When to use: Fault vs. Error Log Indicators (Lightpath Mode)
This appendix gives
highly recommended Service Indicator activation models
for typical system issues, when the Lightpath mode is implemented. The
purpose of this appendix is to get consistency across platforms, and to
answer common questions about how to handle specific issues. The reason that
these are recommended rather than required, is due to the range of systems
that are involved, specifically related to the different types of physical
layouts (for example: deskside, blade and blade chassis, rack-mounted and
particularly high end racks).
This appendix does
not change the architectural requirements specified in
other parts of this document, nor the requirement for implementations to
support those requirements. If there are any inconsistencies between this
appendix and the requirements in the rest of this document, the requirements
take precedence over this appendix. It is very important, therefore, that
designers understand the requirements in this architecture, and more
specifically, those in
.
gives the recommended models. The
general model, though, is still dictated by the following requirement, copied
here from the :
Service Indicator Activation Models for Typical System
Issues (Lightpath Mode)
Component or problem area
Problem or issue
Indicator activation?
(see notes
,
)
Entry made in Service Focal Point error
log?
IBM Call Home?
Comments
FRU Fault indicator?(see notes
,
)
Error Log indicator?(see note
)
All
Any not already covered in this table
Consult with the xipSIA architecture team for the proper
behavior
Power supply or fan
Redundant optional one missing
no
no
no
no
This row for information only (this is not a Serviceable
Event).
Redundant non-optional one missing
no
yes
yes
no
Failed redundant
yes
no
yes
yes
Failed non-redundant
yes
no
yes
yes
Power regulator not in power supply
Failed
yes
no
yes
yes
Treated the same as any other FRU or component on a FRU which
fails.
Power
Insufficient power with all power supplies installed and
operational for power domain
no
yes
yes
no
For components that will not power up due to lack of power
(for example, a blade in a chassis), and when that component
implements a fast blink capability for the green power LED, that
component’s green power LED remains in the fast blink
mode.
Insufficient power with missing power supply
no
yes
yes
no
Disabled button pressed
Unable to power on or off component due to power button being
disabled
no
no
no
no
These rows for information only (these are not a Serviceable
Events).
KVM or media tray button pressed, but not active because
disabled
no
no
no
no
Temperature detected out of tolerance
Warning only (no performance throttling or shut-down)
no
no
brand dependent
no
Performance throttling or shut-down due to over-temp
condition
yes
no
yes
no
FRU Fault indicator must be at component that is throttled or
shut-down (temperature indicators, which are not architected, do
not roll-up to Enclosure Fault indicators).
Memory (DIMMs)
No memory installed at all, or memory that is installed is
mismatched
yes
no
yes
no
Invalid memory configuration above the base memory (missing
memory, mismatched memory, unsupported memory)
no
yes
yes
no
Failed or predicted to fail
yes
no
yes
yes
System VPD
Invalid or missing
yes
no
yes
yes
Lack of some of the required fields in the call-home may
result in the call-home being flagged as a lack of
entitlement.
Battery
Missing or failed
yes
no
yes
brand dependent
Rechargeable battery needing reconditioning
no
yes
yes
no
CPU
Failed or predicted to fail
yes
no
yes
yes
Planar or CEC
Failed
yes
no
yes
yes
Disk
Failed or predicted to fail
yes
no
yes
yes
Boot device
Missing boot device
no
yes
yes
no
Corrupt image
no
yes
yes
no
I/O adapter
Fail or non-recoverable EEH error
yes
no
yes
yes
Service Processor (FSP, BMC, IMM)
Hardware failure
no
yes
yes
yes
Firmware failure
no
yes
yes
brand dependent
BIOS or Flash
Corrupted single side
no
yes
yes
no
Both sides corrupted
yes
no
yes
yes
All
Unisolated event
no
yes
yes
no
See also Requirement
Switch
Failed
yes
no
yes
yes
System chassis controller (AMM, CME, ITME)
Failed
yes
no
yes
yes
Midplane
Failed
yes
no
yes
yes
Cables
Failed or missing
N/A
yes
yes
yes
Notes:
Never activate both a
Fault indicator and an Error Log indicator for the same problem. See also
Requirement
, referenced immediately above
.
Fault indicators
above the FRU Fault indicator are not specified here, but the requirements
specify that a FRU Fault is rolled-up to the next higher level indicator
(specifically, the Enclosure Fault indicator).
Enclosure Fault
indicators and above are only roll-up indicators and are never activated
without a FRU Fault indicator being activated. Therefore the column in
indicates a FRU Fault indicator.
That is, if no FRU Fault indicator exists for the problem, then the Error Log
indicator is used instead (per Requirement
, referenced immediately above
).
The activation of the
Error Log indicator (previously known as the System Information (Attention)
indicator) and Fault indicators are regulated by the following requirements,
among others: