Description

Updating to the affected ZFS Storage Appliance (ZFSSA) Software releases (as listed below) may cause Clustron PCIe cards that are used to connect the two storage nodes to be erroneously reported as faulty, as in the following:

    chassis-000 xxxx01 faulted Oracle Sun ZFS Storage 7320 1142FMM1A3 -- system

This can happen on any ZFS Storage Appliance in clustered configuration (running the affected software releases listed below) if the two storage nodes of the cluster are rebooted in close succession.

Note: Once it is confirmed that all the Clustron links are still active (as described in the Symptoms section below), these erroneous reports can be ignored.

Occurrence

This issue can occur in the following release:

ZFSSA Platform

• 2013.1.6.0 (Version string ak-2013.06.05.6.0) through 2013.1.6.15 (Version string ak-2013.06.05.6.15)

Note: Solaris and ZFSSA 2011 versions are not impacted by this issue.

Symptoms

If the described issue occurs, the following FMA report will be issued:

    chassis-000 xxxx01 faulted Oracle Sun ZFS Storage 7320 1142FMM1A3

To confirm that the FMA reports are erroneous as described in this document, output from the following two CLI commands needs to be reviewed to determine if the reported Clustron cards are in fact 'active' and maintaining the links between the two storage nodes as follows:

    :maintenance problems> ls

      Problems:
      COMPONENT DIAGNOSED TYPE DESCRIPTION

      problem-000 2016-12-2 19:10:01 Critical Fault The cable between the Ethernet ports of each controller is down.
      problem-001 2016-12-2 19:10:05 Critical Fault The cable between serial port 1 of this controller and serial port 0 of the other controller is down.
      problem-001 2016-12-2 19:10:05 Critical Fault The cable between serial port 0 of this controller and serial port 1 of the other controller is down.

    :> configuration cluster links

      clustron:0/clustron_uart:0 = AKCIOS_ACTIVE
      clustron:0/clustron_uart:1 = AKCIOS_ACTIVE
      clustron:0/dlpi:0 = AKCIOS_ACTIVE

Workaround

The erroneous FMA error reports described in this document can be ignored only if the "configuration cluster links" CLI command (as shown above) reports all links as'AKCIOS_ACTIVE'. These erroneous faults related to the Clustron card may be cleared by selecting the problem and marking it repaired, as shown below:

    :> maintenance problems select problem-000 markrepaired

Resolution

This issue is addressed in the following releases:

ZFSSA Platform:

• 2013.1.7.0 (Version string ak-2013.06.05.7.0) or later

History

23-Dec-2016: Document released, status Workaround
15-Feb-2017:Updated Workaround section.
31-Mar-2017: Updated Resolution section. State Resolved

This regression was caused by the putback for ER: 21224255

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

Internal Contributor/Submitter: rahul.nagraj@oracle.com
Internal Eng Responsible Engineer: rahul.nagraj@oracle.com
Oracle Knowledge Analyst: jeff.folla@oracle.com
Internal Eng Business Unit Group: Systems RPE
Internal Associated SRs: 3-13218322011, 3-13254078671, 3-13353513931, 3-13371301814, 3-13427936801, 3-13472509061, 3-13499032891,
3-13525645701, 3-13528545571, 3-13530069042, 3-13535404617, 3-13537929966, 3-13542604078, 3-13543597138, 3-13544539999, 3-13546551768,
3-13550946890, 3-13556675087, 3-13561745754, 3-13563847011, 3-13570015764, 3-13579925107, 3-13582410511, 3-13582991107, 3-13586312714,
3-13588953069, 3-13592324648, 3-13592337481, 3-13593991036, 3-13594297510, 3-13595244453, 3-13595804556, 3-13597634820, 3-13598545701,
3-13599090946, 3-13601678536, 3-13605455706, 3-13607706177, 3-13611722641, 3-13614571566, 3-13618732906, 3-13620580815, 3-13625296536,
3-13627082941, 3-13628792401, 3-13647054321, 3-13656966991, 3-13672030398, 3-13672182175, 3-13716319414, 3-13723202792, 3-13727660051,
3-13774079759, 3-13775211967,
Internal Pending Patches: TBD

References

This solution has no attachment