|
Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com |
all firmware patch digest
From: IT Resource Center (support_feedback
us-support-mail.external.hp.com)
Date: Fri Jun 27 2003 - 11:36:32 CDT
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
HP Support Information Digests
===============================================================================
o IT Resource Center World Wide Web Service
---------------------------------------------------
If you subscribed through the IT Resource Center and would
like to be REMOVED from this mailing list, access the
IT Resource Center on the World Wide Web at:
http://support.itrc.hp.com/
Login using your IT Resource Center User ID and Password.
Then select Support Information Digests (located under
Maintenance and Support). You may then unsubscribe from the
appropriate digest.
To download a patch referenced below, access the
IT Resource Center on the World Wide Web at:
http://support.itrc.hp.com/
Login using your IT Resource Center User ID and Password.
Then select Individual Patches (under Maintenance and Support)
to access the patch. You may also download a patch via anonymous
ftp(1) from ftp.itrc.hp.com.
===============================================================================
Digest Name: weekly all firmware patch digest
Created: Sun Jun 15 3:05:02 PDT 2003
Table of Contents:
Document ID Title
--------------- -----------
PF_CSFW0005 HP Superdome Utility Firmware 7.26 and PDC 36.1
PF_CCANGSPB0220 A6696B GSP firmware B.02.20
PF_CHARGSPC0214 rp24xx revision B GSP firmware C.02.14
PF_CKEYMAT0500 HP rp8400/rp7410 Firmware Version 5.0
PF_D36ESHP05 Seagate 36ES HP05 drive firmware
The documents are listed below.
-------------------------------------------------------------------------------
Document ID: PF_CSFW0005
Date Loaded: 20030320
Title: HP Superdome Utility Firmware 7.26 and PDC 36.1
Patch Name: PF_CSFW0005
Patch Description: HP Superdome Utility Firmware 7.26 and PDC 36.1
Creation Date: 03/03/12
Post Date: 03/03/20
Products:
A5200A HP Superdome
Status: General Release
Critical:
Yes
PF_CSFW0005: OTHER
This release fixes the handling of cell redundant
power faults. Additionally, it contains the
following fixes: Resolved rare corner case issue
of partition failure during large database
tear downs (rare PTLB storm conditions); vPARs
boot time in IODC; Boot support for DS2405 in
IODC; powered off cells causing high alert level
errors throughout complex; put out chassis log
when battery status low; DIMM detection info
checked for corruption; fixed cause of
IODISC_SBA_FLUSH_CACHE_ERR chassis logs; fixed
detection & correction of SBEs on XBCs; and empty
PCI slots reported as non-existent by
manageability tools.
PF_CSFW0004: OTHER
Adjusted cache clock setting for d-cache
tag SAD
PF_CSFW0003: OTHER
This release is to correct possible loss or
corruption of Group C information.
PF_CSFW0002: OTHER
This release is to correct instruction cache
parity error handling. In PDC release 32.4,
multiple parity error events in the instruction
cache of a single processor result in an illegal
instruction trap. As a result, the affected
partition is reset (HD3001).
PF_CSFW0001: OTHER
This release is to minimize the impact of cross bar
hardware problems. It closes all known system
firmware issues for support of SCSI parameters.
Category Tags:
defect_repair hardware_enablement general_release critical
Path Name: /firmware_patches/hp/cpu/PF_CSFW0005
Symptoms:
PF_CSFW0005:
- CPU brick failures should not cause the cell power to
be removed. As a result, issues that could result from a
CPU Brick failure include: Cell boot beyond BIB if Boot
is Blocked; Remote (from the GSP, from scan tools, from
FWUU, or as a result of a PDC command) reads of the
following FRUIDs on the power-faulted cell may fail (Cell
Power Board FRUID, CPU and DIMM FRUIDs); Reading and
writing complex profiles from the MP or from PDC running
on another cell; The GSP's RD/WR commands when these
commands are used to read/write PDH space addresses.
- ServiceGuard or ASR TOC out of PDC procedure or LPMC.
- Non-responsive cells after
BOOT_TABULATE_CPU_SPEED_FAILED chassis log.
- Manageability tools report empty PCI slots as non-
existent.
- Powered-off cells cause high-alert level chassis
codes throughout complex.
- HP-UX PD management tools (parstatus) may hang.
- vPar(s) stall while another vPar boots or shuts down.
PF_CSFW0004:
- CPUs are deconfigured due to d-cache errors
PF_CSFW0003:
- CC_IO_CSR_HF_ERR chassis codes are emitted at regular
intervals while running HP-UX.
- Console numbering on non-consecutive partitions is
incorrect when viewing console logs.
- Loss or corruption of Group C information.
(See Util.7.24.release.notes.text and
PDC.35.3.release.notes.text for full release notes)
PF_CSFW0002:
- During on-line operation, affected partition is reset
(See CUFW0722.text and CPDC0325.text for full release
notes)
PF_CSFW0001:
- During boot, cells doing routing could halt.
- During on-line operation, processors could HPMC and a
partition could reset. All partitions in a complex
eventually go down because EMS monitor calls cross all
fabric links.
PF_CPDC0323:
- HD2952 ERR_HPMC_FALL_THROUGH_RESET halts cell in machine
check.
- The MC handler fails to detect an LPMC and issues an
ERR_HPMC_FALL_THROUGH_RESET chassis code. The affected
cell is halted and the other cells HPMC and restart due
to loss of memory coherency.
PF_CUFW0712:
- The changes in this release include PA8700 cell board
support, IOX cabinet support, iCOD support, 500MHz cell
support, PDC >= 32.2 support, PDC >= 10.0 support, many
minor feature enhancements and many bug fixes. In
addition, a new command, IO, has been added to the GSP
command handler to display the mapping between cell
slots and IO chassis.
- Changed the PS command so it properly locates cell power
and temperature errors for Caribe cells.
- Fixed a bug in the SO command that caused a GSP reboot
if the user answered no to all three of the GSP wide
parameter change questions.
PF_CUFW0078:
- The changes in this release include IOX cabinet support,
iCOD support, 500MHz cell support, PDC >= 10.0 support,
many minor feature enhancements and many bug fixes. In
addition, a new command, IO, has been added to the GSP
command handler to display the mapping between cell slots
and IO chassis.
PF_CPDC0074:
- HD2495 The LPMC handler Icache LPMC detection was being
turned off after 1000 LPMCs on an individual processor.
Only processors with a hard Icache parity error that
results in 1000 parity error events (LPMCs) will
encounter this problem. Up to 1000 parity errors are
handled correctly. After 1000 LPMC's, every fetch of
that Icache line that uses the instruction that contains
the parity error will be affected. PDC is supposed to
turn off signaling of LPMCs and allow the CPU to handle
the event. In this case the CPU does the right thing
and fetches the instruction again and bypasses the cache,
thus ensuring forward progress, though some degradation in
performance. However, since PDC erroneously turned off
LPMC detection, the CPU does not know there was a problem
and could in fact use a bad Icache line. This could
manifest itself in silent data corruption.
- Some data regarding LPMC events in the case of hard cache
errors:
- If it is a cache line that gets hit a lot, LPMC chassis
logs will emit at a terrific rate up to 1000, and then
no more logs. Likelihood is that the OS will panic and
crash, as seen here in the lab.
- If it is a cache line that is not hit a lot, the chassis
logs identifying the first 1000 events will dribble out.
The diagnostics monitoring tool will count these events
and try to de-allocate the offending CPU after 3 such
events.
- LPMC chassis logs have a high alert level that turns on
the front panel error LED, and these logs are captured
in the utilities error log. These mechanisms are
intended to make such problems more visible to the
customer.
- Regarding the diagnostics de-allocation after 3 LPMCs:
CPUs that have IO interrupts attached cannot be
de-allocated, and the monarch CPU cannot be
de-allocated. These processors do pass CPU selftests,
because interrupts are turned off during PDC boot.
- HD2505 Speedy Boot (Fast Boot) skip PDH test option was
skipping testing AND initialization of scratch ram. Now
it just skips testing.
Defect Description:
PF_CSFW0005:
- See Symptoms.
PF_CSFW0004:
- See Symptoms.
PF_CSFW0003:
- See Symptoms.
PF_CSFW0002:
- See Symptoms.
PF_CSFW0001:
- See Symptoms.
PF_CPDC0323:
- See Symptoms.
PF_CUFW0712:
- See Symptoms.
PF_CUFW0078:
- See Symptoms.
PF_CPDC0074:
- See Symptoms.
Enhancement:
No
SR:
0000000000
Patch Files:
Util.7.26.release.notes.text
PDC.36.1.release.notes.text
fwuu
gsp.7.26.frm
clu.7.8.frm
pm.7.16.frm
pdhc.7.10.frm
cio.7.4.frm
hd.bin.36.1.frm
cksum(1) Output:
3198839043 10567 Util.7.26.release.notes.text
3312777333 2626 PDC.36.1.release.notes.text
4056145039 98581 fwuu
2753509328 1969376 gsp.7.26.frm
3151019038 132160 clu.7.8.frm
3126340054 66624 pdhc.7.10.frm
2902747097 66624 pm.7.16.frm
753925881 1049664 cio.7.4.frm
353727197 4195392 hd.bin.36.1.frm
Patch Conflicts: None
Patch Dependencies: None
Hardware Dependencies: None
Other Dependencies:
PHSS_28764 should also be installed on Superdome systems
with the Virtual Partition (vPar) software installed.
Supersedes:
PF_CSFW0004 PF_CSFW0003 PF_CSFW0002 PF_CSFW0001 PF_CPDC0323 PF_CUFW0712
PF_CUFW0078 PF_CPDC0074
Equivalent Patches: None
Patch Package Size: 1771 KBytes
Installation Instructions:
NOTE: Before installing check for servic4 notes.
Applicable service notes will supercede these instructions.
If you have already installed PF_CSFW0004.
On the Superdome:
1. Shutdown UNIX on all partition(s) using "shutdown -RH".
From another UNIX system open two windows:
2. Extract files:
first unzip: "gunzip PF_CSFW0005.tar.gz"
then untar: "tar -xvf PF_CSFW0005.tar"
3. In the first window telnet into the Superdome's GSP
4. In the second window change to the directory containing the fwuu
executable, Utility firmware and PDC firmware provided in this patch.
5. Type: ./fwuu <Superdome's GSP name>
6. The install program will then display two listings. First it will list
all the firmware ROMs located in the current directory, then it will
list the flash handles associated with the system's hardware. The
second section will contain a line for every firmware entity
in the system and a corresponding flash handle which must be used
to update the entity.
7. the program will prompt: FWUU>
8. Update CIO firmware
Type: 'update FLASH_HANDLE(s) cio.7.4.frm'
where FLASH_HANDLE is the CIO flash handle found in step 6
9. When the update is complete the following message will be displayed:
100 %
Flash Update
Cabinet Entity Type Handle Firmware File Status
_______ ___________ ______ ______________________ _________
0 CIO 769 cio.7.4.frm PASSED
Resetting entity(s) requested for update.....
Continue (yes/no)? [Y]: y
Successful reset of all entity(s).
NOTE: All entities of one type may be updated simultaneously by specifying
the FLASH_HANDLES for all entities to be done in the update command.
(i.e. update FLASH_HANDLE_1 FLASH_HANDLE_2 FLASH_HANDLE_3 myclu.frm)
10. Update PDHC firmware
Type: 'update FLASH_HANDLE(s) pdhc.7.10.frm'
11. Update PDC firmware
Type: 'update FLASH_HANDLE(s) hd.bin.36.1.frm'
12. In the first window connected to the GSP power off 48v
on every cabinet in the system.
- GSP:CM> pe
13. Update CLU firmware
Type: 'update FLASH_HANDLE(s) clu.7.8.frm'
14. Update PM firmware
Type: 'update FLASH_HANDLE(s) pm.7.16.frm'
15. Update GSP firmware
Type: 'update FLASH_HANDLE gsp.7.26.frm'
16. In the first window Type: Ctrl ] (that is the Ctrl key on the keyboard)
to break the telnet session and then log back into the GSP.
NOTE: It should take about 2 minutes for the GSP to reboot.
17. Once the GSP has rebooted, AC power cycle cabinet(s).
- AC power off all cabinet(s) in the complex and wait 30 seconds
prior to AC power on to insure TOGO routing is handled by
the new firmware.
When the GSP comes up reset all partitions.
- GSP:CM> rs
then wait 8 minutes for all partitions to boot to BCH.
18. verify that the firmware update is complete by using the sysrev command.
- GSP:CM> sysrev
you should get the following message:
Utility Subsystem FW Revision Level: 7.26
There will also be a list of all firmware revisions in the cabinet
which should be verified for correctness.
-----End of Document ID: PF_CSFW0005-----------------------------------------
Document ID: PF_CCANGSPB0220
Date Loaded: 20030519
Title: A6696B GSP firmware B.02.20
Patch Name: PF_CCANGSPB0220
Patch Description: A6696B GSP firmware B.02.20
Creation Date: 02/11/15
Post Date: 03/05/19
Products:
A6144A HP9000 L3000 Enterprise Server
A6144B HP9000 L3000 Enterprise Server
A6696B L-Class revision B GSP
A3639C HP9000 N4000 Enterprise Server
A6797B HP9000 L1500 Enterprise Server
Status: General Release
Critical: No
Category Tags:
defect_repair hardware_enablement enhancement
general_release
Path Name: /firmware_patches/hp/cpu/PF_CCANGSPB0220
Symptoms:
PF_CCANGSPB0220:
The Following enhancement and fixes are included in GSP
firmware revision B.02.20
************************************************************
The Java plug-in 1.4x is not supported with Web Access. Java
plug-in 1.3x is supportted.
