Processor CATERR message

[Kitem]

Hello,

I received a “Processor - CATERR” message on my Supermicro X11DPH-T MB

I googled and understood it's an issue with the CPU which I have to remove from the MB to boot

However, when I Power On, I can log in via the IPMI but get only a black screen if I'm trying to boot (I don't even get the boot messages.

Before this issue, I experienced failure on the RAM slot (2x under the CPU1).

Any help?

 

[BoredSysadmin]

Clearly, this is CAT ERRor. Time to replace the cat.
Jokes aside, CAT is for Catastrophic Errors with the cpu. Either overheating or permanent CPU damage. Since you had a RAM failure on same board, it could also indicate a board issue, like power related.

 

[Kitem]

My wife is a cat-phobic, so it's always an issue with cats

Damm….how can I check if its the board?
I may get a replacement

I never received any messages related to high temperature or any message at all

Just this critical one, suddenly.

Does the MB boots if I'm using only one CPU?
Must be in the CPU1 slot or can be the other one?

 

[RolloZ170]

CATERR must not be the CPU, but is mostly.

 

[Kitem]

Do you know how I can check what the problem is and how to fix it?
Can you point me to useful documents?

Thanks

 

[RolloZ170]

Can you point me to useful documents?

CATERR point not to a specific issue.
try to check all components on other motherboard

 

[Kitem]

I unplugged everything and kept only 1 CPU and 1 RAM module.....and I was able to boot (getting the screen asking me to plug a bootable device.

I then plugged in the pci card with the 1tb nmve, and I was able to boot

I then plugged in other 3 RAM modules......I was able to boot.

I tried plugging in everything (except the second CPU), and I wasn't able to boot.

I went to the last working configuration (below) but this time it didn't boot

Unplugged everything again.....no luck

I'm going crazy....can't find any log to check what is the issue.

Please help

 

[Kitem]

No one can help?

 

[Kizune]

Really hard to debug such intermittent issues, especially remotely. Do you have access to the IPMI interface and if so were there any errors after the power on? By “does not boot” do you mean “POST does not finish”? With IPMI you can see the Q-codes of the POST process, which was the last code when it stopped?
Are you sure that the CPU cooler is installed with enough tension on the screws? With these big sockets any uneven or wrong tension can cause all sorts of instabilities. On 3167 I would really suggest using screwdriver with torque control and follow the recommended pattern for applying tension. Usually pattern is printed on the cooler itself (at least on the rack mount coolers). If you don’t have such cooler here’s the reference for you.

 

[Kitem]

Thanks for your reply.

1) I do have access to IPMI, and the only errors shown are the ones referring to the "General Chassis Intrusion" when I was removing/installing the components.



2) When I say it doesn't boot, I mean that when I press the "Power On" button from the IPMI, I'm getting this error message




3) I will remove and reinstall the CPU being sure I'm applying enough tension on the screws

 

[Kitem]

I tried to put even more tension while screwing in the CPU. I also update again the BIOS (although same version as before) and reset the IPMI

I noticed in the Miscellaneous menu a “POST Snooping” section.
It shows an “ff” code that, based on the below documentation, points to a “Recovery Error (PEI)” [Pag 7], but it's “Reserved for future AMI error codes” [Pag 10].

I keep trying, but still no luck…

 

[RolloZ170]

“POST Snooping” section.
It shows an “ff” code

FF is the reset value. no port80 action = FF (boards with Q-Code LED have 00 as reset value)

 

[Kitem]

So… that means it's not doing anything?
Or refers to any malfunction?

 

[RolloZ170]

So… that means it's not doing anything?

correct.

 

[Kitem]

Sh*t
How can I log what is preventing the system from booting ?

 

[RolloZ170]

Sh*t
How can I log what is preventing the system from booting ?

at the moment no x86 code is executed.
some voltage failure. defective CPU, short circuits. who knows.

 

[Kitem]

I can also try buying a new CPU, but what if it's not the issue?
I would avoid spending money w/o being sure it will solve the issue....

Also, I have two CPUs (testing them one at a time). Which are the odds both broken almost at the same time?

 

[RolloZ170]

can be the motherboard, but faulty PSU too.
if it worked without change anything something is broken, happens daily.

 

[Kitem]

I will give it a try with a new PSU