleep for now, thank you again!  I really appreciate the information, I hope to have this server up and officially running once the replacement fan arrives.
jander31
O2

Trade Count: (0)
Posts: 12
Threads: 2
Joined: May 2025
Location: United States
Find Reply
05-06-2025, 06:52 AM
#10
RE: How to Incorporate an Origin 300 In Homelab?
I don't know terribly much after this when it comes to putting together the cluster systems. I think what I'm saying is mostly correct, but I've not had first-hand experience, it's all from reading.

Would you need to understand is the server you got is meant to be part of a cluster. Yes it is a complete computer that can boot the operating system to some form of terminal. As some people have mentioned there are compute nodes and some computer notes may actually contain graphics cards. You don't technically need graphics cards, you can use a terminal to do a command line only style interaction.

But basically you're supposed to have a special NumaLink router to attach these heavy cables too them in a star pattern and then designated a startup sequence to get them all online. They each boot in an order and in the end they look like a single system with the ability to perform process migration across nodes. You could easily find yourself with a system that has 30 or 60 CPU's and 10 gigs of RAM or more.

One thing that makes SGI clusters very very different from Beowolf clusters or other such things is that SGI was one of the few companies that actually creates a single system image system. I don't know if they did this before or after they acquired Cray supercomputer. But it was for the same purpose, that's my understanding.

So when you put an SGI cluster together you don't have 30 systems with two processors, you have a single system with 60 processors. It's the same with the memory although it obviously distinguishes between local memory and remote memory. I'm unsure if the linking technology is a variant of Infiniband or not, it seems to work very similarly. Entire processes can be migrated from node to node, I don't know exactly how threading spreads out but I'm assuming it's something similar. However this does mean that the system is not robust enough to withstand a loss of a node during operation.

Unlike modern cloud computing which is a bunch of conventional hardware with enormous amount of redundancy everywhere and an API layer that allows for sudden disappearances of nodes and redundant data being loaded and continuing, SGI systems work more on restores and checkpoints with ultra-reliable hardware. So they're very inflexible to a sudden machine going down. But from my understanding I don't believe it crashes the entire cluster but I do believe it would affect all processes on that cluster node and errors out the applications involved with that failed node...

You could of course still use technologies like MPI if that's how you wish to write it, but with a single system image a normal threaded application or a multi process application should scale linearly across the system due to the single system image design. This is not easy to achieve and you won't find many other platforms that can do this. But what it allowed was an ease of programming instead of relying on message passing or shared memory within infiniband or some sort of networking technology you were abstracted away from that as if you were actually working with a giant SMP architecture.

So that's one of the good things, unfortunately I don't know much about the MIPS processors to know how to get them to perform more or stream anything efficiently that would make them a good candidate to actually run as a cluster. Within the past two years several nearly complete clusters have come up both on online auction sites and on this site with almost 0 takers. That's not surprising because they take an enormous amount of power and what kind of performance are you getting compared to today's standard servers? Putting together two or three of these notes is a great little project. But don't go nuts thinking that if you look 30+ of them together that suddenly you're going to rival what you could get for today's graphics cards and multi CPU x86 style servers.

One thing you might do really depends on if you're involved in the software of things you could definitely build a three or four node cluster that's racked with a large amount of memory and the fastest processors you could find as a way of being a compilation point for someone in the community to very quickly run nightly builds or something for the new freeware software packages being developed by some of the members. I'm not one of those members but I'm simply stating that.

Outside of compiler access unless you're going to utilize any of the old tools for some special kind of video work or something you may just find at the end of this you hit a wall of what to really do with it. There is a glut of these cluster systems because once they didn't have a chance of outperforming a high-end desktop station for a fraction of the price outside of the novelty of the operating system that really wasn't much about them people were interested in. The rackmount systems aren't particularly visually appealing, most of them did have graphical options and so they could be used like a workstation. They are the only ones that can be linked together in more than just a pair. There's a series of rackmount Tezro stations that are allowed to be put into a pair with a NumaLink cable directly between them. None of the workstations, desktops, can be NumaLinked!!!!

But as I said you could get a pair of rackmount Tezros and link them in a special pairing only without a router. But I've never seen it done recently. Rackmount Tezro are actually quite rare. Some of these cluster nodes can be modded, I don't have any information on that end for you.

What I'd suggest is if you're interested in learning and potentially providing compilation space if you get this one working and feel good about all of it look into getting maybe 2 to 3 additional notes and calling it a day. You should also get your hands on an L2 controller, if you can. Look on the various online auction sites normally people post them for a couple hundred dollars but sometimes someone post for literally $75, that's how I got the one I have. So I don't use mine very much but as a troubleshooting tool because they also link to desktop versions of the same architecture, fuel and Tezro... But they have limited usefulness on those platforms.

The L2 runs an embedded Linux system that basically helps give you a terminal to each of your systems, interfaces with the L1 for the different cluster nodes, as well as provide power up sequence and error storage. I think you can boot a cluster system without one but I think they're supposed to be quite handy so consider trying to get your hand on one after you've gotten your first node to work.
weblacky
I play an SGI Doctor, on daytime TV.

Trade Count: (10)
Posts: 1,716
Threads: 88
Joined: Jan 2019
Location: Seattle, WA
Find Reply
05-06-2025, 08:33 AM


Forum Jump:


Users browsing this thread: 1 Guest(s)