0

My server recently rebooted, and the error logs show that there was a multi-bit ECC error. Apparently this has been happening for a while now, so I will be replacing the memory stick. Is there a way to figure out which module is having an issue to get the serial number from it? Attached are images of the error (all the displayed errors have similar information for memory address etc.). It says module 0, so I thought it would be F03D1A46, but it also says that the bank is 0xb so I don't know if that means the bank label (CHANNEL B -> F12FC349).

Errors in event log

A fatal hardware error has occurred.

Component: Memory
Error Source: BOOT
Error Type: Multi-Bit ECC

The details view of this entry contains further information.

Error details

- EventData 

  ErrorSource 7 
  FRUId {00000000-0000-0000-0000-000000000000} 
  FRUText DIMM# Sourced.AY 
  ValidBits 0xc34d 
  ErrorStatus 0x40400 
  PhysicalAddress 0xa1b2c3d4 
  PhysicalAddressMask 0x0 
  Node 0x0 
  Card 0x0 
  Module 0x0 
  Bank 0xb 
  Device 0x0 
  Row 0x1436 
  Column 0x17a 
  BitPosition 0x0 
  RequesterId 0x0 
  ResponderId 0x0 
  TargetId 0x0 
  ErrorType 3 
  Length 1304 
  RawData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

wmic memorychip output:

Attributes  BankLabel  Capacity     Caption          ConfiguredClockSpeed  ConfiguredVoltage  CreationClassName     DataWidth  Description      DeviceLocator  FormFactor  HotSwappable  InstallDate  InterleaveDataDepth  InterleavePosition  Manufacturer  MaxVoltage  MemoryType  MinVoltage  Model  Name             OtherIdentifyingInfo  PartNumber        PositionInRow  PoweredOn  Removable  Replaceable  SerialNumber  SKU  SMBIOSMemoryType  Speed  Status  Tag                TotalWidth  TypeDetail  Version  

2           CHANNEL A  17179869184  Physical Memory  2400                  1500               Win32_PhysicalMemory  64         Physical Memory  DIMM 1         8                                                                               Kingston      1500        0           1500               Physical Memory                        9965669-019.A00G                                                    F03D1A46           26                2400           Physical Memory 0  72          16512                

2           CHANNEL B  17179869184  Physical Memory  2400                  1500               Win32_PhysicalMemory  64         Physical Memory  DIMM 2         8                                                                               Kingston      1500        0           1500               Physical Memory                        9965669-019.A00G                                                    F12FC349           26                2400           Physical Memory 1  72          16512                
5
  • Just to confirm. You only have 2 modules installed, and you are asking essentially, if "Bank 0xb" indicates the problem is in channel B instead of A? – Ramhound Jul 17 '18 at 19:18
  • @Ramhound Correct, I don't know if it's saying "Bank 0xb" because it's channel B or "Module 0x0" because it's in the first slot (Physical Memory 0). – user Jul 17 '18 at 19:27
  • Are you opposed to running something like memtestx86? – Ramhound Jul 17 '18 at 19:42
  • I can pull one of the sticks out and run it in the offline server to see if memtest picks anything up. – user Jul 17 '18 at 20:07
  • I was only suggesting the tool, so you can determine which module you need to replace, if you are pulling out the module then you have a 50/50 chance of being correct. The only way to avoid that 50/50 shot is to use the replacement module, and pull both modules, and test them (which I suppose is acceptable). – Ramhound Jul 17 '18 at 20:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.