Ce 6 Cookbook

CE cookbook
HP-UX
October 2004
Version 6
http://eclipse.bel.hp.com:8888/users/qualification/
Thanks to the persons who made it possible to make this cookbook.
The maker of this cookbook
is not responsible
for the mistakes or the system problems
caused by the use of this cookbook.
Table of contents
Table of contents
Table of contents ....................................................................................2
Chapter 1: LVM and disk replacement ..................................................6

1 Gathering information about your LVM configuration. ................................................................6
2 Main troubleshooting flowchart. ..............................................................................................7
3 Replacing FC disk + FC-controllers ...........................................................................................8
4 Hot-Swap Procedure for Attached Physical Volumes ...................................................................9
5 Hot-Swap Procedure for Unattached Physical Volumes .............................................................11
6 LIF/BDRA Configuration Procedure ........................................................................................12
7 Itanium mirror bootdisk: Procedure for Attached Physical Volumes .............................................13
8 Itanium mirror bootdisk: Procedure for UnAttached Physical Volumes .........................................15
9 Removing a Ghost Disk Using the PV key................................................................................17
10 Removing a PV using its PV key ..........................................................................................18
11 Installation of a mirrored LVM root disk ...............................................................................19
12 Recreating volume group information. .................................................................................20
13 Moving a root disk to to a new disk or another disk..............................................................21
14 Disk recovery using vgexport and vgimport..........................................................................22
15 Splitting mirrors to perform backups....................................................................................23
16 Tech Info Bits....................................................................................................................24
17 Moving an existing root disk to a new hardware path. .........................................................26
Chapter 2: Storage.............................................................................27
1 Overview of most common commands for diskarrays ...............................................................27
ID drive/FRU –........................................................................................................................33
Include - auto ..........................................................................................................................33
Include - manual......................................................................................................................33
2 Special commands for VA 7XY0 ............................................................................................34
3 Argument for diskarray commands.........................................................................................37
4 Autoraid .............................................................................................................................38
5 DS2100 .............................................................................................................................39
6 DS2300 .............................................................................................................................40
7 DS2405 .............................................................................................................................42
8 FC60..................................................................................................................................44
9 SC10 .................................................................................................................................46
10 FC10 ..............................................................................................................................47
2
Table of contents
11 Nike array.......................................................................................................................49
12 VA7xy0...........................................................................................................................51
VA Troubleshooting Do’s and Don’ts .........................................................................................51
VA7100 ................................................................................................................................52
VA7110 – VA7400 ................................................................................................................54
VA7410 ................................................................................................................................55
Chapter 3: Patches .............................................................................56

1 Install..................................................................................................................................56
2 List the contents of a depot ....................................................................................................56
3 Copy a depot ......................................................................................................................56
4 Registration of a depot .........................................................................................................56
5 Create a patch depot from multiple individual patches .............................................................56
6 Patch Clean-up ....................................................................................................................57
Chapter 4: Servers .............................................................................58

1 K-Class rear view and addressing ..........................................................................................59
2 A L N general......................................................................................................................61
Using the 'GSP' on A, L, & N Class...........................................................................................61
SL logs troubleshooting ............................................................................................................61
A, L, and N Class front panel LED .............................................................................................62
3 A 400/500 (RP24xx)...........................................................................................................63
A400/500 Rear view..............................................................................................................63
A400/500 internals................................................................................................................63
A400/A500 IO-paths .............................................................................................................63
4 StormsPeak - rp3410 / rp3440 -- LongsPeak - rx2600 .....................................................64
5 L-Class Rear View and Addressing (RP54xx) ...........................................................................67
6 Mt. Hamilton - rp4440 -- Mt. Diablo - rx4640 .................................................................70
7 N-Class Rear View and Addressing (RP7400) .........................................................................72
8 Matterhorn/keystone MP, BCH commands..............................................................................75
MP ........................................................................................................................................75
BCH ......................................................................................................................................76
9 Matterhorn - rp74X0 -- Eiger – rx7620 ...........................................................................77
10 Keystone - rp8400 -- Olympia - rx8620.......................................................................80
11 V-CLASS TROUBLESHOOTING ..........................................................................................85
12 Superdome ......................................................................................................................87
Chapter 5: Procedures........................................................................91
1 SCSI bus error troubleshooting and log decoding ....................................................................91
2 Interpreting the FC hardware path .........................................................................................92
Types used for FC-addressing ...................................................................................................94
PDA: Peripheral Device Addressing........................................................................................94
LUA: Logical Unit Addressing ................................................................................................94
3
Table of contents
VSA: Volume Set Addressing.................................................................................................94

3 Fibre card troubleshooting ....................................................................................................95
4 Exercise a disk ....................................................................................................................96
5 Boot machine to single user or maintenance mode...................................................................97
6 EMS errors ..........................................................................................................................98
7 Checking for HPMC’s ...........................................................................................................99
8 Using MSTM diagnostics ....................................................................................................100
9 MSTM in single user...........................................................................................................101
10 Stopping mstm if wrong terminal-emulation ........................................................................101
11 Updating firmware from ECU cdrom .................................................................................102
PDC upgrade........................................................................................................................102
Updating Autoraid Controller Firmware ...................................................................................102
Updating GSP Firmware Flowtchart.........................................................................................103
Console path for the A/L/N ...................................................................................................103
Updating GSP Firmware A-revision (Tosca A6696A).................................................................104
Updating GSP Firmware B/C-revision (Maestro) .......................................................................105
12 SS_CONFIG from Tape or ECU cdrom..............................................................................106
13 Diagnostic (STM) install ...................................................................................................106
14 Tar to or untar from a tape...............................................................................................106
15 Service guard.................................................................................................................107
16 Mount a cd ....................................................................................................................108
17 Changing instance numbers.............................................................................................108
18 Cisco routers ..................................................................................................................109
19 DTC16 ..........................................................................................................................110
Normal startup......................................................................................................................110
Errors...................................................................................................................................110
Enter the exchange DTC16 machine type.................................................................................110
Change the DTC16 LAN address ............................................................................................110
20 4 Channel ULTRA 2 SCSI RAID 4SI...................................................................................111
Information ...........................................................................................................................111
Configuration .......................................................................................................................111
KNOW ISSUE ......................................................................................................................111
21 Optical Boxes With Numbered Config and Test Choices.....................................................112
22 ICOD ............................................................................................................................113
iCOD Commands..................................................................................................................113
iCOD Install ( you must use the DEC. 2000 or later Support CD ) ..............................................113
23 Partitioning ....................................................................................................................114
vPar Commands....................................................................................................................114
Commands at MON> prompt .............................................................................................114
Commands at HP-UX prompt ...............................................................................................115
Changing a bootdisk in a vPar ...............................................................................................115
nPar Commands (Hardpartitioning) .........................................................................................116
4
Table of contents
Chapter 6: HPUX Command Summary for CE’s .................................117

1 General HPUX Commands ..................................................................................................117
Important files: ......................................................................................................................118
How to Change your terminal emulation ..................................................................................118
HPUX Boot Commands ..........................................................................................................119
2 LVM Commands ................................................................................................................119
3 LVM FILES and DIRECTORIES ..............................................................................................120
4 Networking Commands ......................................................................................................121
Important files: ......................................................................................................................121
5 Software/Patches...............................................................................................................121
Important files: ......................................................................................................................121
6 FTP commands...................................................................................................................122
7 Vi Commands....................................................................................................................122
5
LVM and disk replacement
Chapter 1: LVM and disk replacement
1 Gathering information about your LVM configuration.
Verify that you have a defective disk that needs to be replaced. Use sysdiag or mstm to diagnose.
Depending on the disk failure, you may need to collect all or some of the following information using a
combination of the various data gathering commands listed here below.
The hardware path and special device file.

# ioscan –fnC disk
Bit out the special device file to pin point the disk hardware path.
# lssf /dev/dsk/cXtXd0
The disktype, example. C2490WD, ST32550W

# diskinfo /dev/rdsk/cXtXd0
The volume group the disk belongs to.

# strings /etc/lvmtab
The logical volumes residing on the volumegroup to which the disk belongs.
# vgdisplay -v /dev/vgXX
Get the map of the logical extends that correspond to the physical extends of each volume.
# pvdisplay -v /dev/dsk/cXtXd0
Get the characteristics and status of each logical volume in the faulty volume group.
# lvdisplay -v /dev/vgXX/lvolX
Get the BDRA information from the boot disk. (Boot Data Reserved Data)
# lvlnboot –v
If the boot disk is mirrored, record the hardware path and the software path of the mirror copy, from the
command above.
Also, check if the mirror has offline diagnostics.

# lifls -il /dev/rdsk/cXtXd0 (/dev/rdsk/cXtXd0 is the path of the mirror disk)
Check how the boot disk booted in normal way.

# lifcp /dev/rdsk/cXtXd0:AUTO - (/dev/rdsk/cXtXd0 is the path of the bootdisk)
hpux
The way to make the disk bootable
# mkboot -a "hpux-lq (;0)/stand/vmunix" /dev/rdsk/cXtXd0
The italic text is the output of the lifcp command
Use the dd command to identify a disk without crawling all over the floor. When you issue the command,
look for the solid blinking green LED. Ctrl+C will stop the read. Be sure disk is CLAIMED.
Note: If the disk is dead this command will not work. Also, dd is a good on line test. If the disk hangs or
if you get an I/O error, then the disk is defective.
# dd if=/dev/rdsk/cXtXd0 of=/dev/null bs=1024k
6
2 Main troubleshooting flowchart.
!!!! When you have to change a FC disk or EFI boot disk read section 3.
Information from Software Recovery Handbook, chapter 16 LVM. http://bali.grc.hp.com/scunix/sw_recovery/
7
3 Replacing FC disk + FC-controllers
The same procedure as for a normal SCSI disk. See flowchart in section 2
BUT When you replace a FC disk or FC controller please be aware of…
THIS IS NEEDED IN CASE IOSCAN GIVES NO_HW.
The ‘World-wide name’ (unique identifier) for device at loop ID xx has changed when you replace a disk. If the device has been replaced intentionally, please use the fcmsutil
replace_dsk command to allow the new device to be used.
The main reason why this command was introduced is to avoid data corruption due to a user accidentally plugging in another device at the same nport_id.
When the system (initiator) establishes a login session (PLOGI) with the device (the target), it authenticates the device.
This authentication is to make sure that it is talking to the same device (at the nport_id).
Once the replace_dsk command is run, there will be no authentication on the device during the next login but the device will be authenticated on all subsequent logins.
# fcmsutil /dev/tdX replace_dsk <remote_N-port_ID>
To locate the N-port ID of the device you replaced use devstat all. Or you can use its Loop-ID.
1. Take out defective disk, if you know N-port_ID or Loop_ID go to step 3
2. Lookup the N-port_ID or Loop_ID. (also check FC-addressing or storage section)
a. # fcmsutil /dev/tdX devstat all | grep loop
b. Try to login into each device

# fcmsutil /dev/tdX echo –l <Loop_ID from step 2.a>
Do this for each Loop_ID until you get the message Unable to login instead of Data
came back intact.
3. Run replace_dsk using the remote-N-port_ID or Loop_ID

