vSphere 7 Update 3 PSOD with Epic Milan 7443p and Asrock ROMED8-2T

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

mirol

Member
Jan 6, 2018
39
4
8
37
Hi Guys yesterday I tried to update my esxi host from VMware ESXi, 7.0.2, 18538813 to ESXi-7.0U3-18644231-standard via ssh metod like i always do.After restart I got PSOD visible on Attached photo from this link Untitled .I tried update it twice and everytime I got the same issue after restart its come back to previous version.Anyone have idea why its happening its some bug with update 3 or some bios settings cause it any clues?
My server spec is:
Amd Epyc 7743P
Asrock ROMED8-2T with Bmc v 1.19.00 and Bios v L3.28
8x 128GB M393A4K40DB3-CWE
And Corsair ax1600i psu
----Pcie Cards:
Chelsio t540-cr
Samsung PM1735 3.2gb
Amd pro w 3100
Nvidia Quadro A4000
 

jpmomo

Active Member
Aug 12, 2018
531
192
43
Hi Guys yesterday I tried to update my esxi host from VMware ESXi, 7.0.2, 18538813 to ESXi-7.0U3-18644231-standard via ssh metod like i always do.After restart I got PSOD visible on Attached photo from this link Untitled .I tried update it twice and everytime I got the same issue after restart its come back to previous version.Anyone have idea why its happening its some bug with update 3 or some bios settings cause it any clues?
My server spec is:
Amd Epyc 7743P
Asrock ROMED8-2T with Bmc v 1.19.00 and Bios v L3.28
8x 128GB M393A4K40DB3-CWE
And Corsair ax1600i psu
----Pcie Cards:
Chelsio t540-cr
Samsung PM1735 3.2gb
Amd pro w 3100
Nvidia Quadro A4000
same here with a supermicro h12ssw-in. i am going to try on another server that has an amd 7452 and another one that has an ice lake. will update this thread with my results.
 

jpmomo

Active Member
Aug 12, 2018
531
192
43
just to clarify, I had the psod with an amd milan 7313p on a supermicro h12ssw-in when trying to upgrade to 7.0.U3. I did not have this issue with either an amd 7452 or an intel 8352s (ice lake xeon).
 

mirol

Member
Jan 6, 2018
39
4
8
37
I got virtualization emabled besides it work with old version.I think this update got issues with milan cpus then :( we prob need to w8 for next update and try.
 

HilkoVMware

New Member
Oct 12, 2021
2
4
3
Netherlands
There is a bug in AGESA MilanPI 1.0.0.1/1.0.0.2 where the LOCK in VM_CR MSR isn't equal between the first core and the other cores.
Ask your motherboard vendor for AGESA 1.0.0.3 or 1.0.0.4 or (not recommended as you're bypassing a problem) use cpuUniformityHardCheckPanic = FALSE ESXi kernel boot option.

1.0.0.1:

1.0.0.4:
 

mirol

Member
Jan 6, 2018
39
4
8
37
Ehh I just dont see asrock will do it if I ask tbh (will try) I will prob need to w8 and pray for updated bios someday then...
Big thx for make us aware about this tho:D
 

HilkoVMware

New Member
Oct 12, 2021
2
4
3
Netherlands
Asrock-rack and Supermicro are amazing at providing BIOS-es. I usually get them within an hour to 24 hours. Even without using my at vmware email address.
 

jpmomo

Active Member
Aug 12, 2018
531
192
43
Asrock-rack and Supermicro are amazing at providing BIOS-es. I usually get them within an hour to 24 hours. Even without using my at vmware email address.
I contacted both but do not have a date for when they will be available. I had much better responses in the past with both companies but not so much lately. I will update this thread as soon as I get any bios that fixes this issue.
 

AugustaLemke

New Member
Sep 6, 2021
19
3
3
I contacted both but do not have a date for when they will be available. I had much better responses in the past with both companies but not so much lately. I will update this thread as soon as I get any bios that fixes this issue.
I am getting this with a 7313P on a Supermicro H12SSL-CT.... Who do we need to call / get ahold of at Supermicro to get this fixed? I am running BIOS version 2.1.

