H610S: erroneous NVRAM error messages

NetApp H610S

This article addresses the sporadic logging of NVRAM faults on H610S nodes, which often results in the eviction of a node's slice (metadata) drive

1. The node logs a variation of one of the following NVRAM (NVDIMM) faults:

NVRAM device critical={Device: 0x1000004 Key: armStatusSaveNArmed, Device: 0x1000004 Key: notEnoughEnergyForCatastrophicSave}
NVRAM device error={Device: 0x1000004 Key: energySourceTemperatureStatus Device: 0x1000004 CurTemp: Unknown HighTemp: Unknown LowTemp: Unknown Device: 0x1000104 CurTemp: Unknown HighTemp: Unknown LowTemp: Unknown} critical={Device: 0x1000004 Key: armStatusSaveNArmed, Device: 0x1000004 Key: notEnoughEnergyForCatastrophicSave, Device: 0x1000004 Key: warningThresholdExceeded}
NVRAM device problems detected, metadata drive will be removed. Contact NetApp Support. error={Device: 0x1000004 Key: energySourceLifetimeStatus} critical={Device: 0x1000004 Key: armStatusSaveNArmed, Device: 0x1000004 Key: energySourceHardwareFailure, Device: 0x1000004 Key: energySourceHealthAssessmentError, Device: 0x1000004 Key: noEnergySourcePresent, Device: 0x1000004 Key: notEnoughEnergyForCatastrophicSave, Device: 0x1000004 Key: warningThresholdExceeded}
NVRAM device warning={[Device: 0x1000105, Key: 0x700cc0]}
NVRAM device problems detected, metadata drive will be removed. Contact NetApp Support. warning={[Device: 0x1000005, Key: 0x300cc0]} error={[Device: 0x1000005, Key: 0x800000]}
platformHardwareEvent: Suppressing Nvdimm status result 1/3 as false positive
platformHardwareEvent: Nvdimm status has stabilized after 1 failed result(s)

2. The node's slice/metadata drive (drive 0) may be ejected, depending on the error that is logged and the version of Element OS the node is running.

If the node is running Element 10.x the slice drive will most likely not be ejected, depending on timing factors within Element.
If the node is running Element 11.0 through 11.1 the slice drive will be ejected, regardless of timing.
If the node is running Element 11.3.0 the slice drive may be ejected, depending on which of the above mentioned faults was experienced.

3. The NVRAM (NVDIMM) fault resolves on its own less than 5 minutes later.

4. If the node's slice/metadata drive (drive 0) was ejected it WILL NOT automatically re-add.

If these alerts are resolved, no further action from support is required.