# fcmsutil /dev/tdX replace_dsk <N-port_ID>
-l <Loop_ID>
4. Insert new disk
Note: After running above command, you will get a message indicating the port will not be
authenticated
8
4 Hot-Swap Procedure for Attached Physical Volumes
NOTE: A physical volume is considered to be attached, if the pvdisplay command is able to report a valid
status (unavailable/available) for it. Otherwise it’s called unattached.
In this LVM chapter we only consider the official procedure to be supported. The reason is that
there are potentially serious problems with replacing an attached device. Although the pvdisplay indicates
the device is unavailable, LVM could still be trying to recover it.
The consequences can be data corruption or obscure problems that can be difficult to track down, due to
the LVM metadata on the device being improperly written.
Follow these steps to replace a hot-swap disk module for attached Physical Volumes, which means that the
disk was not defective/unaccessible at the time the volume group was activated.
Hot-swapping a disk which was defective during activation (unattached) requires a different sequence of
commands. Skip to the alternative procedure, 5. Hot-Swap Procedure for Unattached Physical Volumes.
1. Reduce any logical volumes that have mirror copies on the faulty disk so that they no longer mirror
onto that disk.
NOTE: Be advised to check first, what LVs have mirror extents allocated on the faulty disk (to be
checked with pvdisplay –v /dev/dsk/cXtXdX). Then you should check for each found LV how it is
mirrored (use lvdisplay –v /dev/vgXX/lvolX). If the mirror extents span more than one PV then it is
highly recommended to specifiy all PVs with the lvreduce command that are in the “same mirror set
of disks” as the faulty one. Otherwise LVM may pick the “wrong” disks for reduction, leading to
undesired results (e.g. asymmetrical layouts). Take a note of this PV list, since you need this
information later when you re-establish the mirror using lvextend.
# lvreduce -m 0 -A n /dev/vgXX/lvolX <list of PVs> (for 1 way mirroring)

or
# lvreduce -m 1 -A n /dev/vgXX/lvolX <list of PVs> (for 2 way mirroring)
where list of PVs is the the list of devices determined according to the note above. We use the –A n
option to prevent the lvreduce command from performing an automaticic vgcfgbackup operation,
which is likely to get stuck on accessing a defective disk.
2. Replace the faulty disk. Please refer to the appropriate administration guide for
instructions on how to replace the disk.
Do an ioscan on the replaced disk to insure that it is accessible (CLAIMED) and also as a double
check that it is a proper replacement (see note above).
# ioscan -f /dev/dsk/cXtXdX
3. For fibre channel disks perform the replace_dsk steps described in the section 3.
4. Restore the LVM configuration/headers onto the replaced disk from your backup of the LVM
configuration.
# vgcfgrestore -n VG /dev/rdsk/cXtXdX
5. Attach the new disk to the active volume group with the vgchange command.
# vgchange -a y vgXX
or # vgchange -a e vgXX (for exclusively activated Cluster VGs)
9
6. Important: If the disk is the mirror of a root disk, then you must configure the LIF/BDRA
see LIF/BDRA Configuration Procedure
7. Lvextend the mirrors back onto the replaced disk. This may take several minutes as
it will have to copy all the data from the original copy of the data to the mirrored
extents. The logical volume(s) are still accessible to users' applications during this
command.
# lvextend -m 1 /dev/vgXX/lvolX /dev/dsk/cXtXdX & (for 1 way mirroring)
or
# lvextend -m 2 /dev/vgXX/lvolX /dev/dsk/cXtXdX & (for 3 way mirroring)
To check the progress of the synchronization you could use:

# lvdisplay -v $(find /dev/vgXY -type b) | grep stale | wc –l
A shell loop like this could be used to extend a bunch of lvols automatically:
# for lvol in lvol1 lvol2 lvol3 ... (specify any LV you need to mirror)
> do
> lvextend –m 1 /dev/vgXX/$lvol /dev/dsk/cXtXdX
> done
10
5 Hot-Swap Procedure for Unattached Physical Volumes
Follow these steps to replace a hot-swap disk module for unattached physical volumes.
1. Replace the faulty disk. Please refer to the appropriate administration guide for instructions on how
to replace the disk.
Do an ioscan on the replaced disk to insure that it is accessible (CLAIMED) and also as a double
check that it is a proper replacement .
# ioscan -f /dev/dsk/cXtXdX
2. For fibre channel disks perform the replace_dsk steps described in the section 3.
3. Restore the LVM configuration/headers onto the replaced disk from your backup of the LVM
configuration.
# vgcfgrestore -n VG /dev/rdsk/cXtXdX
4. Attach the new disk to the active volume group with the vgchange command.
or
# vgchange -a e vgXX (for exclusively activated Cluster VGs)
5. Important: If the disk is the mirror of a root disk, then you must configure the LIF/BDRA
see LIF/BDRA Configuration Procedure
6. Resynchronize the mirrors of the replaced disk. This may take several minutes as it will have to
copy all the data from the original copy of the data to the mirrored extents. The logical volume(s)
are still accessible to users' applications during this command.
# vgsync vgXX &
To check the progress of the synchronization you could use:

# lvdisplay -v $(find /dev/vgXY -type b) | grep stale | wc -l
11
6 LIF/BDRA Configuration Procedure
This subprocedure installs/updates information on disk that is mandatory for boot support.
Therefore it is referenced from several other parts of this chapter.
1. Write LIF header and LIF files (ISL, AUTO, HPUX, LABEL):
# mkboot -l /dev/rdsk/cXtXdX
# lifls –l /dev/rdsk/cXtXdX (to ckeck it)
2. Write content of AUTO File: (if autoboot is desired)

# mkboot -a hpux /dev/rdsk/cXtXdX (autoboot with qurom enforced)
# mkboot -a ’hpux –lq’ /dev/rdsk/cXtXdX (autoboot without qurom enforced)
# mkboot -a “/stand/vpmon -a” /dev/rdsk/cXtYd0 (autoboot with vPar)
# lifcp /dev/rdsk/cXtXdX:AUTO - (to ckeck it)
NOTE: By default, LVM enforces a quorum of >50% of a VG’s PVs being available at activation
time. If e.g. the root VG contains 2 PVs, then the system rejects to boot unless you disable the
quorum check using the –lq option.
3. Install ODE files (may be skipped):

# cd /usr/sbin/diag/lif
# getconf HW_CPU_SUPP_BITS (the result is either 32, 32/64 or 64)
# mkboot -b updatediaglif -p ISL -p AUTO -p HPUX -p LABEL /dev/rdsk/cXtXdX (if 32 or 32/64)
# mkboot -b updatediaglif2 -p ISL -p AUTO -p HPUX -p LABEL /dev/rdsk/cXtXdX (if 64)
(the -p option preserve the specified file so that it is not overwritten in LIF)
4. Write content of LABEL file, i.e set root, boot, swap and dump device:
NOTE: This step can be omitted if you replace a failed mirror disk. Then this information has
already been restored by vgcfgrestore. To be sure to have the latest information on the disk, just
do the following steps.
# lvlnboot -r /dev/<rootVG>/lvol3
# lvlnboot -b /dev/<rootVG>/lvol1
# lvlnboot -s /dev/<rootVG>/lvol2
# lvlnboot -d /dev/<rootVG>/lvol2
# lvlnboot –v (to ckeck it)
12
7 Itanium mirror bootdisk: Procedure for Attached Physical Volumes
1. Reduce any logical volumes that have mirror copies on the faulty disk so that no longer mirror onto
that disk.
#lvreduce –m 0 –A n /dev/vg00/lvolX /dev/dsk/cXtYdZs2
2. Replace the faulty disk. Verify with an ioscan.

#ioscan –f /dev/dsk/cXtYdZ
3. Verify the disk partitions (cAtBdC is primary disk, cXtYdZ is mirror disk)
At a cold-installed UX 11.23 system the partition sizes are different compared to UX 11.22.
Use idisk to check the partition sizes.
# diskinfo /dev/rdsk/cAtBdCs1 | grep size
Create a partition description file:
at UX 11.22 at UX 11.23 (cold-installed)

# vi /tmp/partitionfile
# vi /tmp/partitionfile
3
2
EFI 500MB
EFI 100MB
HPUX 100%
HPUX 100%
HPSP 400MB
4. Clear the previously partition configuration

# idisk -R /tmp/partitionfile /dev/rdsk/cXtYdZ
5. Create the disk partitions
Use idisk command to partition the disk according to this file:

# idisk -wf /tmp/partitionfile /dev/rdsk/cXtYdZ
6. Create the new device files for the new partitions (cXtYdsZ, s2, (s3))
# insf -e –Cdisk
7. Verify the partitiontable

# idisk /dev/rdsk/cXtYdZ
8. Use efi_fsinit to initialize the FAT filesystem on the EFI partition:
# efi_fsinit -d /dev/rdsk/cXtYdZs1
NOTE: This step is not neccessary if it can be guaranteed that the mirror disk does not contain a
valid EFI filesystem. In this case efi_fsinit will be done automatically by the subsequent mkboot
command. But if you take e.g. an old UX 11.22 boot disk as mirror disk, mkboot will not
automatically run efi_fsinit. As a result only 100MB of the 500MB EFI partition (s1) can be used.
9. Use mkboot to format the EFI partition (s1) and populate it with the EFI files below /usr/lib/efi/
and to format the LIF volume (part of s2) and populate it with the LIF files (ISL, AUTO, HPUX, LABEL)
below /usr/lib/uxbootlf:
13
# mkboot -e -l /dev/rdsk/cXtYdZ
# efi_ls -d /dev/rdsk/cXtYdZs1 (to check EFI)
# lifls -l /dev/rdsk/cXtYdZs2 (to check LIF)
10. Check the content of AUTO file on EFI partition on primary disk:
# efi_cp -d /dev/rdsk/cXtYdZs1 -u /EFI/HPUX/AUTO /tmp/prim

#cat /tmp/prim
NOTE: Specify the -lq option if prefer that your system boots up without interruption in case of a
disk failure:
# mkboot -a "boot vmunix -lq" /dev/rdsk/cXtYdZ
# mkboot -a "boot vmunix -lq" /dev/rdsk/cAtBdC
11. Check the content of AUTO file on EFI partition on new mirror disk:
# efi_cp -d /dev/rdsk/cXtYdZs1 -u /EFI/HPUX/AUTO /tmp/mirror

# cat /tmp/mirror
12. Copy the HP service partition (UX 11.23 only): (skip this, if you don’t have a service partition)
# dd if=/dev/rdsk/cAtBdCs3 of=/dev/rdsk/cXtYdZs3 bs=1024k
13. Now the disk is configured as boot devices.

Now we have to put the hpux info back.
14. Restore the LVM configuration/headers onto the replaced disk from your backup
#vgcfgrestore –n vg00 /dev/rdsk/cXtYdZs2
15. Attach the new disk to the active volume group with the vgchange command
#vgchange –a y vg00
16. Extend the mirrors back onto the replaced disk.

#lvextend –m 1 /dev/vg00/lvolX /dev/dsk/cXtYdZs2
14
8 Itanium mirror bootdisk: Procedure for UnAttached Physical Volumes
1. Replace the faulty disk. Verify with an ioscan.

#ioscan –f /dev/dsk/cXtYdZ
2. Verify the disk partitions (cAtBdC is primary disk, cXtYdZ is mirror disk)
At a cold-installed UX 11.23 system the partition sizes are different compared to UX 11.22.
Use idisk to check the partition sizes.
Create a partition description file:
3. at UX 11.22 4. at UX 11.23 (cold-installed)
6. # vi /tmp/partitionfile
5. # vi /tmp/partitionfile 3
2 EFI 500MB
EFI 100MB HPUX 100%
HPUX 100% HPSP 400MB
7. Clear the previously partition configuration

# idisk -R /tmp/partitionfile /dev/rdsk/cXtYdZ
8. Create the disk partitions
Use idisk command to partition the disk according to this file:

# idisk -wf /tmp/partitionfile /dev/rdsk/cXtYdZ
9. Create the new device files for the new partitions (cXtYdsZ, s2, (s3))
# insf -e –Cdisk
10. Verify the partitiontable

# idisk /dev/rdsk/cXtYdZ
11. Use efi_fsinit to initialize the FAT filesystem on the EFI partition:
# efi_fsinit -d /dev/rdsk/cXtYdZs1
NOTE: This step is not neccessary if it can be guaranteed that the mirror disk does not contain a
valid EFI filesystem. In this case efi_fsinit will be done automatically by the subsequent mkboot
command. But if you take e.g. an old UX 11.22 boot disk as mirror disk, mkboot will not
automatically run efi_fsinit. As a result only 100MB of the 500MB EFI partition (s1) can be used.
12. Use mkboot to format the EFI partition (s1) and populate it with the EFI files below /usr/lib/efi/
and to format the LIF volume (part of s2) and populate it with the LIF files (ISL, AUTO, HPUX, LABEL)
below /usr/lib/uxbootlf:
# mkboot -e -l /dev/rdsk/cXtYdZ
15
# efi_ls -d /dev/rdsk/cXtYdZs1 (to check EFI)

# lifls -l /dev/rdsk/cXtYdZs2 (to check LIF)
13. Check the content of AUTO file on EFI partition on primary disk:
# efi_cp -d /dev/rdsk/cXtYdZs1 -u /EFI/HPUX/AUTO /tmp/prim

#cat /tmp/prim
NOTE: Specify the -lq option if prefer that your system boots up without interruption in case of a
disk failure:
# mkboot -a "boot vmunix -lq" /dev/rdsk/cXtYdZ
# mkboot -a "boot vmunix -lq" /dev/rdsk/cAtBdC
14. Check the content of AUTO file on EFI partition on new mirror disk:
# efi_cp -d /dev/rdsk/cXtYdZs1 -u /EFI/HPUX/AUTO /tmp/mirror

# cat /tmp/mirror
15. Copy the HP service partition (UX 11.23 only): (skip this, if you don’t have a service partition)
# dd if=/dev/rdsk/cAtBdCs3 of=/dev/rdsk/cXtYdZs3 bs=1024k
16. Now the disk is configured as boot devices.

Now we have to put the hpux info back.
17. Restore the LVM configuration/headers onto the replaced disk from your backup
vgcfgrestore –n vg00 /dev/rdsk/cXtYdZs2
18. Attach the new disk to the active volume group with the vgchange command
#vgchange –a y vg00
19. Resyncronize the mirrors of the replaced disk

#vgsync vg00
16
9 Removing a Ghost Disk Using the PV key
You may come into a situation where you have to remove a PV from a VG that has failed or not even
physically connected but still recorded in the lvmtab. Such a PV is sometimes called a “ghost disk” or
“phantom disk”. You can get a ghost disk if the disk has failed before VG activation, maybe because the
system has been rebooted after the failure.
If you cannot use vgcfgrestore to write the original LVM header back to the new disk because a valid LVM
configuration backup file (/etc/lvmconf/vgXX.conf[.old]) is missing or corrupted you have to remove that
PV from the VG (vgreduce) to get a clean configuration.
NOTE: In such situations the vgcfgrestore command may fail to restore the LVM header, complaining
about a ‘Mismatch between the backup file and the running kernel’. If you are 100% sure that your backup
is valid you may override this check using the –R option.
In order to remove a PV from a VG you have to free it first, i.e. remove all logical extents from it. If the LVs
on such a disk is not mirrored data is lost anyway. If it is mirrored you need to reduce the mirror before
removing the PV.
A ghost disk is usually indicated by vgdisplay reporting more current PVs than active ones.
Additionally LVM commands may complain about the missing PVs:
# vgdisplay vg01
vgdisplay: Warning: couldn't query physical volume "/dev/dsk/cXtXdX":
The specified path does not correspond to physical volume attached to this volume group
vgdisplay: Couldn't query the list of physical volumes.
--- Volume groups ---
VG Name /dev/vg01
…….
Cur PV 2 (number of PVs recorded in the lvmtab)
Act PV 1 (number of PVs recorded in the kernel)
…….
Note that the PV cXtXdX is still recorded in lvmtab:

/dev/vg01
/dev/dsk/c0t0d2
/dev/dsk/c1t2d2
/dev/dsk/cXtXdX
Running vgreduce with the -f option would remove all PVs that are “free”, i.e there is no LV having extents
on that PV. Otherwise - if the PV is not free - vgreduce -f reports an extent map to identify the associated
LVs:
# vgreduce -f vg01
skip alternate link /dev/dsk/c1t2d2
vgreduce: Couldn't query physical volume "/dev/dsk/c0t11d0":
The specified path does not correspond to physical volume attached to this volume group
Not all extents are free. i.e. Out of 508 PEs, only 500 are free.
You must free all PEs using lvreduce/lvremove before the PV can be removed.
Example:
# lvreduce -A n -m 0 /dev/vg01/lvol1.
# lvremove -A n /dev/vg01/lvol1.
17
10 Removing a PV using its PV key
The PV key of a disk indicates its order in the VG. The first PV has the key 0, the second has
the key 1, etc. This does not necessarily have to be the order of appearance in lvmtab altough
it is usually like that, at least when a VG is initially created.
The PV key can be used to address a PV that is not attached to the VG. This usually happens
if it was not accessible during activation, e.g. due to a hardware or configuration problem.
NOTE: The PV may be unattached due to some temporary problem during VG activation which is no
longer present. In this case you should try to re-activate the VG to force LVM to re-scan the
devices listed in lvmtab:
or
# vgchange -a e vgXX (for exclusively activated Cluster VGs)
If the problem persists follow these steps to clear the situation:
1. Obtain the PV key using the -k option of lvdisplay:

# lvdisplay –v –k /dev/vg01/lvol1
...
-- Logical extents ---
LE PV1 PE1 Status 1 PV2 PE2 Status 2
00000 0 00000 stale 1 00000 current
00001 0 00001 stale 1 00001 current
Compared to the output above the ??? of the lvdisplay without –k have been replaced with the PV
key (0, 1).
2. Reduce the mirror with the obtained key as argument:

# lvreduce –k –m 0 /dev/vg01/lvol1 0
3. After that the PV can be removed from the VG:

# vgreduce -f vg01
PV with key 0 sucessfully deleted from vg vg01

Repair done, please do the following steps.....:
1. save and remove /etc/lvmtab
# mv /etc/lvmtab /etc/lvmtab.org
2. Use vgscan -v to re-create /etc/lvmtab

# vgscan –v
3. NOW use vgcfgbackup(1m) to save the LVM setup
4. Check the results:

5. Re-activate the VG and backup the LVM config:

# vgchange -a y vg01
# vgcfgbackup vg01
If the LV was not mirrored, re-create the LV (lvcreate), create a FS on it (newfs) and recover
your data from backup.
18
11 Installation of a mirrored LVM root disk
Perform normal disk installation. You do need the mirroring software installed.
1. Initialize the disk to be a boot mirror.

# pvcreate -f -B /dev/rdsk/cXtXd0
2. Extend the volume group.

# vgextend /dev/vg00 /dev/dsk/cXtXd0
3. Place the boot utilities in the boot LIF area.

# mkboot /dev/rdsk/cXtXd0
4. Add an Auto file in the LIF area.

# mkboot -a "hpux (;0)/stand/vmunix" /dev/rdsk/cXtXd0
NOTE:
The -lq switch could be added to overwrite quorum on HIGH AVAILABILITY systems.
# mkboot -a "hpux -lq (;0)/stand/vmunix" /dev/rdsk/cXtXd0
If vPar use # mkboot -a “/stand/vpmon -a” /dev/rdsk/cXtYd0
Switch on mirroring on all the lvols in the root volume group.

Start with lvol1 and move sequentially in order of ascending lvol numbers.
Don't forget about swap!
5. # lvextend -m 1 /dev/vg00/lvol1 /dev/dsk/cXtXd0 (1-way mirroring)

(cXtXd0 = device file of the new disk)
6. Update BDRA and LIF file on vg00

# lvlnboot –Rv
# lvlnboot -v /dev/vg00/lvol1 (if lvol1 is displayed as boot go to step 7, if not then
# lvlnboot -b /dev/vg00/lvol1 lvlnboot -b will define lvol1 as boot again,
# lvlnboot -Rv /dev/vg00/lvol1 lvlnboot -Rv will update BDRA, verify lvol1 is
defined as boot, if not repeat these 3 commands untill
defined)
7. Display the new mirror.

# lvdisplay -v /dev/vg00/lvol* | more
19
12 Recreating volume group information.
The directory in /dev containing vgXX has been accidentally removed.
1. Manually create the directory.

# mkdir /dev/vgXX
2. Make the /dev/vgXX/group file.

# mknod /dev/vgXX/group c 64 0xXX0000
(XX = volume group number, if possible find out what XX used to be)
3. Move the /etc/lvmtab file.

# mv /etc/lvmtab /etc/lvmtab.old
4. Create the new /etc/lvmtab file.

# vgscan –v
Import the information form the disk that was removed by accident.
Note: If the system has not been rebooted since the accident, you should do it now.
5. # vgimport -v /dev/vgXX /dev/dsk/cXtXd0 <and any other disks>
6. Activate the imported volume group.

# vgchange -a y /dev/vgXX
7. Now mount the file systems or just reboot the system.
If you have naming conflicts, remember that logical volumes can be renamed, but volume groups have to
be exported and imported for renaming.
20
13 Moving a root disk to to a new disk or another disk.
We want to move the the root disk /dev/dsk/c0t6d0 to disk /dev/dsk/cYtXd0 staying within the same
volume group.
1. Make the destination disk a bootable LVM disk.

# pvcreate -f -B /dev/rdsk/cYtXd0
2. Make the disk bootable.

# mkboot /dev/rdsk/cYtXd0
# mkboot -a "hpux -a (;0)/stand/vmunix" /dev/rdsk/cYtXd0
3. Extend your root volume group with the destination disk.