************************************************************
1- Run LED is displayed OFF after "System Boot detected"
In previous revisions of gsp firmware the front panel RUN
led did not always flash and would instead appear to be off.
The following is the correct display from the Vitual Font
Panel.
********** VIRTUAL FRONT PANEL **********
System Boot detected
*****************************************
LEDs: RUN ATTENTION FAULT REMOTE POWER
The following is a sample incorrect display from the Virtual
Front Panel.
********** VIRTUAL FRONT PANEL **********
System Boot detected
*****************************************
LEDs: RUN ATTENTION FAULT REMOTE POWER
OFF FLASH OFF ON ON
B.02.20 has been modified to correct this issue.
2 - LAN hangs when gateway address matches GSP's own IP
address.
In previous revisionss of GSP firmware When GSP's gateway
address and ip address matched and the LAN hung the firmware
pinged the gateway address to dectect if a real hang
occured. If the gateway and ip addresses were the same, the
ping always return OK status incorrectly indicating that
the LAN is active. As a result the LAN would not
reinitialize, leaving the LAN indefinitely in a hang state.
B.02.20 firmware checks the gateway IP is checked against
the GSP IP before attempting to ping gateway. If the gateway
and IP match then the broadcast address is pinged instead.
3 - GSP denies CTRL-Ecf access
During extremely high traffic on the I2C bus CTRL-Ecf did
not work properly and Write access to the GSP would not
be granted.The serial port connection continued to work and
the OS was not affected. This issue was corrected with GSP
firmware revision B.02.20.
4 - Data aborts caused by CRC invalid Voice packets.
In previous revisions of GSP firmware the checksum
verication result on i2c Voice messages is ignored when
returning to upper level calling routines. If the length
byte in the Voice message is corrupted this will result
later in a data abort, when the message is copied to a
secondary location for processing. The following is a
sample error chassis code:
ALERT LEVEL: 8 = Boot Possible, performance impaired
SOURCE: 6 = platform
SOURCE DETAIL: 6 = service processor SOURCE ID: 0
PROBLEM DETAIL: 1 = selftest result
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 03 = console
REPORTING ENTITY TYPE: 1 = service proc REPORTING ID: 00
0x7010028166002030 0B0010FF 00000000 type 14
0x4810028166002030 63436870 6C615266 type 9 fRalphCc
0x4810028166002030 65474F4C 2F637273 type 9 src/LOGe
0x7010028166002030 0102476C 60000013 type 14
0x58100A8166002030 00006504 02143B2F type 11
GSP fimrware revision B.02.20 checks the message type and
length of incoming I2C packets to prevent the issue.
5 - Local port dropping characters on input
Previous revisions of GSP firmware dropped characters during
heavy input. To eliminate the issue GSP firmware revision
B.02.20 establishes a pacing scheme based on hardware timers
preventing the issue.
6 - Added support for SNMP.
PF_CCANGSPB0217:
The Following enhancement and fixes are included in GSP
firmware revision B.02.17
In previous revisions of GSP frimware the modem did not
consistently connect. After multiple connection the GSP
would not recognize the carrier. GSP firmware revision
B.02.20 corrects this problem.
Long packets on the I2C bus potentially causing system
shutdown. The following information describes the issue in
detail.
Problem Details
************************************************************
1) Sudden and unexpected system reboots. Depending on the
setting of the Autoboot flag, the system will either reboot
to HP-UX or will reset to Boot Command Handler (BCH) for no
apparent reason. There will be nothing in the
/var/tombstones directory (e.g. no HPMC events), no evidence
of an OS panic or any core files showing a panic core dump,
nothing in the OS logs (e.g. dmesg, syslog, shutdown logs,
etc.), and nothing obvious in the GSP's Error or Activity
Logs showing the reason for the reboot event(s).
2) Sudden and unexpected DC Power system shutdowns. In this
case, the main customer symptom will be that the system is
"hung" or non-responsive. The front panel "Power" LED will
be flashing GREEN - indicating that the system is running on
DC Standby voltages. Recycling the front DC Power Switch to
off and on positions will restore the DC power and the
system will reboot normally. Alternatively, turning the DC
power back on via the GSP's "PC" command will also typically
restore the DC power and will allow the system to reboot.
Again, there will not be any GSP Errors showing the apparent
reason for the DC power subsystem shutting down.
Problem symptoms
************************************************************
The best way to explain how to determine that your N or
L-Class system is indeed experiencing this problem is to
describe when it is NOT experiencing this problem.
1) The system will typically NOT be experiencing a "solid"
problem. In other words, it will typically NOT be rebooting
every time the OS is booted or reboot very frequently after
the OS is fully operational.
2) The system will typically NOT shutdown to DC standby
voltages and refuse to power back on or remain powered on.
3) There will NOT be any obvious DC Power problems logged
in the GSP Error Logs (i.e. there will NOT be any Alert
Level 14 error events due to Undervoltage or Overvoltage
failures).
4) There will NOT be any obvious reasons in the GSP Error
Logs or Activity Logs indicating any reason for the system
to reboot (e.g. no Alert Level 12 "panics", or any HPMC
events logged). Note: There MAY be alerts that look like
the following events - which are due to a Transfer of
Control (TOC) operation being performed and these log
entries are not to be confused with actual HPMC events. A
TOC event should NOT be considered the root cause of this
unexpected reboot problem.
Typical "TOC" GSP events:
************************************************************
0 PDC 0,0,0 *2 12/09/2002 18:14:41
Log Entry 0: 12/09/2002 18:14:41
Alert Level 2: Non-urgent alert or error; Keyword:
Status: 2
Logged by system firmware 0 during machine check subAct. 51
0x0000022000ff6512 0x0000000000000000
0x58000a2000ff6512 0x0000660909120e29
1 PDC 0,0,0 *2 12/09/2002 18:14:41
Log Entry 1: 12/09/2002 18:14:41
Alert Level 2: Non-urgent alert or error; Keyword:
Status: 2
Logged by system firmware 0 during machine check subAct. 53
0x0000032000ff6532 0x0000000000000000
0x58000b2000ff6532 0x0000660909120e29
Typical L3000 GSP Chassis Code "Footprints"
************************************************************
What WILL typically be logged by the GSP MAY differ between
N-Class and L-Class systems. A typical L3000 failure
signature may look like this in the GSP Activity Logs:
Log Entry # 0 :
SYSTEM NAME: HP000001
DATE: 12/19/2002 TIME: 14:01:41
ALERT LEVEL: 1 = Information only, no action required
SOURCE: 1 = processor
SOURCE DETAIL: 1 = processor general SOURCE ID: 0
PROBLEM DETAIL: 0 = no problem detail
CALLER ACTIVITY: C = system initialization STATUS: 0
CALLER SUBACTIVITY: 00 = implementation dependent
REPORTING ENTITY TYPE: E = HP-UX REPORTING ENTITY ID: 00
0xA0E000101100C000 00000000 000007E1
0x58E008101100C000 00006608 130E0129
Log Entry # 1 :
SYSTEM NAME: HP000001
DATE: 12/19/2002 TIME: 13:57:20
ALERT LEVEL: 1 = Information only, no action required
SOURCE: 6 = platform
SOURCE DETAIL: 6 = service processor SOURCE ID: 0
PROBLEM DETAIL: 1 = selftest result
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 03 = console
REPORTING ENTITY TYPE: 1 = service processor ENTITY ID: 00
0xF010011166002030 00000000 00000000
0x5810091166002030 00006608 130D3914
Log Entry # 2 :
SYSTEM NAME: HP000001
DATE: 12/19/2002 TIME: 13:57:20
ALERT LEVEL: 10 = Boot possible, functionality lost
SOURCE: 3 = PDH
SOURCE DETAIL: 6 = interconnect medium SOURCE ID: 0
PROBLEM DETAIL: 3 = non-responding, may need GSP reset.
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 02 = platform internal interconnect
REPORTING ENTITY TYPE: 1 = service processor ID: 00
0x581008A336002020 00006608 130D3914
************************************************************
This appears to show an Alert Level 10 "Platform Medium Not
Responding" followed by a GSP Selftest Reset followed by a
"Major Change In System State" due to HPUX Initialization.
In some cases, the error only shows up as the following GSP
Chassis Code sequence - which simply shows the system
rebooting:
************************************************************
0 PDC 0,0,0 1 12/09/2002.20:30:32
Log Entry 0: 12/09/2002.20:30:32
Alert Level 1: Forward progress, no failure; Keyword:
Platform 0 ; Status: 0
Logged by firmware 0 during configuration subActivity 6f
System state change: Booting
LEDs : RUN: FLASHING green; ATTENTION: OFF yellow;
STOP: RESERVED red
0xa0000310600026f0 0x0000000000000710
0x58000b10600026f0 0x0000660909111e20
1 PDC 0,0,0 1 12/09/2002.20:30:32
Log Entry 1: 12/09/2002.20:30:32
Alert Level 1: Forward progress, no failure; Keyword:
Processor 0 installed; Status: 2
Logged by system firmware 0 during test subActivity 0
0x0000051c11001002 0x0000000000000000
0x58000d1c11001002 0x0000660909111e20
************************************************************
Typical N4000/rp7400 GSP Chassis Code "Footprint"
The N4000/rp7400 system's GSP Logs typically only shows the
following entries:
************************************************************
0 HPUX 0,0,1 *13 12/04/2002 14:54:19
Log Entry 0: 12/04/2002 14:54:19
Alert Level 13: System hang detected; Keyword:
Processor 0 timeout; Status: 0
Logged by HP-UX 1 during display_activity update subAct 0
Activity Level/Timeout: 0% / Timer expired
0x78e017d41100f000 0x0000000300000000
0x58e01fd41100f000 0x00006609040e3613
1 PDC 0,0,1 1 12/04/2002 14:59:21
Log Entry 1: 12/04/2002 14:59:21
Alert Level 1: Forward progress, no failure; Keyword:
Platform 0 ; Status: 0
Logged by firmware 1 during configuration subActivity 6f
System state change: Booting
LEDs : RUN: FLASHING green; ATTENTION: OFF yellow;
STOP: RESERVED red
0xa0001710600026f0 0x0000000000000710
0x58001f10600026f0 0x00006609040e3b15
2 PDC 0,0,1 1 12/04/2002 14:59:22
Log Entry 2: 12/04/2002 14:59:22
Alert Level 1: Forward progress, no failure; Keyword:
Processor 0 deinstalled; Status: 2
Logged by system firmware 1 during test subActivity 0
0x0000151d11001002 0x0000000000000000
0x58001d1d11001002 0x00006609040e3b16
************************************************************
Another possible chassis code "footprint" would be this
Alert Level 13 Error:
************************************************************
Log Entry # 5 :
SYSTEM NAME: HP000001
DATE: 12/15/2002 TIME: 06:06:48
ALERT LEVEL: 13 = System hang detected via timer popping
SOURCE: 6 = platform
SOURCE DETAIL: 7 = power monitor SOURCE ID: 0
PROBLEM DETAIL: 4 = internal failure
CALLER ACTIVITY: 1 = initialization STATUS: F
CALLER SUBACTIVITY: 01 = self
REPORTING ENTITY TYPE: 2 = power monitor ID: 00
0x002000D46700101F 00000000 00000000
0x582008D46700101F 00006608 0F060630
************************************************************
There may or may not always be an Alert Level 13 event
logged and in those cases the only typical indication of the
reboot condition will typically be the Alert Level 1 "System
stage change: Booting" logged by "system firmware" as is
shown in log event #1 above.
In many cases where there is nothing very obvious in the GSP
Error or Activity Logs,there might be a bit more information
in the OS Chassis Code Logs. These logs can be gathered in 2
ways:
1) Via the "cclogview" utility. This utility, if it is
installed on the system, will format the OS Chassis Code
logs into English and the output can be redirected to an
ASCII file for analysis or emailing as follows:
" cclogview /var/stm/logs/os/ccerrlog > test"
[you can substitute any desired filename for "test" here]
2) Via STM's (CSTM or MSTM or XSTM) LOGTOOL "Chassis / View
Error Log" utility. When using LOGTOOL to gather this data,
be sure to specify "DETAILS" to get the detailed chassis
code data.
PF_CCANGSPB0215:
The Following enhancement and fixes are included in GSP
firmware revision B.02.15
Corrected an issue with the firmware update process that
prevents some systems from updating.
Updating to GSP firmware revision B.02.15 requires
updating to an update only image first. The update only
image has no functionality other than to update to Revision
B.02.15.
PF_CCANGSPB0214:
The Following enhancement and fixes are included in GSP
firmware revision B.02.14
The console context feature has been removed.
PF_CCANGSPB0213:
The following enhancement and fixes are included in GSP
firmware revision B.02.13
1.A user joining a mirrored GSP session can inadvertently
complete another user's command.
When two (or more) users are mirroring in Console mode
now only one user will have write
capability at a time. Messages have been included to let
users know how to get ownership (write capability) to the
console when they need to have write capability.
Users who do not have Console Write Ownership will not be
able to type anything until the user requests (and is
granted) Console Write Ownership. The user needs to type
in ^Ecf (that is, hold control key down while typing
'e', release the control key for the 'c' & 'f') to be able
to write while in console mode.
The last 20 lines of text from the console log are now
output to serve as context. Modem and telnet also print the
last 20 lines of text from the console log when joining a
unix console session.The context will also print when
returning from the SE and CSP commands.
2.Added SSL authentication to the web access.
3.Allow the LAN speed to be forced to either 10BT Full
Duplex or 10BT Half Duplex.
4.In previous versions of GSP firmware the LAN connection
was reset after 1 minute of inactivity. When the LAN port
is connected to various switches this reset was seen as an
error by the switch and created a log entry in the switch.
Firmware revision B.02.13 pings the gateway and subnet
prior to resetting the LAN port.
5.Added a warning to notify the user in the LC and LS
commands that having port 23 selected as the web port will
disable encryption regardless of the security setting
in the SO command.
6.Added a warning message after 3 failed attempts or the
default set with the SO command that the Web Console
connection has been closed.
7.With previous versions of GSP firmware when a MUX is
attached to the UPS port and there was communication on
the port the console would hang and require the system
to be rebooted. Firmware revision B.02.13 corrects this
problem.
8.With Previous versions of GSP firmware when a UPS was
connected to the UPS port an HPMC Would occur when the
GSP was reset with the "XD R" command, an stty reset
command or by pushing the reset button. Firmware revision
B.02.13 corrects this problem.
9. HP has identified a problem on the GSP, which manifests
itself in a number of different ways. Each symptom is
listed below separately in order to facilitate matching
the symptoms to problems experienced. The cause of these
symptoms is fixed with firmware revision B.02.13.
PDC Chassis Code 2083
Intermittent Alert Level 4 events occurring during power
on Selftest/PDC Initialization on L-Series systems with
the A6696B (Maestro GSP). The Alert Level 4 error would
look similar to this:
--------------------------------------------------------
SYSTEM NAME:
DATE: 11/29/2001 TIME: 16:22:52
ALERT LEVEL: 4 = Unexpected configuration change detected
REASON FOR ALERT
SOURCE: 8 = I/O
SOURCE DETAIL: 3 = local bus adapter SOURCE ID: FF
PROBLEM DETAIL: 3 = function failure
LEDs: RUN ATTENTION FAULT REMOTE POWER
FLASH OFF OFF OFF ON
LED State: Running non-OS code. (i.e. Boot or Diagnostics)
0x2000004383FF2083 000000FF FF05FF83 - type 4 = Physical
Location
0x5800084383FF2083 0000650A 1D101634 - type 11 =Timestamp
11/29/2001 16:22:52
---------------------------------------------------------
The most significant "signature" of this Alert/GSP Error
is the chassis code ending in "2083" (e.g.
0x2000004383FF2083). This chassis code is associated
with a PDC Selftest Entry Point of
"CC_IO_PM_NOT_RESPONDING_ERROR". Typically associated
with this Alert would be a "power off" of a PCI I/O slot
- with the I/O slot associated with the HPA address
correlating to the HPA address of the "Physical Location"
field of this chassis code.
EMS Event Code 32
The main symptom is the following EMS error event which
instructs the customer to reset the GSP or replace the
system board. This EMS error may be a recurring event
(in the worst case being logged once daily).
The EMS error will show up with this wording:
>---- Event Monitoring Service Event Notification ----<
Notification Time: Sun Dec 16 17:46:54 2001
niscit sent Event Monitor notification information:
/system/events/core_hw/core_hw is >= 3.
Its current value is SERIOUS(4).
Event data from monitor:
Event Time..........: Sun Dec 16 17:46:54 2001
Severity............: SERIOUS
Monitor.............: dm_core_hw
Event #.............: 32
System..............: niscit
EMS Event Code 31 or 36
An EMS error event similar to the one below which
instructs the customer to reset the GSP or replace the
Platform Monitor or System Board. Note that you
will receive an EMS 36 event for a system with a
"platform monitor" (A-class and L-class) and an EMS
event 31 for a system with a "power monitor" (N-class)
>---- Event Monitoring Service Event Notification ----<
Event Notification Time: Thu Feb 7 06:06:02 2002
jmcdb2 sent Event Monitor notification information:
/system/events/core_hw/core_hw is >= 1.
Its current value is SERIOUS(4).
Event data from monitor:
Event Time..........: Thu Feb 7 06:06:02 2002
Severity............: SERIOUS
Monitor.............: dm_core_hw
Event #.............: 36
System..............: jmcdb2
PDC 14c0 Chassis Code
Intermittent Alert Level 6 GSP errors similar to the one
below:
--------------------------------------------------------
SYSTEM NAME:
DATE: 01/29/2002 TIME: 09:47:14
ALERT LEVEL: 6 = Boot possible, pending failure - action
required
SOURCE: 8 = I/O
SOURCE DETAIL: 9 = serial SOURCE ID: 0
PROBLEM DETAIL: 3 = functional failure
CALLER ACTIVITY: 1 = test STATUS: 0
CALLER SUBACTIVITY: 4C = implementation dependent
REPORTING ENTITY TYPE: 0 = system firmware REPORTING
ENTITY ID: 00
0x00000063890014C0 00000000 00000000 type 0 = Data Field
Unused
0x58000863890014C0 00006600 1D092F0E type 11 = Timestamp
01/29/2002 09:47:14
----------------------------------------------------------
Note! This same GSP Alert (14c0) message can also be
generated when the system console path is set ncorrectly.
The console path for the Maestro (A6696B) GSP is 0/0/4/1.
The console path for the Tosca (A6696A) GSP is 0/0/4/0.
If you experience this problem, verify the correct
console path setting first
Orphan Log Entries
One or more GSP Alert Level 12 events similar to the one
below:
--------------------------------------------------------
SYSTEM NAME:
DATE: 02/12/2002 TIME: 11:10:40
ALERT LEVEL: 12 = Software failure
SOURCE: 1 = processor
SOURCE DETAIL: 1 = processor general SOURCE ID: 0
PROBLEM DETAIL: A = orphan log entry follows
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 00 = unspecified
REPORTING ENTITY TYPE: 1 = service processor REPORTING
ENTITY ID: 00
0x581008CA11002000 00006601 0C0B0A28 type 11 = Timestamp
02/12/2002 11:10:40
Corrupt Log Entries
One or more GSP Alert Level 12 events similar to the one
below:
--------------------------------------------------------
SYSTEM NAME:
DATE: 02/12/2002 TIME: 11:10:40
ALERT LEVEL: 12 = Software failure
SOURCE: 1 = processor
SOURCE DETAIL: 1 = processor general SOURCE ID: 0
PROBLEM DETAIL: A = corrupt log entries
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 00 = unspecified
REPORTING ENTITY TYPE: 1 = service processor REPORTING
ENTITY ID: 00
0x581008CA11002000 00006601 0C0B0A28 type 11 = Timestamp
02/12/2002 11:10:40
Alert 10/2020 Chassis Code
Intermittent GSP Alert Level 10 errors when rebooting the
system multiple times (e.g. running a continuous power
on/off test that reboots the system over and over). The
GSP Alert will appear like this:
--------------------------------------------------------
Log Entry # 0 :
SYSTEM NAME: uninitialized
DATE: 02/05/2002 TIME: 06:38:50
ALERT LEVEL: 10 = Boot possible, functionality lost
SOURCE: 3 = PDH
SOURCE DETAIL: 6 = interconnect medium SOURCE ID: 0
PROBLEM DETAIL: 3 = non-responding, may need GSP reset.
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 02 = platform internal interconnect
REPORTING ENTITY TYPE: 1 = service processor REPORTING
ENTITY ID: 00
0x581008A336002020 00006601 05062632 type 11 = Timestamp
02/05/2002
Alert 8/2030 Chassis Code
Intermittent GSP Alert Level 8 Error Log entries similar
to one or more of the events below. Note that the main
footprint signature in these GSP errors is a chassis
code ending in "2030" and unusual ACSII messages.
--------------------------------------------------------
SYSTEM NAME:
DATE: 02.20/2002 TIME: 19:46:53
ALERT LEVEL: 8 = Boot Possible, performance impaired
SOURCE: 6 = platform
SOURCE DETAIL: 6 = service processor SOURCE ID: 0
PROBLEM DETAIL: 1 = selftest result
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 03 = console
REPORTING ENTITY TYPE: 1 = service processor REPORTING
ENTITY ID: 00
0x7010028166002030 0B0010FF 00000000 type 14 = Problem
Detail
0x4810028166002030 53415332 52616C70 type 9 = ASCII
Message SAS2Ralp
0x4810028166002030 7372632F 4C4F4765 type 9 = ASCII
Message src/LOGe
0x7010028166002030 0106013C 60000013 type 14 = Problem
Detail
0x58100A8166002030 00006601 11132E35 type 11 = Timestamp
02.20/2002 19:46:53
Defect Description:
PF_CCANGSPB0220:
- See Symptoms.
PF_CCANGSPB0217:
- See Symptoms.
PF_CCANGSPB0215:
- See Symptoms.
PF_CCANGSPB0214:
- See Symptoms.
PF_CCANGSPB0213:
- See Sypmtoms.
Enhancement:
Yes
PF_CCANGSPB0220:
- See Symptoms.
SR:
0000000000
Patch Files:
B0220.bin
Resources.out
gsp_upg.cnf
version.dat
cksum(1) Output:
863102829 2175924 B0220.bin
1537136308 504267 Resources.out
3789819483 200 gsp_upg.cnf
4278717776 16 version.dat
Patch Conflicts: None
Patch Dependencies:
s800: 11.00: PHNE_27393 PHCO_27370
s800: 11.11: PHNE_26326 PHCO_27243
Hardware Dependencies: None
Other Dependencies: None
Supersedes:
PF_CCANGSPB0213 PF_CCANGSPB0214 PF_CCANGSPB0215 PF_CCANGSPB0217
Equivalent Patches: None
Patch Package Size: 1173 KBytes
Installation Instructions:
Please review all instructions and the "Hewlett-Packard
Support Tool License Terms" or your Hewlett-Packard support
terms and conditions for precautions, scope of license,
restrictions, and, limitation of liability and warranties,
before installing this patch.
NOTE: It is important that you read and understand these
instructions completely before you begin. This can
determine your success in completing the Firmware update.
Warning:If the currently installed firmware is older than
B.02.15 then updating to GSP firmware revision B.02.20
requires updating to B.02.15 first. B.02.15 is located at
the following URL:
ftp://us-ffs.external.hp.com/superseded_patches
/firmware_patches/hp/cpu/PF_CCANGSPB0215.tar.gz
The Firmware update take no more than 5 minutes.
* * * * * * * * VERIFYING CURRENT REVISION OF FIRMWARE * * * * * * * *
1. telnet into the GSP.
2. logon to the GSP.
3. Type <CNTRL>ecf.
4. Type <CNTRL>b.
5. Type he
If the Revision number of the currently installed firmware is not
B.02.20 then install new firmware.
* * * * * * * * * * UNPACKING THE FIRMWARE * * * * * * * * * * * * * *
1. FTP "PF_CCANGSPB0220.tar.gz" to an empty directory.
2. Use the "gunzip" to unzip the patch.
# gunzip PF_CCANGSPB0220.tar.gz
3. Use the "tar" command to extract the firmware files
# tar -xvf *tar
x B0220.bin, 2175924 bytes, 4250 tape blocks
x Resources.out, 504267 bytes, 985 tape blocks
x gsp_upg.cnf, 200 bytes, 1 tape blocks
x version.dat, 16 bytes, 1 tape blocks
4. Verify the checksum of the files in the temporary directory
using the "cksum" command. Results of the command should be
equal to the following ouput.
# cksum *
863102829 2175924 B0220.bin
1537136308 504267 Resources.out
3789819483 200 gsp_upg.cnf
4278717776 16 version.dat
* * * * * * * * * * * FIRMWARE UPDATE PROCEDURE * * * * * * * * * * *
To update the firmware on the GSP, perform the following steps:
FTP the files to a system with the core I/O lan on the same subnet as the
GSP
1. Access the GSP using the ctrl+b command from the local console keyboard.
2. Ping the LAN where the GSP files are located via the XD command item #3
(LAN access PING)
-----------------------------------------------------------------------------
GSP> xd
XD
Non destructive tests :
1. Parameters checksum
2. I2C access (get Power Monitor status)
3. LAN access (PING)
4. Modem selftests
5. Secondary I2C access (get System status)
Type R to reset the GSP or [Q] to quit the diagnostic menu.
-> Choice: 3
3
Enter IP Address: 192.151.11.37/or your system (core I/O) LAN address
This test must pass on the server being used to supply the update files.
--------------------------------------------------------------------------------
3. Reset the GSP via the XD command (use the R option). This will free up any
unallocated memory for the GSP update.
--------------------------------------------------------------------------------
GSP> xd
XD
Non destructive tests :
1. Parameters checksum
2. I2C access (get Power Monitor status)
3. LAN access (PING)
4. Modem selftests
5. Secondary I2C access (get System status)
Type R to reset the GSP or [Q] to quit the diagnostic menu.
-> Choice: r
r
The GSP is now being reset...
--------------------------------------------------------------------------------
4. Select the GSP command XU. This command activates the upgrade mode.
Update the GSP firmware to the B.02.20 firmware image.
--------------------------------------------------------------------------------
GSP> xu
XU
This command activates the upgrade mode. All connections will be closed,the sess
ion
will be aborted and the modem connection will be dropped immediately, web and
telnet connections will be dropped upon completion. Please, confirm your intenti
on
to activate the upgrade mode (Y/[N]) : y
y
Enter source system IP address: 192.151.11.37/or your system (core I/O) LAN addr
ess
15.8.149.21
Enter file path: path on your system used to ftp files IE:/tmp/GSP
Do you wish to use the default login: anonymous / GSP
hp.com (Y/[N]) :
y (if using docs.hp.com)
n (if using local server and local server login is needed).
IE: if root then supply root password
GSP> xu
-> GSP firmware upgrade in progressM-^E.
Retrieved an upgrade file sucessfully.
Programming ROM. Percent Complete: 100.
Retrieved an upgrade file sucessfully.
Programming ROM. Percent Complete: 100.
-> GSP firmware upgrade complete - Web and telnet connections will
be dropped. GSP will now reset....
