Tyan S8030GM2NE

SINN78

Active Member
Apr 3, 2016
152
30
28
41
Usually there's 9 stand-off's on chassis supporting ATX motherboards -- the Tyan mobo has eight and one of the standoffs is directly under the memory slots. If the standoff isn't removed, it will short out the memory slot(s) (see pic - usually there's a standoff where the red circle is)

View attachment 14755

The other thing is how tightly the motherboard is screwed into the standoffs - if some are differing heights or if the screws are over-tightened, the motherboard can bow and flex - that is REALLY bad as it can cause intermittent shorts and difficult to diagnose faults

Finally, the case and back panel can sometimes have things that touch solder points, components and traces, you need to be careful when installing the I/O shield for example

The good thing is that the motherboard is working, so you just have to go over all this with a fine toothed comb.

What case are you using?
I have young kids so didn't want to leave the board and power supply on my desk out of fear they might try and play with it
so I put it into a corsair air 540 just so I can test everything and what not then it was going into my Lian li pc-p80 (it's in it now)
you ain't kidding about being careful with that I/O shield damn thing gave me a nasty cut on my finger OUCH
 

Cinnabyte

New Member
Jun 26, 2020
2
0
1
You should use the manual and look at the last 2-3 codes that you can see before the 0022. That tells you where in the boot process it is hanging. From the manual:

0x1D – 0x2A: OEM pre-memory initialization codes

So it would seem it is hanging on memory init. Maybe check the CPU socket?
The socket pins seem to be fine, and no debris can be found anywhere. I made sure to mount the CPU correctly multiple times, and I am using the provided torque tool.
 

hmw

Active Member
Apr 29, 2019
236
77
28
The socket pins seem to be fine, and no debris can be found anywhere. I made sure to mount the CPU correctly multiple times, and I am using the provided torque tool.
And the motherboard has been tested outside of any case/chassis to rule out any shorts etc?
 

Kev

Active Member
Feb 16, 2015
439
84
28
37
Who’s willing to test a new IPMI firmware that tyan sent me. It’s supposed to address the mounting iso issue using java. Pm me to get the firmware.
 
  • Like
Reactions: hmw

hmw

Active Member
Apr 29, 2019
236
77
28
I managed to put my motherboard in a Norco RPC-4308 case. It is an extremely tight fit and I wouldn't recommend it - I had to give up using the AX1200i power supply (so no SMBUS PSU monitoring) :(

The case has a lot of empty space at the bottom front, I got a PE foam block and cut out spaces for three 80 mm fans. There's two 80mm fans for the exhaust + the PSU fan is set to 'always on'.

Because the case has a backplane for 8x 3.5" removable HDDs - this interferes with the CPU cooler. I had to mount the fan shroud with some cables ties - in reverse - so the fan is pulling air and doesn't get in the way of the backplane. At least the cable routing is pretty good - no cables blocking the airflow :)

Here's the view from top:

server_top.jpg


The whole thing is amazingly very very quiet. Right now CPU temps are 34C at idle, the temp just below the two exhaust fans is ~ 33C. And drive temps are ~39C, not as low as I'd have liked, but still pretty okay.


And from the side: (you can see the foam block with the fans inset)

server_side.jpg
 

hmw

Active Member
Apr 29, 2019
236
77
28
Who’s willing to test a new IPMI firmware that tyan sent me. It’s supposed to address the mounting iso issue using java. Pm me to get the firmware.
What's the ver of the firmware? I might have a go at it ...
 

Kev

Active Member
Feb 16, 2015
439
84
28
37
I tested out this new firmware because I just realized I didn't need to take down my server if I was only updating BMC firmware. HTML5 KVM now has mount remote media function but didn't test. Java KVM also can mount ISO without crashing but didn't test performance. BIOS settings from BMC page now gone. That's all i've noticed so far. Things are looking good!
 
  • Like
Reactions: hmw

Jordan Perdue

New Member
Jul 15, 2018
28
4
3
This is really excellent news and fixes the main problems I had with the firmware - very glad I stayed with this. Thanks for the updates guys.
 

Jordan Perdue

New Member
Jul 15, 2018
28
4
3
I have not gotten to test anything as its running services in the DC, but I did open the HTML5 KVM and could plug in media.
 

cephyr

New Member
Jul 25, 2020
4
0
1
Hi guys,

is it possible to passthrough those onboard sata ports ( SFF-8643 connectors) in VMWare?
It would be nice if I wouldn't need additional SAS cards for that purpose.

thanks in advance!
 

Kev

Active Member
Feb 16, 2015
439
84
28
37
I asked earlier as there are three sata devices to pass through and I didn’t want to risk tanking my esxi install since it is booting from one of the sata ports. Should be possible if you boot from usb and try.
 

cephyr

New Member
Jul 25, 2020
4
0
1
I asked earlier as there are three sata devices to pass through and I didn’t want to risk tanking my esxi install since it is booting from one of the sata ports. Should be possible if you boot from usb and try.
thanks for your quick reply, I didn't see a post of all possible passthrough devices
could someone post it?

would be great if someone could confirm that it is possible to pass the sata connectors into a VM - would like to use it for freenas
 

cephyr

New Member
Jul 25, 2020
4
0
1
Please report back if you succeed or not.
I can't test it because I don't own the board.
It would've been interesting to know if this works before I buy it.
Otherwise I'm leaning more towards the Asrockrack ROMED8-2T or the Supermicro H12 ATX, because they have more PCI Slots for SAS HBA's.
 

Kev

Active Member
Feb 16, 2015
439
84
28
37
I think the price difference sealed the decision for me. It’s nice to have extra slots but what would I put in the future that I don’t have in there now? I have one slot free after optane AIC and nvme, hba, 40g network.
 

Jordan Perdue

New Member
Jul 15, 2018
28
4
3
So I'm about to roll my 4u 7302p into production in four days and figured I'd double the ram (128gb to 256gb) so... add four sticks and we are good, yeah?


Wrong.

DIMM bank e0 seems completely dead for me. Trying all of the sticks multiple times in a revolving pattern shows all eight are good, and slot E0 is consistently Disabled upon bootup. Is this a known glitch, by rare chance? I wasnt using e0 before and it was rock solid. From what I could tell when upgrading its rock solid with 7/8 DIMMS recognized, but E0 just gets disabled on bootup. Has anyone seen this or know how to proceed?

I'm going to get some canned air and rebuild the motherboard and ram on a flat table, then after blasting the SC846 interior and checking the mounts I will retry screwing it in.

Photos:



Any thoughts and advice are welcomed.