Dell R840 "Critical: memory Incompatible Error-You are using non-production memory"

Hi,

clearing the password via jumper and then using iDRAC Express is not likely to give you a new path around this error. The better path is to treat it as a hardware memory-channel fault and isolate it by removing all DIMMs, then testing the smallest supported configuration, because Dell’s own troubleshooting flow points to DIMM-versus-board isolation as the deciding step.

@Dell-Martin S​ Thanks for the reply.   I recieved this server a few day ago with no memory installed and the power button was stuck in the "on" position.    I unstuck the power button and power up the server and recieved the error "incompatible dimm" even though no memory is installed in the unit.   I installed minimal memory as well as fully populated the server and still got the same error at boot.

I assume the previous person that had the server installed incompatible memory (DDR3, UDIMMs etc.).    They booted up the server and got the error then pulled the power plugs which somehow retained this error.

I tried clearing the NVRAM via jumper in hopes of removing the error with no luck but may not had done it properly and will try again.    

I have a 2nd R840 that is working fine with the same memory.

Hi,

How many processors are there install in the problematic server R840? Error on location C1 seems to be related to CPU3. How about you bring the server to the mininum to post components to check if the error persist, if it does, you might be having a bad mainboard. 

Remove all components except the below, please note that the server needs 2 processors as minimum to post. Then turn on the server to check the issue. 

• Two same processor (Skylake/Cascade lake) in socket processor 1 and processor 2
• One memory module (RDIMM/LRDIMM) in socket A1
• One power supply unit
• PIB and cable (4 power + 2 side band connect to MB)
• System board
• Chassis with Control Panel and cable (to power system on)

@DELL-Joey C​ 

Thanks for the reply.

The unit has 4 x 6132 CPUs installed.   It appears to have iDrac Enterprise not Express.    I noticed a message on the front LCD saying "memory detected but not configurable" when I had the 4 CPUs and 4 x 16GB RDIMMs installed.

I read your post and removed 2 CPUs and am now able to post and boot.   I think the error stills exists if I try to install CPU 3-4 enabling memory bank C.    What should I do next to attempt to remove the assumed existing error.    Should I do a BIOS upgrade or do you suggest another route.

Thanks

Hi,

To isolate if the processor 3 and 4 having issue, try swapping them. Install 3 and 4 processors into socket 1 and 2. And install the other memory module which have been removed out from the system. Based on the steps, you would have 3 memory out from the server as it only need 1 in A1. Hence, try other modules. If the server is able to boot, try updating the BIOS and iDRAC. Make sure they are updated before you power off the server to install back the 2 other CPU and 3 other memory modules. CPU1, Slot A1 / CPU2, Slot B1 / CPU3, Slot C1 / CPU4, Slot D1.

@DELL-Joey C​ 

Hello,

I swapped CPUs 1 and 2 and all 4 CPUs are fine.    I updated the BIOS, Updated iDrac, cleared system logs and cleared NVRAM.

Installing all 4 CPUs brings back the error.    I tried with 2 RDIMMs in A1, D4 then tried 4 RDIMMS in CPU1, Slot A1 / CPU2, Slot B1 / CPU3, Slot C1 / CPU4, Slot D1.

@DELL-Joey C​ 

Thank your assistance, it was looking promising for awhile.

Regards.