Service Processor login:
Service Processor password:
* * * * * * * * * * * * * * END OF UPDATE * * * * * * * * * * * * * *
-----End of Document ID: PF_CCANGSPB0220-------------------------------------
Document ID: PF_CHARGSPC0214
Date Loaded: 20030519
Title: rp24xx revision B GSP firmware C.02.14
Patch Name: PF_CHARGSPC0214
Patch Description: rp24xx revision B GSP firmware C.02.14
Creation Date: 03/05/01
Post Date: 03/05/19
Products:
A6889A rp2430 server
A6890A rp2470 server
Status: General Release
Critical: No
Category Tags:
defect_repair hardware_enablement enhancement
general_release
Path Name: /firmware_patches/hp/cpu/PF_CHARGSPC0214
Symptoms:
PF_CHARGSPC0214:
The Following enhancement and fixes are included in GSP
firmware revision C.02.14
************************************************************
The Java plug-in 1.4x is not supported with Web Access. Java
plug-in 1.3x is supportted.
************************************************************
1- Run LED is displayed OFF after "System Boot detected"
In previous revisions of gsp firmware the front panel RUN
led did not always flash and would instead appear to be off.
The following is the correct display from the Vitual Font
Panel.
********** VIRTUAL FRONT PANEL **********
System Boot detected
*****************************************
LEDs: RUN ATTENTION FAULT REMOTE POWER
The following is a sample incorrect display from the Virtual
Front Panel.
********** VIRTUAL FRONT PANEL **********
System Boot detected
*****************************************
LEDs: RUN ATTENTION FAULT REMOTE POWER
OFF FLASH OFF ON ON
C.02.14 has been modified to correct this issue.
2 - LAN hangs when gateway address matches GSP's own IP
address.
In previous revisionss of GSP firmware When GSP's gateway
address and ip address matched and the LAN hung the firmware
pinged the gateway address to dectect if a real hang
occured. If the gateway and ip addresses were the same, the
ping always return OK status incorrectly indicating that
the LAN is active. As a result the LAN would not
reinitialize, leaving the LAN indefinitely in a hang state.
C.02.14 firmware checks the gateway IP is checked against
the GSP IP before attempting to ping gateway. If the gateway
and IP match then the broadcast address is pinged instead.
3 - GSP denies CTRL-Ecf access
During extremely high traffic on the I2C bus CTRL-Ecf did
not work properly and Write access to the GSP would not
be granted.The serial port connection continued to work and
the OS was not affected. This issue was corrected with GSP
firmware revision C.02.14.
4 - Data aborts caused by CRC invalid Voice packets.
In previous revisions of GSP firmware the checksum
verication result on i2c Voice messages is ignored when
returning to upper level calling routines. If the length
byte in the Voice message is corrupted this will result
later in a data abort, when the message is copied to a
secondary location for processing. The following is a
sample error chassis code:
ALERT LEVEL: 8 = Boot Possible, performance impaired
SOURCE: 6 = platform
SOURCE DETAIL: 6 = service processor SOURCE ID: 0
PROBLEM DETAIL: 1 = selftest result
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 03 = console
REPORTING ENTITY TYPE: 1 = service proc ID: 00
0x7010028166002030 0B0010FF 00000000 type 14
0x4810028166002030 63436870 6C615266 type 9 fRalphCc
0x4810028166002030 65474F4C 2F637273 type 9 src/LOGe
0x7010028166002030 0102476C 60000013 type 14
0x58100A8166002030 00006504 02143B2F type 11
GSP fimrware revision C.02.14 checks the message type and
length of incoming I2C packets to prevent the issue.
5 - Local port dropping characters on input
Previous revisions of GSP firmware dropped characters during
heavy input. To eliminate the issue GSP firmware revision
C.02.14 establishes a pacing scheme based on hardware timers
preventing the issue.
6 - Added support for SNMP.
PF_CHARGSPC0212:
The following enhancements and fixes are included in
GSP firmware revision C.02.12
In previous revisions of GSP frimware the modem did not
consistently connect. After multiple connection the GSP
would not recognize the carrier. GSP firmware revision
C.02.12 corrects this problem.
Long packets on the I2C bus potentially causing system
shutdown. The following information describes the issue in
detail.
Problem Details
************************************************************
1) Sudden and unexpected system reboots. Depending on the
setting of the Autoboot flag, the system will either reboot
to HP-UX or will reset to Boot Command Handler (BCH) for no
apparent reason. There will be nothing in the
/var/tombstones directory (e.g. no HPMC events), no evidence
of an OS panic or any core files showing a panic core dump,
nothing in the OS logs (e.g. dmesg, syslog, shutdown logs,
etc.), and nothing obvious in the GSP's Error or Activity
Logs showing the reason for the reboot event(s).
2) Sudden and unexpected DC Power system shutdowns. In this
case, the main customer symptom will be that the system is
"hung" or non-responsive. The front panel "Power" LED will
be flashing GREEN - indicating that the system is running on
DC Standby voltages. Recycling the front DC Power Switch to
off and on positions will restore the DC power and the
system will reboot normally. Alternatively, turning the DC
power back on via the GSP's "PC" command will also typically
restore the DC power and will allow the system to reboot.
Again, there will not be any GSP Errors showing the apparent
reason for the DC power subsystem shutting down.
Problem symptoms
************************************************************
The best way to explain how to determine that your N or
L-Class system is indeed experiencing this problem is to
describe when it is NOT experiencing this problem.
1) The system will typically NOT be experiencing a "solid"
problem. In other words, it will typically NOT be rebooting
every time the OS is booted or reboot very frequently after
the OS is fully operational.
2) The system will typically NOT shutdown to DC standby
voltages and refuse to power back on or remain powered on.
3) There will NOT be any obvious DC Power problems logged
in the GSP Error Logs (i.e. there will NOT be any Alert
Level 14 error events due to Undervoltage or Overvoltage
failures).
4) There will NOT be any obvious reasons in the GSP Error
Logs or Activity Logs indicating any reason for the system
to reboot (e.g. no Alert Level 12 "panics", or any HPMC
events logged). Note: There MAY be alerts that look like
the following events - which are due to a Transfer of
Control (TOC) operation being performed and these log
entries are not to be confused with actual HPMC events. A
TOC event should NOT be considered the root cause of this
unexpected reboot problem.
Typical "TOC" GSP events:
************************************************************
0 PDC 0,0,0 *2 12/09/2002 18:14:41
Log Entry 0: 12/09/2002 18:14:41
Alert Level 2: Non-urgent alert or error; Keyword:
Status: 2
Logged by system firmware 0 during machine check subAct. 51
0x0000022000ff6512 0x0000000000000000
0x58000a2000ff6512 0x0000660909120e29
1 PDC 0,0,0 *2 12/09/2002 18:14:41
Log Entry 1: 12/09/2002 18:14:41
Alert Level 2: Non-urgent alert or error; Keyword:
Status: 2
Logged by system firmware 0 during machine check subAct. 53
0x0000032000ff6532 0x0000000000000000
0x58000b2000ff6532 0x0000660909120e29
Typical L3000 GSP Chassis Code "Footprints"
************************************************************
What WILL typically be logged by the GSP MAY differ between
N-Class and L-Class systems. A typical L3000 failure
signature may look like this in the GSP Activity Logs:
Log Entry # 0 :
SYSTEM NAME: HP000001
DATE: 12/19/2002 TIME: 14:01:41
ALERT LEVEL: 1 = Information only, no action required
SOURCE: 1 = processor
SOURCE DETAIL: 1 = processor general SOURCE ID: 0
PROBLEM DETAIL: 0 = no problem detail
CALLER ACTIVITY: C = system initialization STATUS: 0
CALLER SUBACTIVITY: 00 = implementation dependent
REPORTING ENTITY TYPE: E = HP-UX REPORTING ENTITY ID: 00
0xA0E000101100C000 00000000 000007E1
0x58E008101100C000 00006608 130E0129
Log Entry # 1 :
SYSTEM NAME: HP000001
DATE: 12/19/2002 TIME: 13:57:20
ALERT LEVEL: 1 = Information only, no action required
SOURCE: 6 = platform
SOURCE DETAIL: 6 = service processor SOURCE ID: 0
PROBLEM DETAIL: 1 = selftest result
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 03 = console
REPORTING ENTITY TYPE: 1 = service processor ENTITY ID: 00
0xF010011166002030 00000000 00000000
0x5810091166002030 00006608 130D3914
Log Entry # 2 :
SYSTEM NAME: HP000001
DATE: 12/19/2002 TIME: 13:57:20
ALERT LEVEL: 10 = Boot possible, functionality lost
SOURCE: 3 = PDH
SOURCE DETAIL: 6 = interconnect medium SOURCE ID: 0
PROBLEM DETAIL: 3 = non-responding, may need GSP reset.
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 02 = platform internal interconnect
REPORTING ENTITY TYPE: 1 = service processor ID: 00
0x581008A336002020 00006608 130D3914
************************************************************
This appears to show an Alert Level 10 "Platform Medium Not
Responding" followed by a GSP Selftest Reset followed by a
"Major Change In System State" due to HPUX Initialization.
In some cases, the error only shows up as the following GSP
Chassis Code sequence - which simply shows the system
rebooting:
************************************************************
0 PDC 0,0,0 1 12/09/2002 17:30:32
Log Entry 0: 12/09/2002 17:30:32
Alert Level 1: Forward progress, no failure; Keyword:
Platform 0 ; Status: 0
Logged by firmware 0 during configuration subActivity 6f
System state change: Booting
LEDs : RUN: FLASHING green; ATTENTION: OFF yellow;
STOP: RESERVED red
0xa0000310600026f0 0x0000000000000710
0x58000b10600026f0 0x0000660909111e20
1 PDC 0,0,0 1 12/09/2002 17:30:32
Log Entry 1: 12/09/2002 17:30:32
Alert Level 1: Forward progress, no failure; Keyword:
Processor 0 installed; Status: 2
Logged by system firmware 0 during test subActivity 0
0x0000051c11001002 0x0000000000000000
0x58000d1c11001002 0x0000660909111e20
************************************************************
Typical N4000/rp7400 GSP Chassis Code "Footprint"
The N4000/rp7400 system's GSP Logs typically only shows the
following entries:
************************************************************
0 HPUX 0,0,1 *13 12/04/2002 14:54:19
Log Entry 0: 12/04/2002 14:54:19
Alert Level 13: System hang detected; Keyword:
Processor 0 timeout; Status: 0
Logged by HP-UX 1 during display_activity update subAct 0
Activity Level/Timeout: 0% / Timer expired
0x78e017d41100f000 0x0000000300000000
0x58e01fd41100f000 0x00006609040e3613
1 PDC 0,0,1 1 12/04/2002 14:59:21
Log Entry 1: 12/04/2002 14:59:21
Alert Level 1: Forward progress, no failure; Keyword:
Platform 0 ; Status: 0
Logged by firmware 1 during configuration subActivity 6f
System state change: Booting
LEDs : RUN: FLASHING green; ATTENTION: OFF yellow;
STOP: RESERVED red
0xa0001710600026f0 0x0000000000000710
0x58001f10600026f0 0x00006609040e3b15
2 PDC 0,0,1 1 12/04/2002 14:59:22
Log Entry 2: 12/04/2002 14:59:22
Alert Level 1: Forward progress, no failure; Keyword:
Processor 0 deinstalled; Status: 2
Logged by system firmware 1 during test subActivity 0
0x0000151d11001002 0x0000000000000000
0x58001d1d11001002 0x00006609040e3b16
************************************************************
Another possible chassis code "footprint" would be this
Alert Level 13 Error:
************************************************************
Log Entry # 5 :
SYSTEM NAME: HP000001
DATE: 12/15/2002 TIME: 06:06:48
ALERT LEVEL: 13 = System hang detected via timer popping
SOURCE: 6 = platform
SOURCE DETAIL: 7 = power monitor SOURCE ID: 0
PROBLEM DETAIL: 4 = internal failure
CALLER ACTIVITY: 1 = initialization STATUS: F
CALLER SUBACTIVITY: 01 = self
REPORTING ENTITY TYPE: 2 = power monitor ID: 00
0x002000D46700101F 00000000 00000000
0x582008D46700101F 00006608 0F060630
************************************************************
There may or may not always be an Alert Level 13 event
logged and in those cases the only typical indication of the
reboot condition will typically be the Alert Level 1 "System
stage change: Booting" logged by "system firmware" as is
shown in log event #1 above.
In many cases where there is nothing very obvious in the GSP
Error or Activity Logs,there might be a bit more information
in the OS Chassis Code Logs. These logs can be gathered in 2
ways:
1) Via the "cclogview" utility. This utility, if it is
installed on the system, will format the OS Chassis Code
logs into English and the output can be redirected to an
ASCII file for analysis or emailing as follows:
" cclogview /var/stm/logs/os/ccerrlog > test"
[you can substitute any desired filename for "test" here]
2) Via STM's (CSTM or MSTM or XSTM) LOGTOOL "Chassis / View
Error Log" utility. When using LOGTOOL to gather this data,
be sure to specify "DETAILS" to get the detailed chassis
code data.
PF_CHARGSPC0210:
The following enhancements and fixes are included in
GSP firmware revision C.02.10
Corrected an issue with the firmware update process that
prevents some systems from updating.
Updating to GSP firmware revision C.02.10 requires
updating to an update only image first. The update only
image has no functionality other than to update to Revision
C.02.10.
PF_CHARGSPC0209:
The following enhancement and fixes are included in GSP
firmware revision C.02.09
Console context has been removed from the firmware.
PF_CHARGSPC0208:
The following enhancement and fixes are included in GSP
firmware revision C.02.08
1.A user joining a mirrored GSP session can inadvertently
complete another user's command.
When two (or more) users are mirroring in Console mode
now only one user will have write
capability at a time. Messages have been included to let
users know how to get ownership (write capability) to the
console when they need to have write capability.