# vgextend /dev/vg00 /dev/dsk/cYtXd0
4. Move all physical extends from the source disk to the destination disk.
# pvmove /dev/dsk/cXtXd0 /dev/dsk/cYtXd0
5. Reduce the source disk from the volume group.

# vgreduce /dev/vg00 /dev/dsk/cXtXd0
6. Shutdown and reboot from the new root disk in maintenance mode.
ISL> hpux -lm (;0)/stand/vmunix
7. Update the BDRA and the LABEL file.

# vgchange -a y /dev/vg00
8. (Always use caution using these commands. Failure to do so can create data
loss!)
# lvlnboot -b /dev/vg00/lvol1
# lvlnboot -s /dev/vg00/lvol2
# lvlnboot -r /dev/vg00/lvol3
# lvlnboot -Rv ( On a 10.20 system verify lvol1 is defined as boot.)
# vgchange -a n /dev/vg00
9. Reboot the system and update the stable storage form the ISL prompt.
21
14 Disk recovery using vgexport and vgimport

Disk Relocation
This procedure shows the basic use of vgexport and vgimport for moving a volume group/disks from one
system to an other. The following pages shows examples of using these two commands for recovery of
data disks which belong to volume groups.
1. Unmount all mounted lvols within the volume group you are exporting. Then deactive the volume
group.
# umount /dev/vgXX/lvolX /mount_point_name (repeat for all lvols)
2. Desactivated the volume group

# vgchange -a n /dev/vgXX
3. Note the group file minor number and all lvols.

# cd /dev/vgXX
# ll
4. Export the volume group LVM info to the map file.

# vgexport -v -m /tmp/vgXX_map /dev/vgXX
5. Create the volume group and the group file on the other system.
# mkdir /dev/vgXX
# cd /dev/vgXX
# mknod /dev/vgXX/group c 64 0xXX0000 (XX = volume group number, info from step
3)
6. Import the new disk or disks into the other system. Note the possible new special device file which
might be different.
# vgimport -v -m /tmp/vgXX_map /dev/vgXX /dev/dsk/cXtXd0
7. Active the the volume group.

# vgchange -a y /dev/vgXX
Add the mount point directories and mount the lvols within that volume group.
Edit the /etc/checklist or /etc/fstab file.
22
15 Splitting mirrors to perform backups.
If you can, make sure the database is not active before the split.
For the example we will use /dev/vg02/lvol1.
Use lvdisplay -v /dev/vg02/lvol1 to verify /dev/vg02/lvol1 is mirrored and current.
1. Split the logical volume.

# sync
# lvsplit -s backup /dev/vg02/lvol1
The system console will display a message of:

Logical volume "/dev/vg02/lvol1backup" has been successfully created with character device
"/dev/vg02/rlvol1backup".
Logical volume "/dev/vg02/lvol1" has been successfully split.
The original logical volume /dev/vg02/lvol1 continues to be available for use.

The new logical volume /dev/vg02/lvolbackup, split off from the original, now exists.
2. Perform fsck on the split logical volume.

# fsck /dev/vg02/rlvol1backup
or
# fsck -F vxfs /dev/vg02/rlvol1backup (for JFS file systems)
3. Perform the backup of the file system, "/dev/vg02/lvol1backup". (favorite backup utility)
4. Unmount the file system (if you had mounted it) in the split-off logical volume when you are ready
to merge the split pairs. When backup is done.
5. Merge the split-off logical volume and the original logical volume back into one mirrored logical
volume.
# lvmerge /dev/vg02/lvol1backup /dev/vg02/lvol1
Note: lvmerge split-off_logical_volume original_logical_volume (do not reverse the order)
6. Verify the merged mirror.

# lvdisplay -v /dev/vg02/lvol1
23
16 Tech Info Bits
*****************************************************************************
Use the "dd" command to identify a disk with out crawling all over the floor.
When you issue the command, look for the solid blinking green LED. "Ctrl+c" will stop
the read. DD can also be used as an online disk read test. Files in should equal files out.
# dd if =/dev/rdsk/cXtXd0 of =/dev/null bs=1024k
*****************************************************************************
Install a LVM file system:

*******Volume Group Creation********
# pvcreate /dev/rdsk/cXtXd0 (-f: force, -B: bootable)
# mkdir /dev/vg01 (vg01 is used for this example)
# mknod /dev/vg01/group c 64 0x010000 (for vg02, the minor number would be 0x020000)
# vgcreate /dev/vg01 /dev/dsk/cXtXd0
*******Logical Volume Creation*******

# lvcreate -L xxx /dev/vg01 (xxx = size of the logical volume in MB, -L = long file names)
# newfs /dev/vg01/rlvol1 C2474S (or any relevant disk type, 9.X) (disk type not used for 10.X)
# newfs -F vxfs /dev/vg01/rlvol1 (use this newfs command if the file system is JFS)
# mkdir /mountdir (mountdir = any moint point name the customer wants)
# mount /dev/vg01/lvol1 /mountdir
You will have to edit /etc/fstab if you want this new file system to mount at boot up.
*******Create a RAW partition:*******

# pvcreate /dev/rdsk/cXtXd0
# mkdir /dev/vg01 (vg01 is used for this example)
# mknod /dev/vg01/group c 64 0x010000
# vgcreate /dev/vg01 /dev/dsk/cXtXd0
# lvcreate -n name1 /dev/vg01
# lvextend -L xxx /dev/vg01/name1
Increase the size of a logical volume from xxx MB to yyy MB:

# lvextend -L yyy /dev/vg01/lvol1 (yyy = size of the new logical volume size in MB)
# umount /mountdir
# extendfs /dev/vg01/rlvol1 (use extendfs -F vxfs /dev/vg01/rlvol1 if you have a JFS file system)
Decrease the size of a logical volume from xxx MB to yyy MB:

NOTE: DATA WILL BE LOST!
*******SAVE THE DATA IN THE DIRECTORY TO BE REDUCED*******
# umount /mountdir
# lvreduce -L yyy /dev/vg01/lvol1 (yyy = size of the new logical volume size in MB)
# newfs /dev/vg01/rlvol1 C2474S (9.X) (disk type not used for 10.X)(newfs -F vxfs for JFS)
*******RESTORE DATA TO THE REDUCED DIRECTORY*******
24
******Create a MIRRORED set*****

NOTE: Mirroring can only be done within a Volume Group.
# pvcreate /dev/rdsk/cXtXd0 (use pvcreate -B if the disk will be a root boot mirror disk)
# vgextend /dev/vgXX /dev/dsk/cXtXd0
# mkboot /dev/rdsk/cXtXd0 (use the mkboot commands if root boot mirror disk)
# mkboot -a "hpux (;0)/stand/vmunix" /dev/rdsk/cXtXd0
# lvextend -m 1 /dev/vg01/lvolX /dev/dsk/cXtXd0 (mirror disk device file)
lvextend all lvols in order from lvol1 - lvolX in proper sequence on a root boot disk
# lvlnboot -Rv (verify lvol1 is defined as boot).
Add a disk drive to an existing Volume Group:

# pvcreate -f /dev/rdsk/cXtXd0 (disk device file of the new disk added)
# vgextend /dev/vg01 /dev/dsk/cXtXd0 (disk device file of the new disk)
Boot from a bad Root Volume Group:

(Typically, if lvlnboot has not been run following changes to the root Volume Group.)
ISL> hpux -lq (;0)/stand/vmunix (10.X)

*******System Boots*******
*******Login as root then:*******
# lvlnboot -Rv /dev/vg00
25
17 Moving an existing root disk to a new hardware path.
Before you shutdown, document /etc/lvmtab, note which disks are in vg00.
Shutdown the system and connect the existing root drive to the new path.
Boot up the system and escape boot sequence, the boot off the root drive at the new hardware path to ISL.
1. Boot into LMV maintenance mode

ISL> hpux -lm (;0)/stand/vmunix
2. Find the new hardware path of the root disk.

# ioscan -fnC disk
3. Install the special file for the root disk at the new hardware path.
# insf -e -H <hardware_path>
4. Find the special device file for the root disk.

# ioscan -fnC disk (example of device file name: /dev/dsk/c2t6d0)
5. Note the minor number of the group file in /dev/vg00.
6. Export the root file system vg00 to a mapefile.

# vgexport -v -m /tmp/vg00_map /dev/vg00
7. Create the volume group and the group file to be used for the vgimport command.
# mkdir /dev/vg00
# mknod /dev/vg00/group c 64 0x000000
8. Import the root disk and any other disks that might be in the vg00 volume group.
# vgimport -v -m /tmp/vg00_map /dev/vg00 /dev/dsk/cXtXd0
<and any other disk in vg00>
9. Activate the volume group.

# vgchange -a y /dev/vg00
10. Update the LIF's Lable file with the information contained in the BDRA.
# lvlnboot -Rv
11. Reboot the system

# shutdown -r 0
DO NOT forget to update the Stable Storage new boot path from the ISL prompt!
26
Storage
Chapter 2: Storage
1 Overview of most common commands for diskarrays
AutoRaid 12 & 12H FC60 VA7XX0

Handle Edison & Icicle Optimis Cassini/Cronus
Monitor ARMServer AM60Srvr sanmgr
/opt/sanmgr/hostagent/sbin/
HA_trigger start
dial_trigger start
Start monitor /sbin/init.d/hparray start /sbin/init.d/hparamgr start New -1.06 will start watchdog timers as
well
/opt/sanmgr/commandview/client/sbin/
HA_Dial_Start
/opt/sanmgr/hostagent/sbin/
HA_trigger stop
dial_trigger stop
monitors may not stop if Storage Area
Manager or measureware products are
running.
New – 1.06 will stop watchdog timers as
Stop Monitor /sbin/init.d/hparray stop /sbin/init.d/hparamgr stop
well
HA_Dial_Stop
New - To prevent commandview from
starting
vi /etc/rc.config.d/ha.conf
Change HOSTAGENT_START=1 to 0
Directory /opt/hparray/bin /opt/hparray/bin /opt/sanmgr/commandview/client/sbin
Log Clear - /opt/hparray/bin/amlog -C -c
controller/drive serial_#
27
Storage
/var/opt/sanmgr/commandview/
server/logs/device/wwn/date
wwn is the 50060B0xxxxxx of controller
Log Directory /var/opt/hparray/log /var/opt/hparray/log date is year-month 2002-10
/var/opt/sanmgr/commandview/
server/logs/performance
xx-usage.log = daily armdsp -a
xx-cntrl.log
Logs files LXXXXXXX AMMXXXXX
where xx is a calander date
example 19-usage.log
logprn -a $id -v or –b
for valt
logprn -t All -v -a s/n > logprn.txt
armlog -c s/n > armlog.txt 1.07
armlog –e s/n > armlog.txt pre 1.07
for performance logs
armperf -c ARRAY -x COMMA s/n >
armperf_array.txt
armperf -c OPAQUE -x COMMA s/n >
armperf_opaque.txt
arraylog -u -e $id amlog -a $id -s -e opt avail
Logs Display $ id must be s/n
armperf -c LUN -x COMMA s/n >
logprint $id -s -e opt avail armperf_lun.txt
New with command view 1.05 &1.06