Did you use their official support portal? I was going to contact Supermicro but if you did already I don't see a point I guess. That is awful if they don't have an ETA. As a consultant, this is exactly why I don't recommend Supermicro to my clients. They are fine for homelab but if I was running a business enterprise this would be completely unacceptable.

Also, I'm not understanding the issue here. If this is a defect in AGESA, why is it now a problem in 7.0U3? What happened / changed in 7.0U3 where this defect now matters?
 
Last edited:

jpmomo

Active Member
Aug 12, 2018
531
192
43
I am getting this with a 7313P on a Supermicro H12SSL-CT.... Who do we need to call / get ahold of at Supermicro to get this fixed? I am running BIOS version 2.1.

Did you use their official support portal? I was going to contact Supermicro but if you did already I don't see a point I guess. That is awful if they don't have an ETA. As a consultant, this is exactly why I don't recommend Supermicro to my clients. They are fine for homelab but if I was running a business enterprise this would be completely unacceptable.

Also, I'm not understanding the issue here. If this is a defect in AGESA, why is it now a problem in 7.0U3? What happened / changed in 7.0U3 where this defect now matters?
I just opened up a normal support ticket. See response below. I told Hoang that I could wait until the official release was available. I will just stay at esx 7.0U2. They should have a new bios by December:

Hoang Phan (TS) 10/21/2021 11:42:30 #8

Hi Joe,

PM has a non-officical release BIOS version 2.3 for H12SSW-iN, however it will NOT allow you to roll back to version 2.0 after upgrade BIOS to version 2.3.

Do you still want this non-officical release BIOS version 2.3?

It is still under testing by the lab don't have the ETA when BIOS 2.3 will be released.

Regards,
HP
 

AugustaLemke

New Member
Sep 6, 2021
19
3
3
I just opened up a normal support ticket. See response below. I told Hoang that I could wait until the official release was available. I will just stay at esx 7.0U2. They should have a new bios by December:

Hoang Phan (TS) 10/21/2021 11:42:30 #8

Hi Joe,

PM has a non-officical release BIOS version 2.3 for H12SSW-iN, however it will NOT allow you to roll back to version 2.0 after upgrade BIOS to version 2.3.

Do you still want this non-officical release BIOS version 2.3?

It is still under testing by the lab don't have the ETA when BIOS 2.3 will be released.

Regards,
HP

Thanks for posting that info. It's a bummer we can't go back if there's an issue for the BIOS so I guess I'll wait to update. In the meantime I'll just do the boot override (cpuUniformityHardCheckPanic=FALSE) to get U3 working.
 

NablaSquaredG

Layer 1 Magician
Aug 17, 2020
1,319
800
113
As a consultant, this is exactly why I don't recommend Supermicro to my clients. They are fine for homelab but if I was running a business enterprise this would be completely unacceptable.
From my experience, no company does give ETAs on this type of issue (required BIOS Update after a bug report)...
 

AugustaLemke

New Member
Sep 6, 2021
19
3
3
From my experience, no company does give ETAs on this type of issue (required BIOS Update after a bug report)...
They are just so slow though... 1.0.0.3 has been out for months and a few of their H12 boards (non-ATX ones) have had the newer AGESA MilanPI since early September.
 

mirol

Member
Jan 6, 2018
39
4
8
37
Hi Guys just to let you know asrock support contact me today and they send me new bios ROMD82T328Z with agesa 1.0.0.6 will try it later its really nice and fast support from ashrock rack team im really positive suprised :D
 

mirol

Member
Jan 6, 2018
39
4
8
37
Hi Guys just to let you know asrock support contact me today and they send me new bios ROMD82T328Z with agesa 1.0.0.6 will try it later its really nice and fast support from ashrock rack team im really positive suprised :D
Just test it and esxi works perfectly.
 

wild_one

New Member
Nov 17, 2021
1
0
1
Looks like Supermicro updated the H12SSL-CT BIOS to 2.3 with MilanPI1.0.0.6 recently, so it appears to be out of beta now.