Help Wanted: Supermicro x10SL7-F Now stuck in infinite boot loop

Vocalpoint

New Member
Mar 25, 2016
23
1
3
Calgary, AB
Today I completed a routine firmware update for one of my two Supermicro x10SL7-F motherboards - simply going from BIOS v3.3 to v3.4 using the IPMI interface - everything seemed to go fine until the machine restarted and now I am caught in an infinite reboot loop

The board displays the usual Supermicro logo and then looks like it's passing off to the UEFI Windows Boot menu - but never makes it.

I was able to get it to boot just once to the OS about an hour ago when I had the box in my office here - then I unplugged and moved it back to it's normal spot in our rack and reconnected everything and no dice - it simply goes round and round.

I have done everything I can think of from a CMOS reset to a battery change to reseating RAM to even forcing a reapplication of the BIOS using their SUM tool.

Everything was working great for months before today. The IPMI FW is current (Redfish 3.88) What else can I try?

Hoping you can make a suggestion or two as this machine looks like it’s in real trouble.

Cheers

VP
 

j_h_o

Active Member
Apr 21, 2015
543
133
43
California, US
Have you confirmed that the board is/is not able to boot another OS via USB? (Or mount an OS via IPMI and try to boot off of that.) Are you able to boot to a Windows USB key and get into Windows set up? That would at least indicate that the board is working correctly -- and just Windows is failing to boot.

You could try repairing the boot configuration, or reinstalling the bare metal OS.
 

Vocalpoint

New Member
Mar 25, 2016
23
1
3
Calgary, AB
J_H_O

Have not moved to this next step - but great suggestion and one that I will surely try.

I believe there is something else odd going on here - the board posts just fine - but will not (for whatever reason) acknowledge the UEFI Windows Boot Manager. It is truly bizarre.

VP
 

Vocalpoint

New Member
Mar 25, 2016
23
1
3
Calgary, AB
Update for everyone - after several painful hours of trial and error - the only solution that is verified was a rollback to BIOS 3.2. This point was really driven home when I returned my attention to the other x10SL7-F (Let's call it Board 1) I have here several hours after the issues with this board I described above. (Board 2)

Just before I posted the original thread (Board 2) I updated Board 1 to BIOS 3.4 and it too exhibited the same issues - no boot into Windows by itself - but in that case - after it did one boot loop - I jumped in with an F11 (Display Boot menu) and forced it to use the UEFI - Windows Boot Manager option - which resulted in a bizarre set of of sounds AFTER I selected the UEFI boot option - this board produced 4 distinct blocks of 6 short beeps each - and then I saw the Windows logo come up.

The odd noises aside (which I still cannot find any info on) - I thought Board 1 was all good - so I just left it running in Windows while I turned my attention to Board 2.

But several troubleshooting hours later - when I decided to see if Board 1 was still actually OK - I restarted it via it's IPMI Interface to see if it would boot to Windows properly and saw that it went right back into it own boot loop.

Right about this time - Supermicro USA did get back to me and suggested a rollback to BIOS 3.3 - but I remembered that Board 1 had 3.2 before this mess started so I reflashed it back to 3.2 - low and behold - the board updated, I went into the BIOS, adjusted a few things (like turning off the PXE boot and the RAID controller) and ensuring the boot sequence was correct - and Board 1 went right into Windows first time. Whew!

In the case of Board 2 - once BIOS 3.4 was on there - and I did it twice - once via the IPMI interface and again in a power down state using the SUM Tool - there was no chance it would ever boot into Windows. I made at least 10-12 attempts (while it was looping) to intervene with F11 and choose the UEFI - Windows Boot Manager option - with this board - I could select it but no weird beeps, no boot, no nothing - the screen would simply go black for like 2 minutes and the machine would eventually restart and display another annoying "System Initializing" pass.

After seeing success with 3.2 on Board 1 - I rolled Board 2 back to 3.2 as well (it had 3.3 on there before this mess started) - and we are all good - both machines are purring away as normal.

I will be sending all this info to Supermicro - but it's a cautionary tale folks - this is the first time in my very long 30 year history of computing that I have ever seen an issue like this with a BIOS file - it's a scary feeling to have a box (my file server no less) in a delicate space of "almost out of commission".

I have no idea WTF is going on with these two specific boards - they are identical in form and function purchased within 3 months of each other in 2017 - why this BIOS simply does not work when all others have been fine remains a mystery. I will have a earful (or email full) for Supermicro on this experience - I cannot imagine this occurring with a enterprise server in a business setting (or a series of them).

I had always thought that Supermicro was a cut above and their stuff always "just worked" - But I guess not. Lesson learned.

VP
 
  • Like
Reactions: Fritz

up3up4

Member
Jun 10, 2018
68
18
8
Same bug with bios 3.4 of my x10slm-f, I rolled back to 3.3 bios. From bios release notes " Fixed failure of chipsec items BLE, BWP, and TCO SMI lock. " I think is the broken bios code related with PCH chipset.
 

JSchuricht

Active Member
Apr 4, 2011
185
65
28
Wish I had seen this thread an hour ago. Same issue going from 3.3 to 3.4 and rolling back to 3.3 to fix the bootloop. BIOS was flashed through IPMI and Windows 2019 server in UEFI boot.
 

Vocalpoint

New Member
Mar 25, 2016
23
1
3
Calgary, AB
As noted I stayed with 3.2 on both servers. I find this BIOS gives dramatically faster booting and IPMI operations than 3.3 ever did. Considering they are end of life and 3.3/3.4 really bring nothing to the table - I would rather keep things fast than deal with low level micro changes that Supermicro deems important. The only thing that is important to me is a working server.
 

PrimeXFN

New Member
Mar 4, 2019
3
0
1
Wish I'd seen this before Friday, when I updated one of my X10SLM-F boards to 3.4 and can't boot anything - except Windows, exactly one time, and never again. Reverted back to 3.3 and all problems went away.

A little bit of a thread necro, but it's just amazing that SuperMicro has kept completely broken BIOS FW up for this long.
 

james23

Active Member
Nov 18, 2014
428
95
28
49
Unbelievable- same exact issue over here, with a new X10SLM-F , as im in the testing / build phase of this 1u server , and it came with bios 3.2, one of my first steps was to update bios to latest, 3.4 - i then spent the next 6 hours or so trying everything just to get into any type of win10 UEFI Installer USB's startup screen (trying everything at the HW and BIOS settings, levels).

Tried several win10 install usbs and Clone'd images copied to ssds , all would just get black screen after post, then about 45sec later board would reboot itself). Fround this thread and it all made sense ( i need a bios older than the current v3.4) so I Then spent ~ 1hr trying to find a legit v3.3 source (link below)

Rolling back to bios 3.3 has resolved the issue.

Very disappointed in supermicro for this. (and im a large SM user).

some amazing STH user has complied an auto-updating archive of all SM bios releases (easily searchable, links are DL'd direct from supermicro.com):
( Supermicro BIOS/IPMI File Listing )
(his STH thread) - what a great guy!

bios DL links for my case (x10slm-f):
v3.4 - current (dont use)
v3.3 - no issues - Bios Download ZIP / Release notes PDF
v3.2 - others say works fine as well (use drunkencat.net link above to download it, if needed)

thanks for making this thread!