armdiag –I –if filename $id = getVAlogs
armdiag -w $id lists non redundant disks
see note – from phil
armperf –c FRONTEND|BACKEND
–x COMMA $id
for trouble shooting backend issues
armperf –c DISK-PORTS –x COMMA $id
= armlog –d disk –p 0xd $id
28
Storage
Flush logs to files amutil -l $id flush logs
Remove logs kill monitor, rm files, touch kill monitor, rm files- lock, touch
/var/opt/hparray/log/LOGCATG catalo g logdel –t all –a s/n
restart monitor rm /var/opt/hparray/log/AMM* option stop time –e MMDDHHMM
arraylog -C -c $id rm /var/opt/hparray/AM60Srvr.lock
touch Note: only logs older than two months can
/var/opt/hparray/log/AMMCATLG be removed
restart monitor
amutil -p remove oldest log file
FIRMWARE AutoRaid 12 & 12H FC60 VA 7XX0
Directory /opt/hparray/bin /opt/hparray/bin /opt/samngr/commandview/client/fwdownload
armdownload -C –O cnt_# fwRev $id
armdownload –B fwRev $id
amdload -C all file_name $id
Download Controller download -C file_name $id armdownload -L JAx/Cx fwRev $id
amdload -S file_name $id bcc
clear advisory lock
armmgr –b FwDownload $id
clear advisory lock
armdownload –I -D drive_# fwRev $id
or clear advisory lock
Download Drive download -D # file_name $id amdload -D file_name $id
armdownload –D –P drv model_# fwRev $id
Failure to clear advisory locks could result
in a rebuild taking place
Migrate Controller armdownload -M # $id
download -M $id from primary
firmware where # is the source FRU M/C1or M/C2
Host commandview sdm

Firmware/software Disk/controller firmware
Disk/controller firmware
29
Storage
ARRAY AutoRaid 12 & 12H FC60 VA 7XX0

Scan for new arrays arraydsp -i amdsp -R armdiscover
dsp
Find array id arraydsp -i amdsp -i armdsp -i
armtopology display wwn & devfiles
Format Array arrayfmt -F $id amfmt $id fmt
armfmt $id
Optimization armmgr –o on|off $id optimization
Prefetch - Cache armmgr –P on $id or vfpmgr –p on
armmgr –P off $id or vfpmgr –p off
Rebuild Configure arrayrbld -a on|off $id amutil -R lun# -f req $id rble -a on|off
arrayrbld -P high|low $id req 1-50 higher is faster 1 default autorebuild
arraymgr -a on|off $id armmgr –a on|off $id
optimization
armmgr –o on|off $id
rebuild progress
armrbld –p $id
set rebuild priority
armrbld –P high | low $id
cancel rebuild
armrbld –c $id
Note you can not cancel a rebuild
Started by auto rebuild
Rebuild Force arrayrbld -r $id 12H only amutil –P lun# $id lun parity recover maps via vfp
Never use these scan recover
commands on an array- if a disk were failed or faulted it could recover maps
Unless all hardware is be put into a rebuild state. All hardware armrecover $id
must be present and functional
Present and Functional start rebuild
Failure to do so will armrbld –r $id
amutil -L lun# $id revive lun
result rebuild progress
use only if all disks fail at the same
in corruption or Data armrbld –p $id
time. This command does not force
30
Storage
loss a rebuild cancel rebuild

armrbld –c $id
Note you can not cancel a rebuild
Started by auto rebuild
Rebuild Progress arrayrbld -p $id amdsp -r $id recover -s recover -c
ammgr -s $id vfpbld -p
armrecover -s $id
Resiliency Configure arraymgr -J SingleController| armmgr -J SingleController|Secure|Normal|
Secure|Normal| RestrictedNormal|
HighPerformance $id HighPerformance $id
Set Array address / id front panel rear panel switches vfpmgr -os $os -c $cnt
arraymgr -c X|Y $id pri contr vfpmgr -L $loop -c $cnt
arraymgr -C X|Y $add $id vfpmgr -t $top -c $cnt
armmgr -L M/C # $id
armmgr -y $TOP $cnt $id
Set Array Alias arraymgr -D alias $id ammgr -D alias $id armmgr -D alias $id
Set Battery date ammgr -b $id reset battery
Set Controllers active ammgr -c AA $id
active
Start / Reset Array front panel enter shutdown front panel switch per controller mgr -R full | partial
armmgr -R full | partial
flush cache, switch mgr -s shut | start
Start / Stop Array armmgr -s shut | start $id
Status of Array arraydsp -a $id amdsp -a $id dsp dsp -s dsp -f
armdsp $id
armdsp -a $id
armdsp -f $id
Status of Controller arraydsp -c $id amdsp -c $id dsp -c dsp -e
armdsp -c $id
Sync Date with Host ammgr -t $id armdiscover
31
Storage
LUN AutoRaid 12 & 12H FC60 VA 7XX0

amcfg -L cnt:lun -d list -r # -s # $id
cnt:lun = A|B : lun #
front panel list = chan:id,chan:id .... cfg -L # -a $size Suffix(GB –g rg#
Create Lun
arraycfg -L # -a $cap $id -r # = raid 1|5 armcfg -L # -a $size –g rg# $id
-s# =stripe 4|8|16|64
recommend you use SAM or STM
Change Controller in arraymgr -c X|Y $id amcfg -M lun# -c A|B $id dsp -L $id
charge/Lun ownership armdsp -L $id
cfg -L # -d
Remove Lun arraycfg -L # -d $id amcfg -D # $id
armcfg -L # -d $id
Replace Lun 0 amcfg -R cnt:0 -d list -r # -s # $id
cnt= A|B
list = chan:id,chan:id...
-r #= raid 1|5
-s #=stripe 4|8|16|64
recommend you use SAM or STM
dsp -L
Status of Lun arraydsp -l $id amdsp -l $id
armdsp -L $id
ammgr -U -l 31 $id enable
UTM Lun 31
ammgr -u $id disable
DRIVE AutoRaid 12 & 12H FC60 VA 7XX0
Force - Fail drive amutil -D chan:d $id
Drive force fix - amutil -d chan:d $id
replaced amutil -rr chan:d $id
replace &
rebuild
Hot Spare arraymgr -h on|off $id ammgr -h chan:id $id add hs armmgr -h on|off $id HP01/HP02
ammgr -d chan:id $id remove hs
armmgr –h None|Automatic|LargestDisk|
32
Storage
LargestTwoDisks $id >HP02

ID drive/FRU – amutil -f chan:id $id flash led armmgr -I enclosure/fru on|off $id
Flash the light amutil -s $id stop flash example # armmgr -I M/D7 on va7400
Include - auto arraymgr –I on|off $id armmgr –I on|off id

Include - manual armcfg –D enclosure/fru –a $i
Loop Map - back end armdsp -t $id
must have commandview 1.06 or later
drivetest -D A|Bx -r 100 $id
-s selftest
Test Drive
-w destructive write test –
don't use
amdsp -d $id
dsp -d or dsp -e
Status of Drive arraydsp -d $id amutil -f chan:id $id flash led
armdsp -d $id
amutil -s $id stop flash
Write Cache wce amutil -w on $id enable wce
amutil -w $id disable wce
TOOLS AutoRaid 12 & 12H FC60 VA 7XX0
Capture and Log ar_log_tool pc based log refl_debug_4.1.zip valt PC based log decoder
Analyzers decoder backend capture script - runs on # logprn -t ALL -v -a S/N > file or
# logprint > lp.txt as data pc # armlog -e $id > armlog.txt or
source ExamDiff.exe # armdiag –I –if filename $id preferred
pc tool - displays diffs of 2 files getVAlogs gather data / logs
PFU web site command view sdm “serial”
other FC60 tools
AutoRaid 12 & 12H FC60 VA 7XX0
33
Storage
2 Special commands for VA 7XY0
Disabling Security –
Setting up security – New with 1.07
Allow normal access to commandview
vi /etc/opt/sanmgr/commandview/server/
secadmin -host my_host
config/PanConfigParams.txt
username Administrator
SECURITY_ENABLED=true change true to false
password administrator
Resetting security on Windows ->Administrator
Commandview 1.07 add –user root –group Administrator –host my_host
remove old
secadmin Note :
C:\Program Files\Hewlett-Packard\sanmgr\
groups and users must be in /etc/passwd & /etc/group
commandview\server\data\UserDBFile
enter password
Resetting security on UNIX ->Adminsistrator
re-enter password
remove old
exit
/var/opt/sanmgr/commandview/server/data/
web GUI will require a valid user and password
UserDBFile
Obtaining license – New with 1.07
60 day grace period – free use Installing License – New with 1.07
License Documentation CMD 1.07 licApp - easy to use GUI – very slow
License web site http://www.webware.hp.com must run from a X window – like reflection X
1. Generate product password ; next requires file already be created
2. use CVSDM107UPGRADE as order # ; next # licApp –i HP_CommandView_SDM
Commandview 1.07 3. check product box for T1086A ; next 1. Browse to /var/opt/OV/HPOvLIC
licApp or licUtil 4. enter system name and select OS ; next 2. select cmdview_lic.txt ; Choose
5. enter customers email address & password ; next 3. import
6. enter license requestor & owner info ; next licUtil - requires more input
7. print and save license requires file already be created in
cut the license info and place in a file on the host /var/opt/OV/HPOvLIC/cmdview_lic.txt
/var/opt/OV/HPOvLIC/cmdview_lic.txt # licUtil –i cmdview_lic.txt HP_CommandView_SDM
save as one line in a file with no embedded cr’s.
7100/ 7400 7110/ 7410
Queue depth entered is assigned to both controllers Queue depth entered is assigned to a controller host
Queue Depth Formula # of disks in RG x Multiplier =que# bus
RG1 odd disks Formula # of disks in RG x Multiplier / 2=que#
RG2 even disks RG1 odd disks
34
Storage
Multiplier 2 -3 RG2 even disks

armmgr –q que# M/C1.H1 $alias Multiplier 2 -3
M/C2.H1 armmgr –q que# M/C1.H1 $alias
M/C1.H2
M/C2.H1
M/C2.H2
set autoformat on off on default armmgr -f on|off $id
set host behavior armmgr -B HpuxFCDriver|Windows/
. LinuxFCDriver 1|2 $id
reset host fibre M/C1.H1 or M/C2.H1 armmgr -c M/C1.H1 $id
Special Configuration
reset jbod fibre M/C1.J1 or M/C2.J1 armmgr -c M/C1.J1 $id
options
reset lock CreateLun|FwDownload|Security|Select|PassThru| armmgr -b option $id
HostPort .
read ahead prefetch on armmgr –P on $id or vfpmgr –p on
read ahead prefech off armmgr –P off $id or vfpmgr –p off
read host info to file armhost -r -f file_name $id
Host interface behavior write host info to file armhost -w -f file_name $id
delete interface wwn entries armhost -d wwn $id
read features licenses & states
URL www.webkey.external.hp.com
armfeature -r $id
HP Authorization Phone US 801-431-1451 M-F 6-6 MST
License/features armfeature -a -f feature_string -k key $id
Europe +31-555-384-210 Netherlands
Asia +81-03-3227-5289
Japan
copy lun 1 to a new lun2 true copies data armcopy -p lun1 -s lun2 -a true $id
copy lun 1 to lun2 false does not copy data armcopy -p lun1 -s lun2 -a false $id
Business copy
assign lun # to copy -x set lun as active inactive armcopy -s # -x true|false -a true|false $id
copies data from business copy to parent lun !!! armcopy -r lun $id
armsecure -n passwd -p passwd $id
virtual front panel
armsecure -e -p passwd $id
secure -R reset password
Security armsecure -d -p passwd $id
secure -e enable
armsecure -r -f file_name -p passwd $id
secure -d disable
armsecure -w -f file_name -p passwd $id
35
Storage
looking at performance now you can use a start -s

