mcelog: Hardware event. This is not a software error

[Florant]

Hello everyone 

Anyone had a situation like this: i see a lot of flooding  on messages as below :

mcelog: Hardware event. This is not a software error.
Mar 13 21:43:33   mcelog: MCE 0
Mar 13 21:43:33  mcelog: CPU 23 BANK 5
Mar 13 21:43:33  mcelog: MISC 21402e2e86 ADDR ae2581c80
Mar 13 21:43:33   mcelog: TIME 1552509813 Wed Mar 13 21:43:33 2019
Mar 13 21:43:33   mcelog: MCG status:
Mar 13 21:43:33  mcelog: MCi status:
Mar 13 21:43:33  mcelog: Error overflow
Mar 13 21:43:33  mcelog: Corrected error
Mar 13 21:43:33  mcelog: MCi_MISC register valid
Mar 13 21:43:33  mcelog: MCi_ADDR register valid
Mar 13 21:43:33  mcelog: MCA: MEMORY CONTROLLER RD_CHANNEL2_ERR
Mar 13 21:43:33  mcelog: Transaction: Memory read error
Mar 13 21:43:33  mcelog: STATUS cc00008000010092 MCGSTATUS 0
Mar 13 21:43:33  mcelog: MCGCAP 1000814 APICID f SOCKETID 0
Mar 13 21:43:33  mcelog: PPIN 8800004f00800092
Mar 13 21:43:33  mcelog: CPUID Vendor Intel Family 6 Model 45

There are a lot of messages like this with different  CPU number BANK number. The server is DL380P G8 and its running Redhat linux. 
I didnt found anything on hp docs about this.

Thanks,
Florant

[Bunsol]

Hi Floran,

Good Day. Check if BIOS is updated to the latest. If it is then perform a hardware diagnostic on the server through the Intelligent Provisioning. The errors you are seeing are possibly because of hardware errors or the firmware.

Redhat article:-

https://access.redhat.com/solutions/67599

Take the diagnostics report and the AHS log and log a case with HPE using the below portal:-

https://support.hpe.com/hpesc/home

Regards,

Bunsol