Users who do not have Console Write Ownership will not be
able to type anything until the user requests (and is
granted) Console Write Ownership. The user needs to type
in ^Ecf (that is, hold control key down while typing
'e', release the control key for the 'c' & 'f') to be able
to write while in console mode.
The last 20 lines of text from the console log are now
output to serve as context. Modem and telnet also print the
last 20 lines of text from the console log when joining a
unix console session.The context will also print when
returning from the SE and CSP commands.
2.Added SSL authentication to the web access.
3.In previous versions of GSP firmware the LAN connection
was reset after 1 minute of inactivity. When the LAN port
is connected to various switches this reset was seen as an
error by the switch and created a log entry in the switch.
Firmware revision C.02.08 pings the gateway and subnet
prior to resetting the LAN port.
4.Added a warning to notify the user in the LC and LS
commands that having port 23 selected as the web port will
disable encryption regardless of the security setting
in the SO command.
5.Added a warning message after 3 failed attempts or the
default set with the SO command that the Web Console
connection has been closed.
6.With previous versions of GSP firmware when a MUX is
attached to the UPS port and there was communication on
the port the console would hang and require the system
to be rebooted. Firmware revision C.02.08 corrects this
problem.
7.With Previous versions of GSP firmware when a UPS was
connected to the UPS port an HPMC Would occur when the
GSP was reset with the "XD R" command, an stty reset
command or by pushing the reset button. Firmware revision
C.02.08 corrects this problem.
8. HP has identified a problem on the GSP, which manifests
itself in a number of different ways. Each symptom is
listed below separately in order to facilitate matching
the symptoms to problems experienced. The cause of these
symptoms is fixed with firmware revision C.02.08.
PDC Chassis Code 2083
Intermittent Alert Level 4 events occurring during power
on Selftest/PDC Initialization on L-Series systems with
the A6696B (Maestro GSP). The Alert Level 4 error would
look similar to this:
--------------------------------------------------------
SYSTEM NAME:
DATE: 11/29/2001 TIME: 16:22:52
ALERT LEVEL: 4 = Unexpected configuration change detected
REASON FOR ALERT
SOURCE: 8 = I/O
SOURCE DETAIL: 3 = local bus adapter SOURCE ID: FF
PROBLEM DETAIL: 3 = function failure
LEDs: RUN ATTENTION FAULT REMOTE POWER
FLASH OFF OFF OFF ON
LED State: Running non-OS code. (i.e. Boot or Diagnostics)
0x2000004383FF2083 000000FF FF05FF83 - type 4 = Physical
Location
0x5800084383FF2083 0000650A 1D101634 - type 11 =Timestamp
11/29/2001 16:22:52
---------------------------------------------------------
The most significant "signature" of this Alert/GSP Error
is the chassis code ending in "2083" (e.g.
0x2000004383FF2083). This chassis code is associated
with a PDC Selftest Entry Point of
"CC_IO_PM_NOT_RESPONDING_ERROR". Typically associated
with this Alert would be a "power off" of a PCI I/O slot
- with the I/O slot associated with the HPA address
correlating to the HPA address of the "Physical Location"
field of this chassis code.
EMS Event Code 32
The main symptom is the following EMS error event which
instructs the customer to reset the GSP or replace the
system board. This EMS error may be a recurring event
(in the worst case being logged once daily).
The EMS error will show up with this wording:
>---- Event Monitoring Service Event Notification ----<
Notification Time: Sun Dec 16 17:46:54 2001
niscit sent Event Monitor notification information:
/system/events/core_hw/core_hw is >= 3.
Its current value is SERIOUS(4).
Event data from monitor:
Event Time..........: Sun Dec 16 17:46:54 2001
Severity............: SERIOUS
Monitor.............: dm_core_hw
Event #.............: 32
System..............: niscit
EMS Event Code 31 or 36
An EMS error event similar to the one below which
instructs the customer to reset the GSP or replace the
Platform Monitor or System Board. Note that you
will receive an EMS 36 event for a system with a
"platform monitor" (A-class and L-class) and an EMS
event 31 for a system with a "power monitor" (N-class)
>---- Event Monitoring Service Event Notification ----<
Event Notification Time: Thu Feb 7 06:06:02 2002
jmcdb2 sent Event Monitor notification information:
/system/events/core_hw/core_hw is >= 1.
Its current value is SERIOUS(4).
Event data from monitor:
Event Time..........: Thu Feb 7 06:06:02 2002
Severity............: SERIOUS
Monitor.............: dm_core_hw
Event #.............: 36
System..............: jmcdb2
PDC 14c0 Chassis Code
Intermittent Alert Level 6 GSP errors similar to the one
below:
--------------------------------------------------------
SYSTEM NAME:
DATE: 01/29/2002 TIME: 09:47:14
ALERT LEVEL: 6 = Boot possible, pending failure - action
required
SOURCE: 8 = I/O
SOURCE DETAIL: 9 = serial SOURCE ID: 0
PROBLEM DETAIL: 3 = functional failure
CALLER ACTIVITY: 1 = test STATUS: 0
CALLER SUBACTIVITY: 4C = implementation dependent
REPORTING ENTITY TYPE: 0 = system firmware REPORTING
ENTITY ID: 00
0x00000063890014C0 00000000 00000000 type 0 = Data Field
Unused
0x58000863890014C0 00006600 1D092F0E type 11 = Timestamp
01/29/2002 09:47:14
----------------------------------------------------------
Note! This same GSP Alert (14c0) message can also be
generated when the system console path is set ncorrectly.
The console path for the Maestro (A6696B) GSP is 0/0/4/1.
The console path for the Tosca (A6696A) GSP is 0/0/4/0.
If you experience this problem, verify the correct
console path setting first
Orphan Log Entries
One or more GSP Alert Level 12 events similar to the one
below:
--------------------------------------------------------
SYSTEM NAME:
DATE: 02.14/2002 TIME: 11:10:40
ALERT LEVEL: 12 = Software failure
SOURCE: 1 = processor
SOURCE DETAIL: 1 = processor general SOURCE ID: 0
PROBLEM DETAIL: A = orphan log entry follows
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 00 = unspecified
REPORTING ENTITY TYPE: 1 = service processor REPORTING
ENTITY ID: 00
0x581008CA11002000 00006601 0C0B0A28 type 11 = Timestamp
02.14/2002 11:10:40
Corrupt Log Entries
One or more GSP Alert Level 12 events similar to the one
below:
--------------------------------------------------------
SYSTEM NAME:
DATE: 02.14/2002 TIME: 11:10:40
ALERT LEVEL: 12 = Software failure
SOURCE: 1 = processor
SOURCE DETAIL: 1 = processor general SOURCE ID: 0
PROBLEM DETAIL: A = corrupt log entries
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 00 = unspecified
REPORTING ENTITY TYPE: 1 = service processor REPORTING
ENTITY ID: 00
0x581008CA11002000 00006601 0C0B0A28 type 11 = Timestamp
02.14/2002 11:10:40
Alert 10/2020 Chassis Code
Intermittent GSP Alert Level 10 errors when rebooting the
system multiple times (e.g. running a continuous power
on/off test that reboots the system over and over). The
GSP Alert will appear like this:
--------------------------------------------------------
Log Entry # 0 :
SYSTEM NAME: uninitialized
DATE: 02/05/2002 TIME: 06:38:50
ALERT LEVEL: 10 = Boot possible, functionality lost
SOURCE: 3 = PDH
SOURCE DETAIL: 6 = interconnect medium SOURCE ID: 0
PROBLEM DETAIL: 3 = non-responding, may need GSP reset.
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 02 = platform internal interconnect
REPORTING ENTITY TYPE: 1 = service processor REPORTING
ENTITY ID: 00
0x581008A336002020 00006601 05062632 type 11 = Timestamp
02/05/2002
Alert 8/2030 Chassis Code
Intermittent GSP Alert Level 8 Error Log entries similar
to one or more of the events below. Note that the main
footprint signature in these GSP errors is a chassis
code ending in "2030" and unusual ACSII messages.
--------------------------------------------------------
SYSTEM NAME:
DATE: 02/17/2002 TIME: 19:46:53
ALERT LEVEL: 8 = Boot Possible, performance impaired
SOURCE: 6 = platform
SOURCE DETAIL: 6 = service processor SOURCE ID: 0
PROBLEM DETAIL: 1 = selftest result
CALLER ACTIVITY: 2 = operation STATUS: 0
CALLER SUBACTIVITY: 03 = console
REPORTING ENTITY TYPE: 1 = service processor REPORTING
ENTITY ID: 00
0x7010028166002030 0B0010FF 00000000 type 14 = Problem
Detail
0x4810028166002030 53415332 52616C70 type 9 = ASCII
Message SAS2Ralp
0x4810028166002030 7372632F 4C4F4765 type 9 = ASCII
Message src/LOGe
0x7010028166002030 0106013C 60000013 type 14 = Problem
Detail
0x58100A8166002030 00006601 11132E35 type 11 = Timestamp
02/17/2002 19:46:53
Defect Description:
PF_CHARGSPC0214:
- See Symptoms.
PF_CHARGSPC0212:
- See Symptoms.
PF_CHARGSPC0210:
- See Symptoms.
PF_CHARGSPC0209:
- See Symptoms.
PF_CHARGSPC0208:
- See Sypmtoms.
Enhancement:
Yes
PF_CHARGSPC0214:
- See Symptoms.
SR:
0000000000
Patch Files:
C0214.bin
Resources.out
gsp_a_upg.cnf
version.dat
cksum(1) Output:
3495409698 2173644 C0214.bin
1537136308 504267 Resources.out
2488838608 200 gsp_a_upg.cnf
4278717776 16 version.dat
Patch Conflicts: None
Patch Dependencies:
s800: 11.00: PHNE_27393 PHCO_27370
s800: 11.11: PHNE_26326 PHCO_27243
Hardware Dependencies: None
Other Dependencies: None
Supersedes:
PF_CHARGSPC0208 PF_CHARGSPC0209 PF_CHARGSPC0210 PF_CHARGSPC0212
Equivalent Patches: None
Patch Package Size: 1172 KBytes
Installation Instructions:
Please review all instructions and the "Hewlett-Packard
Support Tool License Terms" or your Hewlett-Packard support
terms and conditions for precautions, scope of license,
restrictions, and, limitation of liability and warranties,
before installing this patch.
NOTE: It is important that you read and understand these
instructions completely before you begin. This can
determine your success in completing the Firmware update.
Warning:If the currently installed firmware is older than
C.02.10 then updating to GSP firmware revision C.02.14
requires updating to C.02.10 first. C.02.10 is located at
the following URL:
ftp://us-ffs.external.hp.com/superseded_patches
/firmware_patches/hp/cpu/PF_CHARGSPC0210.tar.gz
The Firmware update take no more than 5 minutes.
* * * * * * * * VERIFYING CURRENT REVISION OF FIRMWARE * * * * * * * *
1. telnet into the GSP.
2. logon to the GSP.
3. Type <CNTRL>ecf.
4. Type <CNTRL>b.
5. Type he
If the Revision number of the currently installed firmware is not
C.02.14 then install new firmware.
* * * * * * * * * * UNPACKING THE FIRMWARE * * * * * * * * * * * * * *
1. FTP "PF_CHARGSPC0214.tar.gz" to an empty directory.
2. Use the "gunzip" to unzip the patch.
# gunzip PF_CHARGSPC0214.tar.gz
3. Use the "tar" command to extract the firmware files
# tar -xvf *tar
x C0214.bin, 2173644 bytes, 4246 tape blocks
x Resources.out, 504267 bytes, 985 tape blocks
x gsp_a_upg.cnf, 200 bytes, 1 tape blocks
x version.dat, 16 bytes, 1 tape blocks
4. Verify the checksum of the files in the temporary directory
using the "cksum" command. Results of the command should be
equal to the following ouput.
# cksum *
3495409698 2173644 C0214.bin
1537136308 504267 Resources.out
2488838608 200 gsp_a_upg.cnf
4278717776 16 version.dat
* * * * * * * * * * * FIRMWARE UPDATE PROCEDURE * * * * * * * * * * *
To update the firmware on the GSP, perform the following steps:
FTP the files to a system with the core I/O lan on the same subnet as the
GSP
1. Access the GSP using the ctrl+b command from the local console keyboard.
2. Ping the LAN where the GSP files are located via the XD command item #3
(LAN access PING)
-----------------------------------------------------------------------------
GSP> xd
XD
Non destructive tests :
1. Parameters checksum
2. I2C access (get Power Monitor status)
3. LAN access (PING)
4. Modem selftests
5. Secondary I2C access (get System status)
Type R to reset the GSP or [Q] to quit the diagnostic menu.
-> Choice: 3
3
Enter IP Address: 192.151.11.37/or your system (core I/O) LAN address
This test must pass on the server being used to supply the update files.
--------------------------------------------------------------------------------
3. Reset the GSP via the XD command (use the R option). This will free up any
unallocated memory for the GSP update.
--------------------------------------------------------------------------------
GSP> xd
XD
Non destructive tests :
1. Parameters checksum
2. I2C access (get Power Monitor status)
3. LAN access (PING)
4. Modem selftests
5. Secondary I2C access (get System status)
Type R to reset the GSP or [Q] to quit the diagnostic menu.
-> Choice: r
r
The GSP is now being reset...
--------------------------------------------------------------------------------
4. Select the GSP command XU. This command activates the upgrade mode.
Update the GSP firmware to the C.02.10 firmware.