Performance armperf -c LUN|OPAQUE|ARRAY $id
-e end time MMddhhmm
http://url_or_ip:4096/Launcher.html
commandview 1.07 will require a valid user &
Web Tool SDM
password
see secadmin
36
Storage
3 Argument for diskarray commands
$add scsi address

$cap capacity = size see suffix
$cnt controller slot 1 or 2
$id array serial number or alias
$loop Loop id
$os hpux nt win2k linux
$top Topology 0 = Auto detect 1 = Private Loop 2 = Pulic Loop 3 = Fabric
$TOP Topology automatic|private|public|pointtopoint|fabric
chan:id fc60 channelpath 1-6 : scsi id 0-4 8-12
feature_string string that describes license
file_name file_name - note for armhost use fully qualified path
fw_rev firmware file revisio
key string that enables feature
log_directory /var/opt/sanmgr/commandview/server/logs/wwn/date
passwd default password is AUTORAID
raw_dev Raw device file in form /dev/rdsk/cxtyd0
Suffix K Kilbytes M megabytes G Gigabytes
# slot or a lun number
armcopy lun1 parent lun2 secondary -a true copies data false does not copy data
armhostfile Example: Host ID Table-the wwn node name of interface card from fcmsutil
HpuxFCDriver WindowsFCDriver LinuxFCDriver
wwn port Host port
10000060b0fa3139 HpuxFCDriver
securityfile Example: Lun Security Table

W write R read WC write/configure 0 invisible
Type WWN node lun permissions
DEFAULT null 1 W
DEFAULT null 2 W
DEFAULT null 5 W
NODEWWN 10000060b0fa3543 4 W
Note: These are not all the possible arguments - check the man pages for more
options.
37
Storage
4 Autoraid
Commands
arraydsp -i gives the serial number of all the arrays
export <ID>=<serial number> gives a an arrayID (a name you assign so that you do not to use
the long serial number with the commands
arraydsp -a <arrayID> all (status)
arraydsp -c <arrayID> controller status
arraydsp -d <arrayID> disk status
arraydsp -h <arrayID> hardware status
arraylog -e <arrayID> displays the contents of disk array controller event log
arraymgr -s shut<arrayID> Shutdown array
drivetest -D AX -r 100<arrayID> performs read/verify test on 100% of disk AX on
<arrayID> non-destructive
dteststat -D AX -? <arrayID> displays results of drivetest on AX
logprint -d {-s starttime} {-e stoptime} <arrayID> logfile is put in default location for the
specified time (mmddhhmm)
/var/opt/hparray/log
Partial Menu Selection
Push "Menu", then use the "+" or "-" buttons to scroll through the menus, "enter " to activate a selection
Shutdown → Confirm (or Cancel)
View Settings → Disk States gives the state of each disk in the array
Rebuild in progress or complete
System State
Warnings
Number of disks
Primary Cntrl
SCSI ID x
SCSI ID y
Firmware
Array S/N
ActivSpare
L-Drv States
Cntrl Changes → Reset Cntrl
[Copy Firmware]
[Recover] Restores map of drives
[Switch Cntrl]
Reset Cntrl
38
Storage
5 DS2100
Notes
• Only 1 enclosure in supported in the SE mode.
Up to 3 enclosures for a maximum of 12 drives otherwise.
• Maximum cable length is 12 meter for LVD
• Disk system requires a terminator at the end of the bus.
If you connect DS2100 to TWO HBA (SCSI-id 6 and 7), switch can NOT be set to A
There are NO disk fault led's on the DS2100
LED State Description

System Power
Green Power is on
Blinking Malfunction- Either fan is not operating or
…………………………………..internal voltage is too low
Off Power is off
Disk On Installed and spinning up. If .the LED stays

…………………………………..on for 3 minutes after initial power up disk
…………………………………..may be bad
Flashing Disk i/o
Off Not installed, not active, or no i/o
39
Storage
6 DS2300
Switch Bank 1 Off –1 On - 0

1 Full Bus Create a single bus up to 14 disk Splits the bus into buses up to 7
drives disk each
2 SES / SAF-TE SAF-TE Mode (SAF-TE mode is the SES mode
default. For HP-UX using EMS, set
to SES)
Switch Bank 1 Off –1 On – 0
1 Bus Reset: Hot Swap disk Automatically issues a SCSI bus Lets the host detect change and
reset when a disk is removed or determine action
inserted
2 Bus Reset: Power Fail Issues a SCSI reset when the
power supply/fan module
indicates it will go offline
3 Bus Reset: Hot Swap BCC Automatically issues a SCSI bus Lets the host detect change and
reset when a BCC is removed or determine action
inserted
40
Storage
SCSI ID
Physical Disk Slot
Full Buss Split Bus (Right BCC) Split Bus (Left BCC)
1 0 0
2 1 1
3 2 2
4 3 3
5 4 4
6 5 5
7 6 6
8 8 8
9 9 9
10 10 10
11 11 11
12 12 12
13 13 13
14 14 14
Right BCC 15 15
Left BCC 15
41
Storage
7 DS2405
LoopID = [(EnclID x 16) + (Disk slot -1)]

Encl ID 0 1 2
Loop ALPA Loop ALPA Loop ALPA
Disk slot Bus Tgt Bus Tgt Bus Tgt
ID (hex) ID (hex) ID (hex)
1 0 0 0 EF 16 1 0 CD 32 2 0 B2
2 1 0 1 E8 17 1 1 CC 33 2 1 B1
3 2 0 0 E4 18 1 2 CB 34 2 2 AE
4 3 0 3 E2 19 1 3 CA 35 2 3 AD
5 4 0 4 E1 20 1 4 C9 36 2 4 AC
6 5 0 5 E0 21 1 5 C7 37 2 5 AB
7 6 0 6 DC 22 1 6 C6 38 2 6 AA
8 7 0 7 DA 23 1 7 C5 39 2 7 A9
9 8 0 8 D9 24 1 8 C3 40 2 8 A7
10 9 0 9 D6 25 1 9 BC 41 2 9 A6
11 10 0 10 D5 26 1 10 BA 42 2 10 A5
12 11 0 11 D4 27 1 11 B9 43 2 11 A3
13 12 0 12 D3 28 1 12 B6 44 2 12 9F
14 13 0 13 D2 29 1 13 B5 45 2 13 9E
15 14 0 14 D1 30 1 14 B4 46 2 14 B4
SES 15 0 15 CE 31 1 15 B3 63 3 15 73
Encl ID 3 4
Loop ALPA Loop ALPA
Disk slot Bus Tgt Bus Tgt
ID (hex) ID (hex)
1 48 3 0 98 64 4 0 72
2 49 3 1 97 65 4 1 71
3 50 3 2 90 66 4 2 6E
4 51 3 3 8F 67 4 3 6D
5 52 3 4 88 68 4 4 6C
6 53 3 5 84 69 4 5 6B
7 54 3 6 82 70 4 6 6A
8 55 3 7 81 71 4 7 69
9 56 3 8 80 80 4 8 67
10 57 3 9 7C 73 4 9 66
11 58 3 10 7A 74 4 10 65
12 59 3 11 79 75 4 11 63
13 60 3 12 76 76 4 12 5C
14 61 3 13 75 77 4 13 5A
15 62 3 14 74 78 4 14 59
SES 79 4 15 56 79 4 15 56
42
Storage
Encl ID 5 6
Loop ALPA Loop ALPA
Disk slot Bus Tgt Bus Tgt
ID (hex) ID (hex)
1 80 5 0 55 96 6 0 3A
2 81 5 1 54 97 6 1 39
3 82 5 2 53 98 6 2 36
4 83 5 3 52 99 6 3 35
5 84 5 4 51 100 6 4 34
6 85 5 5 4E 101 6 5 33
7 86 5 6 4D 102 6 6 32
8 87 5 7 4C 103 6 7 31
9 88 5 8 4B 104 6 8 2E
10 89 5 9 4A 105 6 9 2D
11 90 5 10 49 106 6 10 2C
12 91 5 11 47 107 6 11 2B
13 92 5 12 46 108 6 12 2A
14 93 5 13 45 109 6 13 29
15 94 5 14 43 110 6 14 27
SES 95 5 15 3C 111 6 15 26
43
Storage
8 FC60
Commands
amdsp -i Gives the arrayID number of the machine
amdsp -a <arrayID> Gives the status of all the drives, bcc, & power
supplies. Also gives the firmware levels
amdsp -d -S <arrayID> Status information of all disk
swlist -l product | grep -e "HP Array" Identifies patch level & patches on
HPUX 10.20 & 11.0
ammgr -b <arrayID> resets battery cache, must be done after replacing battery
and after battery is charged, will not work if disk rebuild is in progress .
Identifying Split and Full Bus
• Split bus will have a will have 2 SCSI cables coming from the FC60 Controller (channels 1-6) to
each SC10. Each FC60 can handle up to 3 SC10's in a split bus configuration.
Ex: Channel 1 on the FC60 controller connected to BCC A on the first SC10 and channel 2
connected to BCC B on the same SC10.
• Full bus will have one cable coming from the FC60 Controller (channels 1-6) to each SC10. Each
FC60 Controller can handle up to 6 SC10's in full bus
Notes:
• Controller upper A, lower B, Each has a unique FC address assigned through dip switches.
Controllers are hot swappable but you must follow the proper procedures for replacing them.
Dimms and Simms may need to be transferred to new controller
• Cache Dimm must be the same on both controllers
• Unused LVD ports must be terminated
• Power supplies, MIA, fan controller module, BBU, disk, disk fan modules and power supply fan
modules are hot swappable
• JBOD - address 4, 5, 6 & 7 reserved for host controller. ARRAY - 5,6, & 7 reserved for host
controller
44
Storage
Status LED's FC60

