3.18.120 32113 - Uncorrectable ECC memory error

Alarm Group:
PLAT
Description:
This alarm indicates the chipset has detected an uncorrectable (multiple-bit) memory error the ECC (Error-Correcting Code) circuitry in the memory is unable to correct.
Severity:
Critical
Instance:
May include AlarmLocation, AlarmId, AlarmState, AlarmSeverity, and bindVarNamesValueStr
HA Score:
Normal
Auto Clear Seconds:
0 (zero)
OID:
eagleXgDsrTpdEccUncorrectableErrorNotify
Alarm ID:
TKSPLATCR14
Cause:
This alarm indicates chipset has detected an uncorrectable (multiple-bit) memory error the ECC (Error-Correcting Code) circuitry in the memory is unable to correct.
Diagnostic Information:
Syscheck can be manually executed using the following methods:
  • Login as syscheck. When logging in, syscheck runs and the login connection is dropped. This account does not have shell access.
  • From the root account the Command Line Interface can be used directly.
    • Execute syscheck -h for usage information.
  • In DSR 6.0 and later, from the admusr account the Command Line Interface can be used directly when called using sudo.
    • Execute syscheck -h for usage information.
  • Through the platcfg user interface.

Note:

In versions later than TPD 6.5, root access using SSH is disabled. The admusr should be used instead. If the command needs to be run as admusr, sudo must be prepended to the command and the full path to the command must be used.

Recovery:

  1. It is recommended to contact My Oracle Support to request hardware replacement.