--------------------------------------------------------------------------------
GSP> xu
XU
This command activates the upgrade mode. All connections will be closed,the session
will be aborted and the modem connection will be dropped immediately, web and
telnet connections will be dropped upon completion. Please, confirm your intention
to activate the upgrade mode (Y/[N]) : y
y
Enter source system IP address: 192.151.11.37/or your system (core I/O) LAN address
15.8.149.21
Enter file path: path on your system used to ftp files IE:/tmp/GSP
Do you wish to use the default login: anonymous / GSP
hp.com (Y/[N]) :
y (if using docs.hp.com)
n (if using local server and local server login is needed).
IE: if root then supply root password
GSP> xu
-> GSP firmware upgrade in progress.
Retrieved an upgrade file sucessfully.
Programming ROM. Percent Complete: 100.
Retrieved an upgrade file sucessfully.
Programming ROM. Percent Complete: 100.
-> GSP firmware upgrade complete - Web and telnet connections will
be dropped. GSP will now reset....
Service Processor login:
Service Processor password:
* * * * * * * * * * * * * * END OF UPDATE * * * * * * * * * * * * * *
-----End of Document ID: PF_CHARGSPC0214-------------------------------------
Document ID: PF_CKEYMAT0500
Date Loaded: 20030327
Title: HP rp8400/rp7410 Firmware Version 5.0
Patch Name: PF_CKEYMAT0500
Patch Description: HP rp8400/rp7410 Firmware Version 5.0
Creation Date: 03/03/25
Post Date: 03/03/27
Products:
A6752A HP Server rp7410
A6093A HP Server rp8400
Status: General Release
Critical: No
Category Tags:
defect_repair hardware_enablement enhancement
general_release
Path Name: /firmware_patches/hp/cpu/PF_CKEYMAT0500
Symptoms:
PF_CKEYMAT0500:
Announcements, supported features, fixes and
enhancements appear in Version 5.0.
- Announcements:
This release notice provides information and
installation/upgrade instructions for the
rp8400 and rp7410 Firmware Release Version
5.0.
-The following new features have been added:
Support for 2 GB DIMMs has been added.
- Known Problems and Workarounds:
* Use of the Genesis profile command at the
MP command menu returns Invalid Sequence
IDs for complex profile groups B and C.
This result is expected, but users will
note a MP error log entry resulting from
this command.
* Chassis code errors may appear in the MP
error log. This problem may manifest
itself as an orphan chassis code
(ORPHAN_CC log entry) or an error level
13 entry.
* rp8400 Only: Several help commands
available at the Boot Console Handler
(BCH) Information Menu are not visible
from within the Information Menu.
The help commands for the following menu
items are affected:
- ChipRevision
- FwrVersion
- MEmory
- PRocessor
- BOot
- DIsplay
- HElp
- REBOOT
- RECONFIGRESET
- MAin
* When updating MP firmware from version
2.08 (Version 3.01) or version 3.005
(Version 4.0/4.11) to 4.009
(Version 5.0)the MP LAN configuration
will be overwritten with default
values. Before updating the MP
firmware, execute the MP command ls
to document the LANs configuration.
Upon completion of a successful
MP 4.009 version update, it is
necessary to connect a serial console
and execute the MP lc command to
reconfigure the MP LAN.
* Decreased installation failures when
using IGNITE over a LAN.
- Fixes in this Version:
- PDC revision 16.011 contains the following
fixes:
* Left-justified DIMM signatures in FRUID
are not recognized.
* PDC does not allow Superdome qualified
memory.
* Several PDC Procedures run out of ROM
space and should run out of RAM to
increase speed.
* Vpars fails to load specific IODC
drivers properly.
* Intermittent Single-bit Errors cause
X-Bar slices to be different, which in
turn cause the system to panic.
* PDC does not validate frequency
information received from System
Utilities.
* PDC is not robust enough handling
early PDH events.
* NVM copy of DIMM Serial-Presence Detect
information is not checked for
corruption after HPMCs.
* Flushing of LBA Cache fails during LBA
reset.
* Powered-off PCI slots report as
non-existent in cell information.
* Parstatus and Parmodify hangs in certain
configurations.
* Boot Path does not show trailing 0s
when showing Fiber-Channel path.
* Core I/O LAN fails to boot in certain
configurations.
* Booting after a search command fails if
LIF is not the first on the disk.
* PDC returns incorrect physical location
after doorbell/latch events.
* PDC event returns wrong current
latch/intrusion status.
* Core I/O does not find some bootp
servers.
* Tachlite IODC driver failures when
Brocade switches are set for exclusive
point-to-point topologies.
* Boot support for DS2405 in IODC.
* 2GB boot support of the A6795A 2GB FC
card.
- MP revision 4.009 contains the following
fix:
* Errors existing in chassis codes
generated by the loss or removal of
bulk power supplies have been
corrected.
* The MP Command Menu now contains a
help entry for Onboard Scan
Programming Utility (OSP).
* DIMM and CPU FRU Identification
information is now available at the
Operating System level.
* Bulk power supplies now correctly
report the loss of 12V or 48V systems.
* Existing complex profile and LAN
setting information is now retained
when replacing a core IO board.
* An error preventing telnet access
through the MP port and a loss of LAN
configuration information using the
'MP:CM>lc' command has been fixed.
* Total Power Available is now reported
correctly based on the number of AC
power cords available.
* The MP Command Menu help (MP:CM>he)
now directs users to the MP Main
Menu help (MP>he) for detailed help
information.
* The Display FRU Command 'MP:CM>df>all'
no longer lists components that do not
exist in the system.
* Console access is now available to the
Operating System while the MP is in
reset. Previous attempts to access the
console during reset resulted in
CC_SELECT_CONSOLE_ERROR_CONSOLE_RETRYING
chassis codes.
* rp7410 Only: Multiple fan failure
messages in Standby (48V off) no longer
prevent subsequent power-up.
PF_CKEYMAT0411:
Announcements, supported features, fixes and
enhancements appear in Version 4.11.
- Announcements:
This release notice provides information and
installation/upgrade instructions for the
rp8400 and rp7410 Firmware Release Version
4.11.
-The following new features have been added:
* Support for the 750 & 875 MHz PA-8700
version 3.0 and 3.1 processors has been
added.
* The state of the cells is now verified
before updating the PDC. If a cell is
running the OS, a Warning is now given.
* Multi-partition machines may now operate
each partition at different CPU
frequencies (e.g., Partition 0 operating
at 750 MHz with Partition 1 operating at
875 MHz).
* The new 875 MHz processors may be mixed
with PA-8700 version 2.2 and 2.3
processors on the same cell board or
within the same partition.
* Boot over Ultra160 SCSI is now supported.
* Support for 256 MB DIMMs has been added.
* The ability to update cell board FPGAs
(LPM, Bob and Drain) using the Onboard
Scan Programming (OSP) Utility has been
added.
* Three new MP command menu commands have
been added to support Secure Sockets Layer
(SSL) used for web encryption:
- Enable SSL (ES)
- Disable SSL (DS)
- SSL Certificate/Keypair Generation (CG).
- Known Problems and Workarounds:
* Use of the Genesis profile command at the
MP command menu returns Invalid Sequence
IDs for complex profile groups B and C.
This result is expected, but users will
note a MP error log entry resulting from
this command.
* Use of the Help on OSP command at the MP
prompt returns Unrecognized command or
topic message. This problem is solely
a help menu issue, and no degradation
exists in the OSP Utility functionality.
* Chassis code errors may appear in the MP
error log. This problem may manifest
itself as an orphan chassis code
(ORPHAN_CC log entry) or an error level
13 entry.
* rp8400 Only: Several help commands
available at the Boot Console Handler
(BCH) Information Menu are not visible
from within the Information Menu.
The help commands for the following menu
items are affected:
- ChipRevision
- FwrVersion
- MEmory
- PRocessor
- BOot
- DIsplay
- HElp
- REBOOT
- RECONFIGRESET
- MAin
- Fixes in this Version:
- PDC revision 16.009 contains the following
fixes:
* Numerous minor PDC-to-operating system
interface errors were corrected.
* Typographical and revision errors were
fixed in the Boot Console Handler (BCH)
menu.
* PDC now isolates single bit errors to a
single DIMM during bootstrap to the DIMM.
* CCI FIFO contents are now preserved on a
soft reset.
* The detection of address/parity errors now
results in CPU deconfiguration only and
will not halt the entire cell.
* Cells booted to BCH following CPU
replacement no longer stop at Boot is
Blocked (BIB).
* A multi-bit error in a non-root cell no
longer brings down the entire partition.
* Event Monitoring System crossbar hardware
checks no longer result in system hangs
that monopolize CPU time.
* Diskless boot (ignite) failures were
reduced by creating a larger buffer to
compensate for lost packet information.
* A copy instruction has been changed to
allow hard coding of the d-cache tag
clock settings.
- SINC revision 2.002 contains the following
fix:
* A PDC hardware semaphore request was
removed to minimize boot delays while
reading chassis codes.
- MP revision 3.005 contains the following
fixes:
* The web console now works with Netscape
6.x and Internet Explorer 5.x. It will
not work with Netscape 4.x or Netscape 5.x.
* The MP 'Console Log' option now returns the
correct partition log information.
* Command menu help now includes listings for
Disable Web Console Access (DW) and Enable
Web Console Access (EW).
* The occurrence of ethernet Media Access
Controller (MAC) hangs due to slow LAN
traffic has been minimized by changes made
to the LAN driver.
* All FPGA and firmware revision numbers are
now reported to three digits (MP 3.005
versus the previous MP 3.05).
- CIO revision 1.001:
* Was modified to add enhancements for
future generations of computers.
- Cell LPM revision 2.002:
* Was modified to improve VRM noise
immunity.
PF_CRP84000400:
Announcements, supported features, fixes and
enhancements appear in Version 4.0.
- Announcements:
This release notice provides information and
installation/upgrade instructions for the
rp8400 Firmware Release Version 4.0.
- The following new features have been added:
* Support for the 875 MHz PA-8700 version
3.0 and 3.1 processors has been added.
* Multi-partition machines may now operate
each partition at different CPU
frequencies (e.g., Partition 0 operating
at 750 MHz with Partition 1 operating at
875 MHz).
* The new 875 MHz processors may be mixed
with PA-8700 version 2.2 and 2.3
processors on the same cell board or
within the same partition.
* Boot over Ultra160 SCSI is now supported.
* Support for 256 MB DIMMs has been added.
* The ability to update cell board FPGAs
(LPM, Bob and Drain) using the Onboard
Scan Programming (OSP) Utility has been
added.
* Three new MP command menu commands have
been added to support Secure Sockets Layer
(SSL) used for web encryption:
- Enable SSL (ES)
- Disable SSL (DS)
- SSL Certificate/Keypair Generation (CG).
- Known Problems and Workarounds:
* Use of the Genesis profile command at the
MP command menu returns Invalid Sequence
IDs for complex profile groups B and C.
This result is expected, but users will
note a MP error log entry resulting from
this command.
* Use of the OSP help command at the MP
prompt (MP> he osp) returns Unrecognized
command or topic message.
This problem is solely a help menu issue,
and no degradation exists in the OSP
Utility functionality.
* Chassis code errors may appear in the MP
error log. This problem may manifest itself
as an orphan chassis code
(ORPHAN_CC log entry) or an error level
13 entry.
* Several help commands available at the
Boot Console Handler (BCH) Information Menu
are not visible from within the Information
Menu. The help commands for the following
menu items are affected:
- ChipRevision
- FwrVersion
- MEmory
- PRocessor
- BOot
- DIsplay
- HElp
- REBOOT
- RECONFIGRESET
- MAin
- Known Problems and Workarounds:
* Use of the Genesis profile command at the
MP command menu returns Invalid Sequence
IDs for complex profile groups B and C.
This result is expected, but users will
note a MP error log entry resulting from
this command.
* Use of the Help on OSP command at the MP
prompt returns Unrecognized command or
topic message. However, users can go to
the MP command menu and use the OSP command
without any problems. In fact, this
document provides an example on using the
OSP command to update FPGA firmware.
- Fixes in this Version:
- PDC revision 16.009 contains the following
fixes:
* Numerous minor PDC-to-operating system
interface errors were corrected.
* Typographical and revision errors were
fixed in the Boot Console Handler (BCH)
menu.
* PDC now isolates single bit errors to a
single DIMM during bootstrap to the DIMM.
* CCI FIFO contents are now preserved on a
soft reset.
* The detection of address/parity errors now
results in CPU de-configuration only and
will not halt the entire cell.
* Cells booted to BCH following CPU
replacement no longer stop at Boot is
Blocked (BIB).
* A multi-bit error in a non-root cell no
longer brings down the entire partition.
* Event Monitoring System crossbar hardware
checks no longer result in system hangs
that monopolize CPU time.
* Diskless boot (Ignite) failures were
reduced by creating a larger buffer to
compensate for lost packet information.
* A copy instruction has been changed to
allow hard coding of the d-cache tag clock
settings.
- SINC revision 2.002 contains the following
fix:
* A PDC hardware semaphore request was
removed to minimize boot delays while
reading chassis codes.
- MP revision 3.005 contains the following
fixes:
* The web console now works with Netscape
6.x and Internet Explorer 5.x. It will
not work with Netscape 4.x or Netscape
5.x.
* The MP 'Console Log' option now returns
the correct partition log information.
* Command menu help now includes listings
for Disable Web Console Access (DW) and
Enable Web Console Access (EW).
* The occurrence of Ethernet Media Access
Controller (MAC) hangs due to slow LAN
traffic has been minimized by changes
made to the LAN driver.
* All FPGA and firmware revision numbers
are now reported to three digits
(MP 3.005 versus the previous MP 3.05,
for example).
- SINC revision 2.002 contains the following
fix:
* A PDC hardware semaphore request was
removed to minimize boot delays while
reading chassis codes.
- MP revision 3.005 contains the following
fixes:
* The web console now works with Netscape
6.x and Internet Explorer 5.x. It will
not work with Netscape 4.x or Netscape 5.x.