Codes alternate between (where listed) one of the codes in the "code" column and one of the codes in the
"alternating" column approximately every two seconds
CODE ALTERNATING DESCRIPTION CAUSE/SOLUTION
OX20/21/22/23/24/2
5/26/27/28/29/2A/2B Kernel problem
/2C/2D/2E/2F/30
0x33/36/37/38 0x00
Diags Running
0x44 0x01/02/03/12
replace/reseat controller
0x54/55 0x00
0x65 0x00/10/20 Host SCSI/Fibre diags running
0x67 CRC Failure
0x6B 0x01/02/03/04/05/11/12/13/14/15/21/22/23/24/25
Drive SCSI channel X diags running
0x6C 0x01/02/03/04/05
0x61/64/65 Download Problem
Firmware file is bad/download good firmware
0x80 In between Programming steps
0x80 0x00 Normal State (heartbeat)
0x81/82/83/84/89/ Erasing/Writing/Verifying Flash or Boot Firmware file is downloading or firmware file
8A/8B/8C Segments is bad/ download good firmware
0x6E 0xEE Normal during firmware download. May be in
Passive Controller this state after power on, use AM60 software
to set to Active
0xAA PCI Bus Fault
0xE1 Controller Exception
0xE7 Download a matching combination of
Kernel missing or kernel CRC mismatch controller Bootware & controller Application
firmware
0XE8 Application Firmware Missing or Download Application firmware that matches
Application CRC mismatch current Bootware
0xFF Controller held in hardware reset/
reseat/replace controller
•Reading the codes ex: 0x2A = Ignore the 0x ," 2 " represents the leftmost set of four status LED's in binary ( Ο Ο Ο ) , "A" represents the right set of four status LED's in
binary ( Ο Ο ) A = 10
SC 10 Addressing - Split Bus

ch 1 1:0 2:0 1:1 2:1 1:2 2:2 1:3 2:3 1:4 2:4
ch 2 3:0 4:0 3:1 4:1 3:2 4:2 3:3 4:3 3:4 4:4
ch 3 5:0 6:0 5:1 6:1 5:2 6:2 5:3 6:3 5:4 6:4
SC 10 Addressing - Full Bus

ch 1 1:0 1:8 1:1 1:9 1:2 1:10 1:3 1:11 1:4 1:12
ch 2 2:0 2:8 2:1 2:9 2:2 2:10 2:3 2:11 2:4 2:12
ch 3 3:0 3:8 3:1 3:9 3:2 3:10 3:3 3:11 3:4 3:12
ch 4 4:0 4:8 4:1 4:9 4:2 4:10 4:3 4:11 4:4 4:12
ch 5 5:0 5:8 5:1 5:9 5:2 5:10 5:3 5:11 5:4 5:12
ch 6 6:0 6:8 6:1 6:9 6:2 6:10 6:3 6:11 6:4 6:12
BCC DIP Switch Settings 1=on, 0=off

SW ON OFF Explanation
1 X Full Bus
1 X Split Bus
2 X Active JBOD Mode, rotary switches are ignored
2 X Active Array Mode, rotary switch in full use (needed for FC60)
3 X Automatic reset of bus upon power failure
4 X High SCSI disk addressing is used in split bus mode, (switch 1 must be off)
4 X Low SCSI disk addressing is used in split bus mode, (switch 1 must be off)
5 X SCSI Reset is generated whenever a disk module is removed
45
Storage
9 SC10
46
Storage
10 FC10
Enclosure Ids (0 through 9) are set with a dial on the LCC. This settings determines the FC-Al Ids and
corresponding HW-paths for each slot and LCC based on the Enclosure ID.
The Enclosure Ids on both LCCs in the same disk system must be identical
Disk LoopID = [(EnclID x 12) + Disk slot]

LCC LoopID = [(EnclID x 12) + 10]
Encl ID 0 1 2 3 4
HW path HW path HW path HW path HW path

Loop Loop Loop Loop Loop
Slot #
ID ID ID ID ID
Bus Targt Bus Targt Bus Targt Bus Targt Bus Targt
0 0 0 0 12 0 12 24 1 8 36 2 4 48 3 0
1 1 0 1 13 0 13 25 1 9 37 2 5 49 3 1
2 2 0 2 14 0 14 26 1 10 38 2 6 50 3 2
3 3 0 3 15 0 15 27 1 11 39 2 7 51 3 3
4 4 0 4 16 1 0 28 1 12 40 2 8 52 3 4
5 5 0 5 17 1 1 29 1 13 41 2 9 53 3 5
6 6 0 6 18 1 2 30 1 14 42 2 10 54 3 6
7 7 0 7 19 1 3 31 1 15 43 2 11 55 3 7
8 8 0 8 20 1 4 32 2 0 44 2 12 56 3 8
9 9 0 9 21 1 5 33 2 1 45 2 13 57 3 9
LCC 10 0 10 22 1 6 34 2 2 46 2 14 58 3 10
RSVD 11 0 11 23 1 7 35 2 3 47 2 15 59 3 11
47
Storage
Encl ID 5 6 7 8 9
HW path HW path HW path HW path HW path

Loop Loop Loop Loop Loop
Slot #
ID ID ID ID ID
Bus Targt Bus Targt Bus Targt Bus Targt Bus Targt
0 60 3 12 72 4 8 84 5 4 96 6 0 108 6 12
1 61 3 13 73 4 9 85 5 5 97 6 1 109 6 13
2 62 3 14 74 4 10 86 5 6 98 6 2 110 6 14
3 63 3 15 75 4 11 87 5 7 99 6 3 111 6 15
4 64 4 0 76 4 12 88 5 8 100 6 4 112 7 0
5 65 4 1 77 4 13 89 5 9 101 6 5 113 7 1
6 66 4 2 78 4 14 90 5 10 102 6 6 114 7 2
7 67 4 3 79 4 15 91 5 11 103 6 7 115 7 3
8 68 4 4 80 5 0 92 5 12 104 6 8 116 7 4
9 69 4 5 81 5 1 93 5 13 105 6 9 117 7 5
LCC 70 4 6 82 5 2 94 5 14 106 6 10 118 7 6
RSVD 71 4 7 83 5 3 95 5 15 107 6 11 119 7 7
48
Storage
11 Nike array
Accessing GRID Manager
1. At the function keys on the top of the keyboard

User system
F8 config
F5 Terminal Config
Tab to TermMode and push F2 (next choice) until TermMode changes to EM100
F1 (save config)
2. Ctrl L you should be in Grid Manager
3. To enter FE Mode
Ctrl P then type fe and enter
4. Don’t forget to return TermMode to original setting after leaving Grid Manager
To do an inventory on a nike array;
Get into FE mode, go to the Presentation Utility and type lower case i or the # key.
The i key gives you a hardware inventory of the boards and hard drive sizes.
The # key gives you more detail about the hard drives themselves.
To Exit out of FE mode on a Nike array type;
CTRL P at the same time then holding down the SHIFT key type USER.
To update firmware on a Nike array or drive:
1. Boot from the dfdutil utility tape or CD that has Nikearry file.
2. Go into ODE
3. Disable one of the Nike processors. (Phoenix controllers are done in grid manager by going to
Change SP parameters menu, then go into Select soft disable).
4. Type nikearry to run the firmware utility.
5. Enter the password.
6. Put in the path or paths of the sp controllers to update.
7. Type download
8. Get address of sp controllers still enabled (if updating more than 1 controller
9. Type SP_925_00 for Suana or older controller or SP_955_01 for phoenix or newer controller.
49
Storage
10. Once download is complete enable the sp controller that you disabled earlier. If firmware is not
moved over from updated controller to the controller that was disabled repeat from step 3 disabling the
controller that was just updated and update the other one.
11. After the sp controllers are updated you can now update the nike drives is need be.
12. Type download
13. Put in firmware file NIKE2GHH03 for ST15550N drives or NIKE2GLP08 for ST32550N drives.
14. Can do a reset command when done to check if the updates took.
15. Check grid manager for drive updates.
50
Storage
12 VA7xy0
VA Troubleshooting Do’s and Don’ts
When troubleshooting a VA disk array (VA7100, VA7400, or VA7410), it is very important to keep the following
points in mind to avoid data loss and minimize downtime to the customer.
1) DO NOT power off the array. Disk arrays are designed to be serviced while powered up, so powering
down the array to replace parts can cause problems and even data loss under rare circumstances. When at
all possible, leave it powered up.
2) DO obtain and analyze all log data before replacing or even removing any parts. This should be done
with logprn on the host to get historical data, armlog to pull the current logs from the array, and vfplog from
the serial port VFP in the event that the array is not accessible from the host.
3) DO find out if the host can communicate with the array with armdsp and armlog. If the array is not
accessible from the host, disconnect the fibre cables to the host and connect a laptop to the serial port of each
controller one at a time. Collect logs and look for any error indications displayed while logging the output to
a file. If the array appears to be continually resetting itself, the initialization steps will be displayed along
with any messages like “Abterm.” Only the controller experiencing an error or abterm will display that error
message.
4) DO NOT hesitate to get help if you see “No Map,” “No Quorum,” or “Abterm,” and are not sure exactly
what to do. These conditions are very serious, and data loss can result if the wrong decisions and steps are
taken. When in doubt, escalate.
5) DO know what will be required when you escalate. An armdsp –a, logprn, and armlog or vfplog will be
asked for to proceed with an escalation. DO everything you can to get that log data.
6) DO NOT pull batteries off of any controller removed from the array unless you are sure the array is in a
Ready or Single Controller Warning state with one functioning controller installed in the array; or, you are
sure there is no usable map in memory; or, you want to purposely clear the maps in memory. This step is
used to safely move the memory DIMM(s) from a suspect controller to a new replacement controller.
7) DO NOT troubleshoot a controller problem without parts. If it means you have to wait several hours to get
parts, and then wait. If you can get the array working in the Single Controller Warning state by pulling one
of the controllers’ part way out, then leave the array in this configuration until parts arrive.
8) DO NOT swap memory DIMM(s) from one controller to another if you suspect the problem is memory. You
must keep at least one controller working in the array at all times. If you have to wait for parts, then wait.
Replace the memory on the one controller removed, reinstall and make sure the array comes up to the Ready
state.
9) DO NOT use the vfprecover command unless you are absolutely sure the maps cannot be recovered from
battery backed memory on any other controller. This command along with the online version, armrecover, is
very dangerous. The VA array does not work the same way as the Autoraid Model 12 & 12H. If any portion
of the maps are recoverable but ultimately the command fails, then a corrupt, partial version of the maps will
become the most current. This means the maps are no good and the data is effectively gone anytime the
command is not completely successful. A MULT_FAIL_EH event will likely be found in the logs.
10) DO update firmware and software as soon as possible and keep current.
11) DO NOT install a new disk into an array if a rebuild is in progress.
51
Storage
VA7100
52
Storage
A B
Host FC LED's Indication
Green Amber
off off unit not powered or link down or GBIC not installed
on off normal operation
off on controller fault
on on controller fault
immediate intervention required, GBIC requires attn. or operator
off flashing
induced fault
A B
Controller LED's (on Host) Indication
Green Amber
CONTROLLER
off off unit not powered or link down or GBIC not installed
off on array controller fault
on on array controller fault
off flashing immediate intervention required. check logs or operator induced fault
flashing off array controller self-test in progress or controller I/O activity
BATTERY
off off unit not powered battery not installed
off on module fault
on on module fault
off flashing operator induced manually
flashing off battery self-test in process or battery is powering NVRAM contents
DIMM
A B
A-DIMM 1 B-DIMM 1 Indication
C-DIMM 2 D-DIMM 2
off off unit not powered, battery not installed
off on DIMM fault
on on DIMM fault
off flashing operator induced fault manually
flashing off DIMM self-test in process
Power/Fan Module A B Indication

off off unit not powered on or power/fan module fault
off on module fault
on on module fault
immediate intervention required, (eg, if fan in the same
off flashing module fails, the power supply shuts down for self
preservation), operator induced fault
Disk Drives A B Indication

off off no power or disk drive passed self-test
off on disk drive fault
on on disk drive fault
disk drive problem (eg disk not included in volume set), or
off flashing
operator induced fault with system software
flashing off disk drive self-test in process or I/O activity
System Status A B Indication

off off unit not under power
off on CRU fault. check rear panel for other lit amber LED's
on on CRU fault. check rear panel for other lit amber LED's
a CRU may or may not be faulty. check logs. A CRU my not
off flashing
be faulty if no Quorum or Maps occur.
flashing off array self-test or I/O activity on one or two controllers
Logs: # armdiag -I -if </tmp/armdiag.txt> <array-id> (-I = -i uppercase)
53
Storage
VA7110 – VA7400
54
Storage
VA7410
55
Patches
Chapter 3: Patches
1 Install
UX 10.20 swinstall –x autoreboot=true –x match_target=true –s <depot>

