HP StorageWorks 7100 - Virtual Array Appendix page 16

FC-SCSI Interface SpecificationDescribes the IceCube SCSI interface. Targetted as part of OEM manual.Disk Array Con
Event Event
Number Name
(dec/hex)
345/0x159 Bus A TWSI communication
failure
346/0x15a
Data Loss Detector entry
347/0x15b Member Disk Drive Added
Back Into Disk Set
348/0x15c Frontend Fibre Channel ABTS
Event
349/0x15d Cache Version Mismatch In
RAM Image
350/0x15e RAM Version Mismatch
351/0x15f
Disk Format Version Mismatch
352/0x160 ShutDown Due To Power
Supply Failure
353/0x161 Recovery from battery backed
RAM Loss Started
354/0x162 Recovery from Battery Backed
RAM Loss complete
355/0x163
Redundancy corrected
Predictive Suspected
Maintenance Components
Implication
Y Controller,
battery, power
supply,
midplane
Y Indicates Disk drive(s)
multiple failure
causing data loss
Y Ignore - Operator None
or host activity
Y Ignore None
Y See errors None
associated with
shut down
Y N.A. None
Y N.A. None
Y Single Power Supplies,
Occurrance Controller,
Backplane
Y Single Controller
Occurrance
Y Ignore None
Y Suspect data if Disk Drive,
followed by error Back End FC
code 93 (0x5d) Link, Controller
Error! No text of specified style in document.
Description
Y Either local controller or remote controller SRAM
on two-wire serial bus A could not be accessed for
controller sync. Bus B SRAM will be used instead
(if Bus B SRAMs are also inaccessible, both
controllers will reset and and try again; finally one
controller will be disabled and the system will
continue in single controller mode)
N This error code indicates that the data loss detector
module was called because of multiple failures. A
data loss has occurred.
N This error code is a system change event indicating
that a member of a volume set that was either down,
failed, or missing has been returned to the
unrestricted use (aka ready) state. This can occur
via the add physical drive command, or via hot plug
insertion of the drive.
N This error code indicates that the Host sent a Fibre
Channel ABTS (Abort Sequence) BLS frame to the
abort an IO. The array will log this event for
informational and debug purposes only. It does not
necessarily indicate a problem with the array.
N A firmware download has a new version of the
cacheVersionNumber. This is OK if there are no
writes stuck in cache but it causes upload to fail
with this error code if there are writes stuck in
cache. To fix this it is necessary to revert to the old
version of firmware and solve whatever problem
was causing writes to be stuck in cache (probably
one or more disks have failed).
N The upload routine was unable to upload part of the
disk NVRAM image because the current firmware
does not support that use of RAM. (Most likely an
older version of firmware is trying to upload a disk
image posted by some other firmware version.)
N The NVRAM format of the disks does not match the
format used by the controller. This event is logged
when an icicle controller attempts to use disks
previously shutdown by an Ice controller or visa-
versa. The controller cannot use an NVRAM image
in the incorrect format.
N This error code indicates that the Background
Manager (BGM) has discovered that there are not
enough good power supplies to run the system. The
BGM is shutting down the subsystem to minimize
system operation with inadequate power.The
NVRAM is posted to disk.
N This error code indicates that recovery from a
battery backed RAM loss has been initiated.
N This error code indicates that recovery from a
battery backed RAM loss completed with at least
partial success. Maps were recovered. Multiple
failures may or may not have occurred. Occurrance
of multiple failures are reported in the log between
the RECOV_STARTED and RECOV_DONE
events.
N This error code indicates that the parity scan which
executes during recovery from RAM loss found an
instance of incorrect redundant data. Data which
was being updated when RAM was lost may
produce this error in single failure scenarios. This
error can also result from multiple failure
conditions. Firmware differentiates the two
scenarios by placing a limit on the number of these
conditions which can occur before recovery is
terminated due to multiple component failures. If
this error is followed by error 93 (0x5d) then the
contents of data blocks reported in this error should
be considered suspect. The block address of the