C220 M3 no CIMC, BMC Failed, IPMI errors, and HUU fails

[MPorter17]

Long time lurker, first time poster!

So I picked up a 2nd had C220 M3 for use in my home lab.

I am running into issues that I hope people can help with, so thanks in advance!

Things I notice::

So here is the deal.  The server takes anywhere from 20-30 minutes to go through the 'Configuring and Testing Memory' and Configuring Platform Settings' stages of initial boot.

Once it gets through that, it starts booting ESXi (what was installed when I bought it), and halfway through ESXi boot, it hangs forever on some sort of IPMI driver.

I wanted to set up CIMC for remote access of course, and when I press F8 on boot (after waiting the 20-30 minutes past the 2 'Configuring' stages) it also flags an error relating to IPMI - something about IPMI protocol not found or along those lines.

Also, when looking at the BIOS (again, after waiting the 20-30 minutes above) I see that the BIOS reports 'BMC Self Test Status - FAILED', (screenshot attached), and there is a place in the BIOS (cant remember where at the moment) where you would typically find Device MAC and such, it is all blank (well blank in the sense that it displays all the same characters, strongly hinting at some sort of failure)

So with all this, and with research, I downloaded the HUU utility to try and clean up/repair/upgrade components to try and fix.  Again, after waiting the 20-30 minutes above, I hit F6, boot to the HUU bootable USB I made, and it starts its process.  It does its thing, but then fails, with an error of::

'ERROR: Invalid Platform'

ISO is for Target Machine - C220

Actual Machine is - CXXX

Actual PID - ToBeFilledbyO.E.M.

(Screenshot of above attached as well)

Now, if this were a customer, I would TAC it and see what happens, but this is mine, with no support.  Im suspecting its likely a MotherBoard failure, but want to exhaust all options before looking for a 2nd hand MotherBoard.

I have attached 2 screenshots of the errors, and if additional screens or pictures are needed or are helpful, please let me know.

Once again, thanks for looking, and everyones time is greatly appreciated!!

Porter

[mojafri]

Most likely motherboard issue. Since BMC test is failing, in C220M3  model we don't have option to boot alternate cimc image using jumper (Preset in M5 servers). Wondering, which HUU image you are using? Because while running HUU, you are getting message for "Invalid platform".

Also, after you press F8, this what stage you reach? Are you getting option to restore is to factory defaults? 

Regards,

MJ

[MPorter17]

mojafri,

Thanks for your time!

The HUU images I tried were:

3.0(4j)

2.0(9o)

1.5(7f)

Basically tried all the latest, even after checking the compatibility matrix.

When pressing F8 on boot to access CIMC settings, It tries, then I get an error along the lines of 'IPMI Protocol not found'.  I will have to try the boot again and take a screenshot at a later time, as I am at work right now.

Another thing I forgot to mention....

So I thought maybe I could try using the BIOS and CIMC recovery.cap to maybe fix these issues.  I follow the instructions (changing jumper), and plug in a USB (formatted FAT32) with the recovery.cap file extracted from HUU (tried from the 3.0 and 2.0 HUU's listed above).

That failed as well - Its like the BIOS doesnt see the .cap file to let it upgrade.  

Used this guide:

https://www.cisco.com/c/en/us/td/docs/unified_computing/ucs/c/hw/C220/install/C220/install.html#pgfId-1417014

I never see the 'Found a valid recovery file' in the BIOS - I assume it skips right over that, because I dont see it try and do that anywhere in the boot process with a .cap and USB plugged in.

[Wes Austin]

The HUU is failing because the server SEEPROM is burned in with "ToBeFilledbyOEM" and not C220-M3S etc.

If you are able to login CIMC, you can see what the PID is. If it is not UCS-C220-M3...the HUU will fail. You may need to consult with who you bought the gear from and understand why the PID is not a Cisco one.

[MPorter17]

Thanks for your time Wes!

I am not able to get into CIMC in any way - dont know its IP, its creds, nothing.

Thats what has led me to where I am now - by trying to reset CIMC to default, but running into all these issues.

I can see the MoBo has Cisco part numbers on it, so I would suspect where I got the server probably wasnt aware of this issue.

Is there anyway to correct the SEEPROM?

[MPorter17]

Well I ended up getting a 2nd hand motherboard, and as one would expect, everything seems to be working fine.

Thanks for looking!

[MPorter17]

And to add another update that might help someone in the future....

I installed the replacement mobo, and things began working fine (had it in my office with 'Monitor1')  All was working great.

Moved it into my basement, and connected 'Monitor2' (different monitor than 'Monitor1'), and issues began happening once again.  BMC Failed in BIOS, and took 20-30 minutes to boot.

Baffled, I took the server back to my office with Monitor1, and everything worked perfect.

Moved it back to my basement, using Monitor1, and everything has been spot on perfect - used HUU's 1.5.4, 2.0.9, and finally 3.0.4 to successfully update my server all the way to current.

This has been such a bizarre issue, that turned out to be a problem with whatever monitor I was using at first. I dont know what was up with that monitor, or the cable, but this C220-M3S DID NOT LIKE IT!!!

So needless to say, if you are having these goofy issues, try a different monitor LOL

Take Care everyone, and have a great holiday season!!

Porter