Description

 SPARC M5-32 and M6-32 servers running firmware versions prior to 9.6.20.b may erroneously diagnose a CMU board as faulty for memory link failover.

Occurrence

 This issue can occur in the following platforms: 

SPARC Platform

• SPARC M5-32 and M6-32 Servers without firmware version 9.6.20.b (as delivered in patch 27043440) or later

Note: To determine the firmware version installed on the system, use the following ILOM command:

    -> show /HOST sysfw_version
    
    /HOST
    Properties:
    sysfw_version = Sun System Firmware 9.6.20.b 2017/07/19 16:24

Symptoms

If the described issue occurs, the server will improperly mark a CMU board faulty with FMA code SPSUN4V-8000-GX and Problem class: fault.memory.memlink-failover.

The fault diagnosis will be evident in the service processor event log.  The event log can be displayed via the below ILOM command:

    -> show /SP/logs/event/list

If a Fault message similar to the following is in the event log, then the described issue has occured:

    1471  Wed Nov 15 18:26:04 2017  Fault     Fault     critical
    Fault detected at time = Wed Nov 15 18:26:04 2017. The suspect
    component: /SYS/CMU8 has fault.memory.memlink-failover with
    probability=100. Refer to http://support.oracle.com/msg/SPSUN4V-8000-GX
    for details.

The ILOM command 'show faulty' will contain output similar to the following:

    -> show faulty
    Target             | Property              | Value
    -------------------+-----------------------+-----------------------------------
    /SP/faultmgmt/0    | fru                   | /SYS/CMU8
    /SP/faultmgmt/0/   | class                 | fault.memory.memlink-failover
    faults/0           |                       |
    /SP/faultmgmt/0/   | sunw-msg-id           | SPSUN4V-8000-GX
    faults/0           |                       |
    /SP/faultmgmt/0/   | component             | /SYS/CMU8/CMP1/MLINK00/LANE1
    faults/0                                   |
    /SP/faultmgmt/0/   | uuid                  | 55c3d04c-9fb0-495a-dcda-c3f2673484
    faults/0           |                       | c6
    /SP/faultmgmt/0/   | timestamp             | 2017-11-15/18:26:04
    faults/0           |                       |
    /SP/faultmgmt/0/   | fru_part_number       | 7070507
    faults/0           |                       |
    <truncated>

Workaround

 There is no workaround for this issue.

This issue is addressed in the following releases:

SPARC Platform

• SPARC M5-32 and M6-32 Servers with firmware version 9.6.20.b (as delivered in patch 27043440) or later

Note: Loading new system firmware requires the host to be stopped and restarted to deploy the fixed hypervisor.  

See <Document: 1542610.1> for additional information regarding system firmware upgrade procedures.

Patches

History

15-Dec-2017: Document released, status Resolved

Internal Section: Comments:

This is a software issue.  Hypervisor fails to correctly identify whether the memory link failover is due to a persistent or transient error.

Questions regarding this document should be addressed to
sunalertpublication_us_grp@oracle.com and copy the
submitter/responsible engineer listed below:

Internal Contributor/Submitter: david.lafko@oracle.com
Internal Eng Responsible Engineer: stephen.guidon@oracle.com
Oracle Knowledge Analyst: jeff.folla@oracle.com
Internal Eng Business Unit Group: Server OS
Internal Associated SRs: 3-16179716748, 3-16125438261
Internal Resolution Patches: 27043440

References

This solution has no attachment