Supermicro x10QBi
ver 3.2a
4x Xeon cpu E7 4850 v2 @ 2.30Ghz
(these are 10core for a total of 40cores)
memory 262144 MB (ddr3 1067MHz)
LSI3008-IT HBA adapter

 

This machine was working with no issues, friend was upgrading and I was able to aquire this one with no hard drives.
I have everything in my rack and unraid booting fresh with my key. I only have one sata drive installed to start. After logging into unraid for the first time on this machine the drive was not reconized. I tried to put into different bays, was set to auto start and nothing.
I checked connectors, cables, and the drive I was using I put in another machine and it was seen with no problems.

Here is another concern, I went to look at my core count and in total with hyper thread I have 95? and I should only have 80. I talked to the previous owner and they said it always showed 80 for them.
I have tried different versions or the software, I even tried installing on another usb without my key and it was still showing 95 total and would not see the hard drive.
I am at a loss and wanted to post for any ideas.
I have to this server off right now as my head feels like it wants to explode hehe. I did read in bugs that there is a spot to save a log file and that would help. I just have to wait until tomorrow and I will try to post that too.

 

thanks

it is very likely that the HBA that comes with the machine is not supported by UNRAID.

You need to supply diagnostics to reveal more details.

 

I am sorry, when I said working with no issues I should of said "no issues with their unraid" and for this I would think the HBA controller would be fine.

 

I am going to try and attach the diag zip file, remember this is a pretty fresh unraid, no drives and under settings/drives i set auto to yes as I believe it was no at install.

 

the zips will have 2 log files as I tried first with oldest version i could get (6.12.9) and then upgraded to (6.12.13) and did reboot and did the diag again. Both log files seem to be the same. Here are a couple of the lines that I dont understand.

 

Aug 31 06:53:29 TowerG kernel: smpboot: Allowing 96 CPUs, 0 hotplug CPUs         (there should only be 80 max)

 

Aug 31 06:53:29 TowerG kernel: MDS CPU bug present and SMT on, data leak possible     (cpu bug?)

there is a url under this : See  https://www.kernel.org/doc/html/latest/admin-guide/hw-vuln/mds.html for more details.  for some reason I couldnt get it to load.

 

I look forward to any insight someone may have

 

thank you

 

towerg-diagnostics-20240831-0959.zip

well much to my surprise I found out why I was getting so many cores.

4x Xeon cpu E7 4850 v2 @ 2.30Ghz      v2=12 cores wich is 96 or 0-95 i guess v1 =10 core so much to my surprise we figured out why

still no hard drives found, and working to figure that out next, any suggestions?

Here is another update: I did a bios reset and i started to get a memory error, after taking a few hours to walk away from everything I was able to pull a memory board and re-seat the memory and this resolved the memory error. 

I had tried to 3d print a 3.5 to 2.5 caddy for my ssd that failed. I then just took the ssd and slid it in the tray until a connection was made. To my surprise, unraid noticed the drive, I even did a reboot and again to my surprise my HBA controller had seen the drive. I was able to mount the ssd, start the array and format. The drive I had been trying to get to to work I put back in my main pc and it worked with no issues. So this seems to show that the HBA is working, the backplane is working (at least the bay I have it plugged into) and I will look to order me a couple sas drives and cross my fingers that there will be no issues.

I hope some of this was informative and helpful to someone else. It really was a headache for me

Here is a follow up.

to me and a few other people our surprise was there was something that was burnt on the backplane. What made this more confusing was that it must  of been something that controlled drive spin ups as ssd drive would still work.

After talking to other techs that build these machines it was said that moving the server can sometimes shift the backplane enough where it can short on the metal support cage that is there. They mentioned that this has happened to them a few times.

I had to dig up the money to purchase another backplane and after installing everything is working with no issues.

hope nobody else has to have this experience as it caused me sleepless nights and I am getting to old to not get my sleep hehe