* The MP 'Console Log' option now returns the
correct partition log information.
* Command menu help now includes listings for
Disable Web Console Access (DW) and Enable
Web Console Access (EW).
* The occurrence of ethernet Media Access
Controller (MAC) hangs due to slow LAN
traffic has been minimized by changes made
to the LAN driver.
* All FPGA and firmware revision numbers are
now reported to three digits (MP 3.005
versus the previous MP 3.05).
PF_CRP74100400:
Announcements, supported features, fixes and
enhancements appear in Version 4.0.
- Announcements:
This release notice provides information and
installation/upgrade instructions for the
rp7410 Firmware Release Version 4.0.
-The following new features have been added:
* Support for the 875 MHz PA-8700 version
3.0 and 3.1 processors has been added.
* Multi-partition machines may now operate
each partition at different CPU
frequencies (e.g., Partition 0 operating
at 750 MHz with Partition 1 operating at
875 MHz).
* The new 875 MHz processors may be mixed
with PA-8700 version 2.2 and 2.3
processors on the same cell board or
within the same partition.
* Boot over Ultra160 SCSI is now supported.
* Three new MP command menu commands have
been added to support Secure Sockets Layer
(SSL) used for web encryption:
- Enable SSL (ES)
- Disable SSL (DS)
- SSL Certificate/Keypair Generation (CG).
- Known Problems and Workarounds:
* Use of the Genesis profile command at the
MP command menu returns Invalid Sequence
IDs for complex profile groups B and C.
This result is expected, but users will
note a MP error log entry resulting from
this command.
* Use of the Help on OSP command at the MP
prompt returns Unrecognized command or
topic message. However, users can go to
the MP command menu and use the OSP command
without any problems. In fact, this
document provides an example on using the
OSP command to update FPGA firmware.
- Fixes in this Version:
- PDC revision 16.009 contains the following
fixes:
* Numerous minor PDC-to-operating system
interface errors were corrected.
* Typographical and revision errors were
fixed in the Boot Console Handler (BCH)
menu.
* PDC now isolates single bit errors to a
single DIMM during bootstrap to the DIMM.
* CCI FIFO contents are now preserved on a
soft reset.
* The detection of address/parity errors now
results in CPU deconfiguration only and
will not halt the entire cell.
* Cells booted to BCH following CPU
replacement no longer stop at Boot is
Blocked (BIB).
* A multi-bit error in a non-root cell no
longer brings down the entire partition.
* Event Monitoring System crossbar hardware
checks no longer result in system hangs
that monopolize CPU time.
- SINC revision 2.002 contains the following
fix:
* A PDC hardware semaphore request was
removed to minimize boot delays while
reading chassis codes.
- MP revision 3.005 contains the following
fixes:
* The web console now works with Netscape
6.x and Internet Explorer 5.x. It will
not work with Netscape 4.x or Netscape 5.x.
* The MP 'Console Log' option now returns the
correct partition log information.
* Command menu help now includes listings for
Disable Web Console Access (DW) and Enable
Web Console Access (EW).
* The occurrence of ethernet Media Access
Controller (MAC) hangs due to slow LAN
traffic has been minimized by changes made
to the LAN driver.
* All FPGA and firmware revision numbers are
now reported to three digits (MP 3.005
versus the previous MP 3.05).
PF_CRP74100301:
Announcements, supported features, fixes and
enhancements appear in Version 3.01.
- Announcements:
* This release notice provides information and
installation/upgrade instructions for the rp7410
Firmware Release Version 3.01. The only time that
the rp7410 might require a firmware upgrade to
Version 3.01 is if either a cell board or a PCI
backplane is being replaced. The rp7410 system
uses the same cell board and the same PCI
backplane as the rp8400 system. The rp8400 system
started shipping in Nov. 2001 with Firmware
Release Version 1.0. Then there was a second
release, Version 2.0. And now, a third release,
Version 3.01. This is the version of firmware
that rp7410 will ship with.
- The following features are supported by rp7410:
* Support for the 650 MHz (version 2.3) and
750 MHz (version 2.3) PA-8700 processors.
* The Onboard Scan Programming Utility allows
firmware to be updated on system FPGA's,
such as PCI backplane LPM.
- Fixes in this Version:
- PDC revision 15.007 contains the following
fixes:
* Code syntax was changed to correct a SCSI
Parameters error. Address fields are now
validated before being accessed.
* Corrections were made to accurately display
all of the characters for Cell board and CPU
state errors found at the Boot Console
Handler (BCH) display.
* A CPU rendezvous execution was modified to
copy buffer information from scratch RAM to
memory. This fixes a loss of memory coherency
leading to a ERR_HPMC_FALL_THROUGH_RESET
chassis code and subsequent HPMC.
* A procedure call used to add a de-configured
CPU was modified. De-configured CPUs may now
be added to a cell without halting on
subsequent reboot.
* The Instruction Cache single bit error LPMC
handler was fixed to properly clear the
icache tag arrays and prevent illegal
instruction traps.
* Disabled SERR assertion by cards with bridges
such as Procurium and 5838 Combo Card, when
the discard timer expires.
- SINC revision 2.00 contains the following fix:
* A time delay was added to fix incorrect
reporting of VRM Status and VRM Fault when
using the PS command.
- MP revision 2.08 contains the following fixes:
* Improvements have been made to the following
commands: PS, VFP, HE, FW, SO.
* Encryption has been added to the web
interface when using Embedded Secure Web
Console (ESWC). This interface works better
with Internet Explorer than with Netscape.
* Changes have been made to increase the
console input data rate and to prevent
dropped characters.
* The login message now uses the system name
rp7410. The model string (seen by HP-UX) has
not changed - it is still 9000/800/rp7410.
* The FPGA firmware update function is now
operational.
* This function allows the following FPGAs
to be programmed over the network:
*System Backplane Global Power Monitor (GPM)
*System Backplane Fan Monitor (FM)
*System Backplane Onboard Scan
Programming (OSP)
*PCI Backplane Local Power Monitor (LPM)
*PCI Backplane Hot Swap (HS)
*Core IO
*Cell Local Power Monitor (LPM)
*Cell BOB
*Cell Drain
*Cell Sinc
*Cell Processor Dependent Code (PDC)
* The "PCI Domain" name has been changed to
"IO Chassis" in the status displays under
the PS command.
* Complex profile and IP address information is
now retained following Core IO replacement.
Defect Description:
PF_CKEYMAT0500:
- See Symptoms.
PF_CKEYMAT0411:
- See Symptoms.
PF_CRP84000400:
- See Symptoms.
PF_CRP74100400:
- See Symptoms.
PF_CRP74100301:
- See Symptoms.
Enhancement:
Yes
PF_CKEYMAT0500:
* Support for 2 GB DIMMs has been added.
PF_CKEYMAT0411:
* Support for the 750 & 875 MHz PA-8700
version 3.0 and 3.1 processors has been
added.
* The state of the cells is now verified
before updating the PDC. If a cell is
running the OS, a Warning is now given.
* Multi-partition machines may now operate
each partition at different CPU
frequencies (e.g., Partition 0 operating
at 750 MHz with Partition 1 operating at
875 MHz).
* The new 875 MHz processors may be mixed
with PA-8700 version 2.2 and 2.3
processors on the same cell board or
within the same partition.
* Boot over Ultra160 SCSI is now supported.
* Support for 256 MB DIMMs has been added.
* The ability to update cell board FPGAs
(LPM, Bob and Drain) using the Onboard
Scan Programming (OSP) Utility has been
added.
* Three new MP command menu commands have
been added to support Secure Sockets Layer
(SSL) used for web encryption:
- Enable SSL (ES)
- Disable SSL (DS)
- SSL Certificate/Keypair Generation (CG).
PF_CRP84000400:
* Support for the 875 MHz PA-8700 version
3.0 and 3.1 processors has been added.
* Multi-partition machines may now operate
each partition at different CPU
frequencies (e.g., Partition 0 operating
at 750 MHz with Partition 1 operating at
875 MHz).
* The new 875 MHz processors may be mixed
with PA-8700 version 2.2 and 2.3
processors on the same cell board or
within the same partition.
* Boot over Ultra160 SCSI is now supported.
* Support for 256 MB DIMMs has been added.
* The ability to update cell board FPGAs
(LPM, Bob and Drain) using the Onboard
Scan Programming (OSP) Utility has been
added.
* Three new MP command menu commands have
been added to support Secure Sockets Layer
(SSL) used for web encryption:
- Enable SSL (ES)
- Disable SSL (DS)
- SSL Certificate/Keypair Generation (CG).
PF_CRP74100400:
* Multi-partition machines may now operate
each partition at different CPU
frequencies (e.g., Partition 0 operating
at 750 MHz with Partition 1 operating at
875 MHz).
* The new 875 MHz processors may be mixed
with PA-8700 version 2.2 and 2.3
processors on the same cell board or
within the same partition.
* Boot over Ultra160 SCSI is now supported.
* Three new MP command menu commands have
been added to support Secure Sockets Layer
(SSL) used for web encryption:
- Enable SSL (ES)
- Disable SSL (DS)
- SSL Certificate/Keypair Generation (CG).
PF_CRP74100301:
* The Onboard Scan Programming Utility allows
firmware to be updated on system FPGA's,
such as PCI backplane LPM.
SR:
0000000000
Patch Files:
celllpm202b.bin
cio101b.bin
fm102a.bin
gpm102.bin
matcio101c.bin
mp409.bin
osp101.bin
pcilpm102.bin
pdc16.011.bin
rp8400_rp7410_V5_0_ReleaseNotes.pdf
sinc202.bin
cksum(1) Output:
2438641994 311594 celllpm202b.bin
2973335795 310152 cio101b.bin
137652359 172005 fm102a.bin
4172902729 225145 gpm102.bin
4216606015 310882 matcio101c.bin
3469303240 8388628 mp409.bin
940318655 164071 osp101.bin
147466829 225145 pcilpm102.bin
1397118046 4194324 pdc16.011.bin
3034626718 871943 rp8400_rp7410_V5_0_ReleaseNotes.pdf
1070528429 65550 sinc202.bin
Patch Conflicts: None
Patch Dependencies: None
Hardware Dependencies:
9000/800/rp8400
9000/800/rp7410
Other Dependencies: None
Supersedes:
PF_CKEYMAT0411 PF_CRP84000400 PF_CRP74100400 PF_CRP74100301
Equivalent Patches: None
Patch Package Size: 3238 KBytes
Installation Instructions:
PLEASE BACK UP YOUR SYSTEM BEFORE YOU INSTALL
THE PATCH.
NOTE: It is important that you read and
understand these instructions completely
before you begin. This can determine your
success in completing the Firmware update.
Be sure to prepare the system and its users before
beginning a Firmware update. Since the system will not be
in operation during the update, you should notify
appropriate managers and users as to when it will be
shutdown and for how long it will remain inoperative.
* * * * * * UNPACKING THE FIRMWARE on HPUX* * * * * * * * * *
1. FTP "PF_CKEYMAT0500.tar.gz" to an empty directory.
2. Use the "gunzip" to unzip the patch.
# gunzip PF_CKEYMAT0500.tar.gz
3. Use the "tar" command to extract the firmware files
# tar -xvf PF_CKEYMAT0500.tar
x celllpm202b.bin, 311594 bytes, 609 tape blocks
x cio101b.bin, 310152 bytes, 606 tape blocks
x fm102a.bin, 172005 bytes, 336 tape blocks
x gpm102.bin, 225145 bytes, 440 tape blocks
x matcio101c.bin, 310882 bytes, 608 tape blocks
x mp409.bin, 8388628 bytes, 16385 tape blocks
x osp101.bin, 164071 bytes, 321 tape blocks
x pcilpm102.bin, 225145 bytes, 440 tape blocks
x pdc16.011.bin, 4194324 bytes, 8193 tape blocks
x rp8400_rp7410_V5_0_ReleaseNotes.pdf, 871943 bytes, 1704 tape blocks
x sinc202.bin, 65550 bytes, 129 tape blocks
4. Verify the checksum of the files using the "cksum" command.
Results of the command should be equal to:
# cksum *
1700902107 15257600 PF_CKEYMAT0500.tar
2438641994 311594 celllpm202b.bin
2973335795 310152 cio101b.bin
137652359 172005 fm102a.bin
4172902729 225145 gpm102.bin
4216606015 310882 matcio101c.bin
3469303240 8388628 mp409.bin
940318655 164071 osp101.bin
147466829 225145 pcilpm102.bin
1397118046 4194324 pdc16.011.bin
3034626718 871943 rp8400_rp7410_V5_0_ReleaseNotes.pdf
1070528429 65550 sinc202.bin
* * * * * * UNPACKING THE FIRMWARE on Microsoft Windows* * * * * *
Note: The Microsoft Windows environment which the file is
going to be downloaded to must have WinZip installed
to un-package the "PF_CKEYMAT0500.tar.gz" file.
1. FTP "PF_CKEYMAT0500.tar.gz" to an empty folder.
2. Open a Windows Explorer window, select the drive and folder
that "PF_CKEYMAT0500.tar.gz" was downloaded to.
3. Double click on the "PF_CKEYMAT0500.tar.gz" file to start
the WinZip program and the un-packaging process.
4. The WinZip program will ask if the file should be decompressed
into a empty folder and opened. Click on the "Yes" button.
5. The WinZip program will start and display all of the files within
the "PF_CKEYMAT0500.tar.gz" file. Click on the "Extract" button.
6. WinZip will ask for the folder that the files should be extracted
to. Using the Folders/Drives portion of the window select the
desired location to extract the files to and the click on the
"Extract" button.
7. Using the Windows Explorer window, select the drive and folder
that "PF_CKEYMAT0500.tar.gz" was downloaded to.
