Help with Machine Check Events detected on your server from Fix Common Problems

[SDguy]

I've gotten this error message a few times now and haven't really seen what this message really pertains to.  I've had a couple of machine lock-ups in the past and would really love some help figuring out what might be the cause of this error report. Thanks for any help in advance!

towerofpower-diagnostics-20170326-2215.zip

[Squid]

Looks like a memory issue

 

Quote

Mar 26 04:30:31 TowerOfPower root: Fix Common Problems: Error: Machine Check Events detected on your server
Mar 26 04:30:31 TowerOfPower root: Hardware event. This is not a software error.
Mar 26 04:30:31 TowerOfPower root: MCE 0
Mar 26 04:30:31 TowerOfPower root: CPU 6 BANK 8 
Mar 26 04:30:31 TowerOfPower root: MISC 5cfbfb0000006040 ADDR c2c340b40 
Mar 26 04:30:43 TowerOfPower root: TIME 1490441889 Sat Mar 25 04:38:09 2017
Mar 26 04:30:43 TowerOfPower root: MCG status:
Mar 26 04:30:43 TowerOfPower root: MCi status:
Mar 26 04:30:43 TowerOfPower root: Corrected error
Mar 26 04:30:43 TowerOfPower root: MCi_MISC register valid
Mar 26 04:30:43 TowerOfPower root: MCi_ADDR register valid
Mar 26 04:30:43 TowerOfPower root: MCA: MEMORY CONTROLLER RD_CHANNELunspecified_ERR
Mar 26 04:30:43 TowerOfPower root: Transaction: Memory read error
Mar 26 04:30:43 TowerOfPower root: Memory read ECC error
Mar 26 04:30:43 TowerOfPower root: Memory corrected error count (CORE_ERR_CNT): 1
Mar 26 04:30:43 TowerOfPower root: Memory transaction Tracker ID (RTId): 40
Mar 26 04:30:43 TowerOfPower root: Memory DIMM ID of error: 0
Mar 26 04:30:43 TowerOfPower root: Memory channel ID of error: 0
Mar 26 04:30:43 TowerOfPower root: Memory ECC syndrome: 5cfbfb00
Mar 26 04:30:43 TowerOfPower root: STATUS 8c0000400001009f MCGSTATUS 0
Mar 26 04:30:43 TowerOfPower root: MCGCAP 1c09 APICID 20 SOCKETID 1 
Mar 26 04:30:43 TowerOfPower root: CPUID Vendor Intel Family 6 Model 44

 

( @RobJ - FYI, FCP added output last month from mcelog if it's installed , and I'm impressed with the detail that it gives )

Edited March 30, 2017 by Squid

[RobJ]

25 minutes ago, Squid said:

( @RobJ - FYI, FCP added output last month from mcelog if it's installed , and I'm impressed with the detail that it gives )

 

I saw you had added it, was glad to see it, and yes, I've been very pleased with the reporting it gives you, most of the time.  I don't remember it, but I believe I've seen a report or 2 of other hardware subsystem MCE issues that were too cryptic for me, and couldn't find good advice online.  But mostly it's great.

 

SDguy, I believe it's reporting that your ECC RAM detected and corrected a memory error, so this looks harmless.  But the fact it's having them, could possibly be related to your lockups in the past, a memory error that *couldn't* be corrected.  You may want to try the PassMark Memtest on your RAM (it has ECC RAM support).

[SDguy]

Thank you both! I had run memory check twice when first setting up the server without errors, but I suspected that this might be the case... I'll pull the suspected memory module and see if that fixes my problem. Thanks again!