8. Individually "Right Click" on each of the files located in this
folder and select "Properties". Verify that each file has the
following properties:
celllpm202b.bin
Size: 311,594 bytes
Wednesday, March 19, 2003, 11:03:48 PM
cio101b.bin
Size: 310,152 bytes
Wednesday, March 19, 2003, 11:03:48 PM
fm102a.bin
Size: 172,005 bytes
Wednesday, March 19, 2003, 11:03:48 PM
gpm102.bin
Size: 225,145 bytes
Wednesday, March 19, 2003, 11:03:48 PM
matcio101c.bin
Size: 310,882 bytes
Wednesday, March 19, 2003, 11:03:52 PM
mp409.bin
Size: 8,388,628 bytes
Wednesday, March 19, 2003, 11:04:00 PM
osp101.bin
Size: 164,071 bytes
Wednesday, March 19, 2003, 11:04:00 PM
pcilpm102.bin
Size: 225,145 bytes
Wednesday, March 19, 2003, 11:04:00 PM
pdc16.011.bin
Size: 4,194,324 bytes
Wednesday, March 19, 2003, 11:04:04 PM
rp8400_rp7410_V5_0_ReleaseNotes.pdf
Size: 871,943 bytes
Created: Today, March 25, 2003, 11:03:38 PM
sinc202.bin
Size: 65,550 bytes
Created: Wednesday, March 19, 2003, 11:04:06 PM
* * * * * * * * * * * FIRMWARE UPDATE PROCEDURE * * * * * * * * * * *
See rp8400_rp7410_V5_0_ReleaseNotes.pdf for firmware update instructions.
-----End of Document ID: PF_CKEYMAT0500--------------------------------------
Document ID: PF_D36ESHP05
Date Loaded: 20030307
Title: Seagate 36ES HP05 drive firmware
Patch Name: PF_D36ESHP05
Patch Description: Seagate 36ES HP05 drive firmware
Creation Date: 03/03/04
Post Date: 03/03/07
Products:
A6537A Add on 18.2GB 10K RPM Ultra3 SCSI Drive
A6538A Add on 36 GB 10K RPM Ultra3 SCSI Drive
Status: General Release
Critical: No
Category Tags:
defect_repair hardware_enablement enhancement
general_release
Path Name: /firmware_patches/hp/disk/PF_D36ESHP05
Symptoms:
PF_D36ESHP05:
There have been several instances reported by the field of
SCSI bus hangs , NO_HW in ioscans ,and SCSI bus resets
occurring with Seagate Cheetah 36ES disk drives ST318406LC
and ST336706LC. All of the reported cases have been on
DS2100 disk systems with disks at HP03 revision levels.
Defect Description:
PF_D36ESHP05:
There were two defects that could cause I/O timeouts and
resets which could lead to disk hangs and other SCSI bus
errors. These defects were fixed in HP04 and HP05 firmware
and are described below.
1) Change from HP03>HP04.
SMART fly height measurement took a long time
I/O timeouts occurred on some drives because the SMART fly
height measurement took a long time (over 30 seconds, in
some cases). The SMART fly height measurement is
performed as part of the SCSI REZERO command, and as part
of the SMART frame measurements that the drive takes every
two hours. Fundamentally, the SMART fly height measurement
took too long because the firmware did not bother to clear
certain controller error conditions that can happen when
the servo falls off track. Compounding the fundamental
problem, servo-off-track was made much more likely by an
extra read channel modification (during SMART fly height)
that was an accidental carry-over from earlier product
generations. Both problems (controller non-recovery from
off-track, controller irritation of the servo) have been
fixed in HP04 firmware.
2) Change from HP04>HP05
Reselection Timeout -
This is related to unwanted auto-seek during hardware
streaming. Two write commands are merged together
through hardware sequential streaming. The next
command that comes in is a non-merge command. This command
never gets reselected and it times out. In this case, two
writes were merged together and they ended up on the end
of a track. The next write command is a non-merging
command. A disk interrupt is generated when an end of a
track is reached AND then a servo interrupt occurs(for
same reason). But in this failing case, a host interrupt
is generated between the disk and servo interrupt. The host
interrupt was due to the next non-merging write command.
The servo ISR blocks autoseek. This means that we no
longer need autoseek. When the disk interrupt for the next
end of track occurs, the firmware still thinks that
autoseek is in progress and then hangs waiting for a
servo event. So before doing this, it should make sure
that autoseek is terminated or not. The way it can check
that it to look at lsi registers & a global flag-
servo_int_anticipated. This defect has been fixed in
revision HP05
Enhancement:
Yes
PF_D36ESHP05:
SR:
0000000000
Patch Files:
PF_D36ESHP05.frm
cksum(1) Output:
4051632651 3349504 PF_D36ESHP05.frm
Patch Conflicts: None
Patch Dependencies: None
Hardware Dependencies: None
Other Dependencies: None
Supersedes: None
Equivalent Patches: None
Patch Package Size: 3271 KBytes
Installation Instructions:
Please review all instructions and the "Hewlett-Packard
Support Tool License Terms" or your Hewlett-Packard
support terms and conditions for precautions, scope of
license, restrictions, and, limitation of liability
and warranties, before installing this patch.
PLEASE BACK UP YOUR SYSTEM BEFORE YOU INSTALL THE PATCH.
NOTE: It is important that you read and understand
these instructions completely before you
begin. This can determine your success in
completing the Firmware update.
Be sure to prepare the system and its users before
beginning a Firmware update. Since the system will
not be in operation during the update, you should
notify appropriate managers and users as to when it
will be shutdown and for how long it will remain
inoperative.
The Firmware update times will vary depending on the
number of disks installed and the numbers of disks
requiring firmware update. The update should take no
longer than 4 minutes per disk to complete.
* * * * * VERIFYING CURRENT REVISION OF FIRMWARE * * *
VERIFYING FIRMWARE REVISION ON-LINE
The current revision of Firmware can be determined
using the Ioscan and diskinfo commands. The ioscan
command show the device file used in the diskinfo
command. The diskinfo command shows the firmware
revision level.
1. Use the "ioscan" command to find the raw
device files of installed disk drives as
follows:
#ioscan -fnC disk
Class I H/W Path Driver S/W State H/W Type Description
===========================================================
disk 9 0/3/0/1.1.0 sdisk CLAIMED DEVICE HP 36.4GST336706LC
/dev/dsk/c5t1d0 /dev/rdsk/c5t1d0
^^^^^^^^^^^^^^^^
raw device file
2. Use the "diskinfo" command to determine the firmware
revision level of the installed disks as follows:
# diskinfo -v /dev/rdsk/c5t1d0
SCSI describe of /dev/rdsk/c5t1d0:
vendor: HP 36.4G
product id: ST336706LC
type: direct access
size: 35566480 Kbytes
bytes per sector: 512
rev level: HP03 <<<<<<<Firmware revision
blocks per disk: 71132960
ISO version: 0
ECMA version: 0
ANSI version: 2
removable media: no
response format: 2
* * * * * CREATING THE FIRMWARE UPDATE TAPE * * * *
1. Unpack the PF_D36ESHP05.frm file from the PF_D36ESHP05.tar.gz
file using the commands:
gunzip PF_D36ESHP05.tar.gz
tar xvf PF_D36ESHP05.tar
2. Verify the checksum of the file " PF_D36ESHP05.frm"
using the "cksum" command. Results of the command
should be equal to
4051632651 3349504 PF_D36ESHP05.frm
.
Use the "dd" command to copy the file
" PF_D36ESHP05.frm " to tape, as follows:
dd if= PF_D36ESHP05.frm of=/dev/rmt/0m bs=2k
^^^^^^^^^^^
(device path dependent)
NOTE: Blocksize (bs) must be 2k
* * * * *CREATING AN BOOTP SERVER TO UPDATE FROM * * * *
Preparation
Determine or acquire the following:
a.Server to update hostname.
b.Server to update IP address.
c.Server to update LAN Address.
d.Server that is available for use as a
tftp and bootp server for the server
to update blade.
Setup for Diagnostic Load
NOTE: In this example we are using the following values
for the server to update:
Hostname = hpserver
IP = 15.8.134.110
LAN Address = 00306e0cfa60
1.Locate a server on the same network subnet that can
serve as a tftp and bootp server.
2. Log on to the tftp/bootp server as root and enable
tftp and bootp access using SAM or vi:
Using SAM:
a. Run SAM
#sam
b. Now select "Networking and Communications"
c. Then select M-^SNetwork ServicesM-^T
d. Make sure that TFTP is enabled. If not, enable
it now.
e. Make sure that bootp is enabled. If not, enable
it now.
f. Exit sam
Using vi:
#vi /etc/passwd
Add the following line to /etc/passwd:
tftp:*:510:1:Trivial FTP user:/home/tftpdir:/usr/bin/false
3. Create the directory that is to be used for the
tftp/bootp process. In this example the directory will be
/home/tftpdir:
#cd /
#mkdir /home/tftpdir
#cd /home/tftpdir
NOTE: The tftp directory is the directory specified in
the /etc/passwd file. If you want to change the
default directory, you must edit the /etc/passwd
file and replace the "/home/tftpdir" entry
with the directory you want to use.
4. Copy or ftp the firmware LIF file into the tftp
directory just created:
if you are ftp-ing the LIF file:
# ftp <name or IP of ftp server>
login: anonymous
password: anonymous
ftp> lcd /home/tftpdir
ftp> cd <directory with LIF file>
ftp> bin
ftp> mget PF_D36ESHP05.frm
ftp> quit
5. Now on the tftp/bootp server edit the file
/etc/bootptab and enter information about the server
that will be booting the Offline Diagnostic Environment
(ODE):
Add an entry in the /etc/bootptab file in the format of:
hpserver:\
ht=ether:\
ha=00306e0cfa60:\
ip=15.8.134.110:\
hn:\
hd=/:\
bf= PF_D36ESHP05.frm
* * * * * * * * * FIRMWARE UPDATE PROCEDURE * * * * * * *
NOTE: ENSURE YOUR SYSTEM IS FULLY BACKED UP BEFORE
PROCEEDING WITH THIS FIRMWARE UPDATE.
1. Log on as root, and enter "shutdown -r". This command
will shutdown the Operating System and reboot the
system.
2. If AUTOBOOT is on, you will receive the following
message:
"Process is starting autoboot process
To discontinue, press any key within 10 seconds."
PRESS ANY KEY WITHIN 10 SECONDS TO INTERRUPT
THE BOOTING PROCESS.
Boot from the device that contains the firmware update
tape you just created.
4. Interact with IPL (Y or N)?>
Respond with "N".
A series of messages will be displayed. No user
intervention is required. Proceed to step 5.
CAUTION: FROM THIS POINT FORWARD, THIS PROCESS CANNOT
BE INTERRUPTED UNTIL COMPLETION. INTERRUPTING
THIS PROCESS WILL CAUSE YOUR SYSTEM TO BE
INOPERATIVE.
5. The update will take several minutes to complete.
The following is a shortened sample output:
Boot IO Dependent Code (IODC) revision 2
HARD Booted.
ISL Revision A.00.43 Apr 12, 2000
ISL booting ode dfdutilu; exit all; reset
**********************************************************
****** *****
****** Offline Diagnostic Environment *****
****** *****
****** (C) Copyright Hewlett-Packard Co 1993-2001 *****
****** All Rights Reserved *****
****** *****
****** HP shall not be liable for any damages *****
****** resulting from the use of this program. *****
***** *****
***** *****
****** TC Version A.02.23 *****
****** SysLib Version A.00.75 *****
****** Loader Version A.00.62 *****
****** Mapfile Version A.01.41 *****
****** *****
**********************************************************
Type HELP for command information.
ISL_CMD> dfdutilu; exit all; reset
**********************************************************
**** *****
****Disk Firmware Download Utility for User(DFDUTILU)*****
**** *****
****Copyright (C) 1994 - 2002 by Hewlett-Packard *****
**** Company *****
**** Version UB.02.15X2 (03/18/2002) *****
**** *****
**********************************************************
Type HELP for command information.
Please wait while I scan the device busses...
....
**********************************************************
* HP Supported Disks Found *
**********************************************************
Indx Path Product ID Bus Size Rev
--- ------------ ------------------- ------ ------ ------
0 0/3/0/1.1.0 HP 36.4GST336706LC SCSI 36.4GB HP03
Legend:
Indx = Index number used for referencing the device
Rev = Firmware Revision of the device
Note: Due to different calculation methods used, the size
of the device shown is only a rough approximation.
Please wait while I search for all the firmware files....
Note: This may take a while if you are booting from tape.
.......
**********************************************************
* Firmware Files Found (not disks) *
**********************************************************
File name Intended Product ID Rev. Size
---------- ------------------------------ ------ --------
ST318406L5 HP 18.2GST318406LC disk drive HP05 601152
ST336706L5 HP 36.4GST336706LC disk drive HP05 601152
Legend:
File name = name of the firmware file
Intended Product ID = firmware file's intended product name
Rev. = firmware Revision of the firmware file
Size = exact byte size of the firmware image
Updating for the firmware ST336706LC ...
.
.
.
Update Finished. Run dispmap to check the result.
**********************************************************
*Do not forget to power cycle (switch off & on) the disk *
*drives you just downloaded firmware to. Power off the *
*computer if the disk is internal. This will ensure the *
*use of the new firmware downloaded. *
**********************************************************
Now REBOOTing the system....
6. The system will automatically reset, and the boot
process will begin. Boot your system with your
normal process.
If there is a problem with your system after the
update, please contact your service representative.
* * * * * * * * * * END OF UPDATE * * * * * * * * * * *
-----End of Document ID: PF_D36ESHP05----------------------------------------
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]