UX 11.00 swinstall –x autoreboot=true –x patch_match_target=true –s <depot>
2 List the contents of a depot
swlist –l fileset –a state –s <path>
3 Copy a depot
swcopy –s <source path> \* @ <destination path>
Without the @ <destination path>

it will copy it to the default directory /var/spool/sw
4 Registration of a depot
swreg –l depot <path of depot>
5 Create a patch depot from multiple individual patches
Creating a SD-UX depot containing 3 different patches

PHxx_A, PHxx_B and PHxx_C:
Retrieve the patches for instance from a patch server or web site:
mkdir /tmp/patches
cd /tmp/patches
Copy patches via ftp, rcp ... into this directory.
Unpack all patches :
sh PHxx_A
sh PHxx_B
sh PHxx_C
This can be done in one operation with the following commands:
# cd /tmp
# for i in PH*
> do
> sh $i
> done
#
56
Patches
Create a patch depot :
mkdir /tmp/patch_depot
swcopy -v -s /tmp/patches/PHxx_A.depot PHxx_A @ /tmp/patch_depot
swcopy -v -s /tmp/patches/PHxx_B.depot PHxx_B @ /tmp/patch_depot
swcopy -v -s /tmp/patches/PHxx_C.depot PHxx_C @ /tmp/patch_depot
This can also be done with the following commands in one operation:
# mkdir /tmp/patch_depot
# for i in PH*.depot
> do
> swcopy -s ${PWD}/$i \* @ /tmp/patch_depot
> done
#
If you need to copy the depot to a tape use the following command :
swpackage -s /tmp/patch_depot -x target_type=tape -d /dev/rmt/0m "*"
6 Patch Clean-up
Useful if there is not enough space for the new patches.
# cleanup –c X (X = number of time the patch is superceded)
example: cleanup –c 2 : the patch is al least 2 times superceded.
57
Servers
Chapter 4: Servers
Name Description Product Code name Op Sys

rx9610 IA64 16 CPUs A6754A Olympic
Superdome enterprise server A5200A
Superdome 32 CPUs A5201A
Superdome
Add on cabinet 32 CPUs A5202A
Superdome 16 CPUs A6113A
rp8400 16 CPUs A6093A Keystone
rp7410 8 CPUs PA8700 650/750/875 MHz 2 partitions A6752A Matterhorn
rp7405 2/4/8 CPUs PA8700 650/750/875 MHz 2 partitions A7111A / A7112A / A7113A Matterhorn
rp7400 N4000 8 CPUs A3639A/B/C Prelude
e3000 N4000 (220MHz) A6449B MPE
e3000 N4000 (330MHz) A6450B MPE
e3000 N4000 (440MHz) A6451B MPE
e3000 N4000 (440MHz *3) A6452B MPE
e3000 N4000 (550MHz *3) A6453B MPE
rx5670 L class IA64 900 MHz / 1 GHz A6837A / A6838A Everest
rp5470 L3000 A/B 4 CPUs A6144A/B Marcato / Cantata
rp5450 L2000 4 CPUs A5191A/B Rhapsody
rp5430 L1500 A6797B Cantata
rp5405 L Class 2 CPUs- 5 PCI / 4 CPUs-10 PCI PA8700 650 / 750 / 875 MHz A7117A / A7118A
rp5400 L1000 2 CPUs A5576A/B Rhapsody
IA64 4 CPUs 733MHz A6153A
rx4610 Ironman
IA64 4 CPUs 800MHz A6447A
rx2600 IA64 1-2 CPUUs 900MHz / 1 GHz A6870A / A6873A Longs peak
rp2470 A500 PA8700 650 / 750 MHz A6890A Harmony
rp2450 A500 1-2 CPUs PA8600 440/550 MHz A5570A/B Piranha
e3000 A500 A6399B MPE
e3000 A400 A6398B MPE
rp2430 A400 1 CPU PA8700 650 MHz A6889A Harmony
rp2405 A400 1 CPU- 2 PCI or 2 CPUs- 4 PCI PA8700 650 MHz A7121A / A7122A
rp2400 A400 1 CPU PA8600 440 MHz A6109A/B Crescendo
A180 / A180C 1 CPU A5582A / A5183A Staccato
T600 Jade
58
Servers
1 K-Class rear view and addressing
59
Servers
60
Servers
2 A L N general
Using the 'GSP' on A, L, & N Class
For revision A GSP, the console path must be assigned to path 0/0/4/0.
For revision B/C GSP, the console path must be assigned to path 0/0/4/1.
Type CTRL+B
At login and password just hit enter or ask login of CU
At the ' GSP>' prompt type one of the following commands
GSP> ac Turns off the automatic display of GSP errors

he Help
ls Displays the current status of the GSP lan
pc Machine can be turned off or on thru the GSP. It can be done remotely also.
ps Gives the status of the power management module and fans
sl To view the system logs. type in E to view the system error logs. See below
ss Shows the status of the processors
tc Transfer of control - resets system & causes the processors to dump their state of the
way down
xd • select I2C - test the support bus that connects the system proc., GSP, &
power monitor
• select R - to reset the bus then the HPUX command
‘stty+resetGSP/GSPdiag1’ or press the GSP reset button. Unplugging the
machine will reset the GSP also.
co Returns machine to console mode
SL logs troubleshooting
Alert Level: How the problem has affected the system operation Alert Levels
0 No failure detected
Source: What major part of the system the alert is referring to (i.e, 1 Service processor alert, no failure detected. Used by the
platform, memory, processor, etc.) SP to perform a specific action, such as changing the
system state from boot start to boot complete
Source Detail: What subpart of the system the alert is referring to 2 Non-urgent operator attention required (i.e, Notification that a
(i.e, cabinet fan, Dimm, high voltage DC power, etc.) previous failed fan is now functioning)
3 System blocked waiting for operator input.
Source ID: Specific FRU referred to in Source & Source Detail (i.e, 4 Unexpected configuration change detected
cabinet fan #4) 5 Boot possible, environmental problem exist
6 Boot possible, pending failure or environmental problem with
Problem Detail: Specific problem info (i.e, power off, functional action required (i.e, power failure, overtemp, etc.)
failure, etc. 7 HPMC detected
8 Boot possible, performance impaired
Timestamp: When the problem occurred. 9 Reserved
10 Boot possible, functionality lost
11 Reserved
12 Software failure (i,e, HPUX panic)
13 System hang detected via timer popping
14 Fatal power or environmental problem prevents operation
15 Fatal hardware or configuration problem prevents operation
SOURCE SOURCE DETAIL SOURCE ID PROBLEM DETAIL FRU ACTION
1 - Processor Processor check processor status,
"in pr" in BCH or "ss"
in GSP
2 - ProcessorCache Processor check proc status
3 - PDH System Board replace sytem board or
run ss_update if new
system board
4 - Power 1 - AC Mains 9 - Power off AC Power check AC supply
4 - Power 3 - Low voltage DC CPU Support CPU Support replace CPU support
power Module # Module module ref. in Source ID
4 - Power 4 - High voltage DC Power supply # Power Supply replace power supply
power ref. in Source ID
61
Servers
6 - Platform 3 - Cabinet fan Fan # Fan replace fan ref. in

Source ID
6 - Platform 6 - Service Proc. Core I/O replace Core I/O
7 - Memory 1 - Controller System Board replace Syst. Brd.
7 - Memory 2 - Carrier Memory Carrier if 1 memory carrier
installed replace it, if
multiple carriers
check BCH "in me" to
isolate
7 - Memory 4 - Simm or Dimm Memory Dimm isolate failing dimm
using "in me" in BCH
8 - I/O 6 - Disk Various Values Disk Subsystem use BCH or ODE to
check disk subsystem
A, L, and N Class front panel LED
LED'S STATE (front panel)

STATE POSSIBLE ACTION
RUN ATTN FAULT
on off off System running normally or no prompt
system hung
on off on Invalid indication
on off flash System crashed & rebooted itself Check chassis code or SP logs
on flash off System interruption that did not take the Check chassis code or Sentinel logs
system down
on flash on Invalid indication
on flash flash System running , unexpected reboot Check chassis code, GSP console logs, ,check
occurred , a non-critical error has been Sentinel logs
detected
flash off off • Executing no OS code-no problem • Check console for responses
detected • System hung, execute TC from GSP, check
• System hung or waiting for BCH response logs
• PDC never executed or HPMC while PDC • Bring system to minimum config &
config. troubleshoot
flash off on Boot failed or executing non-OS code Check chassis codes or read console
messages
flash off flash Unexpected reboot occurred or executing Check chassis code or read console messages
non OS code
flash flash off Executing non-OS code or non-critical error Check chassis codes
occurred
flash flash on Boot failed, Executing non-OS code, non- Check chassis code
critical error
flash flash flash Unexpected reboot/system recovering, Check chassis codes or read console
Executing non-OS code, Non-critical error messages
off off off Check AC power, Turn back power switch on
or try executing PC command from GSP,
cannot execute PDC check for bad processor
or system board
off off on Boot failed, OS not up & running PDC Check chassis codes, read console messages
detected failure
off off flash Invalid indication
off flash off No code executing, Non-critical error Processor or system board, check chassis
occurred codes
off flash on Boot failed, OS not up and running PDC Check chassis codes, Read console messages
detected failure, Non-critical error detected
off flash flash Invalid indication
on Invalid indication
Remote LED- yellow - remote console enabled via modem
Power LED-Green- system power on, flashing-system in standby mode
62
Servers
3 A 400/500 (RP24xx)
A400/500 Rear view
A400/500 internals
A400/A500 IO-paths
Beschrijving Hardware Path

10/100BT Onboard LAN Port 0/0/0/0
External Ultra2 SCSI VHDCI Connector 0/0/1/0
External Ultra SCSI 50Pin HD 0/0/2/0
Top Hot Plug Disk 0/0/1/1.15
Bottom Hot Plug Disk 0/0/2/1.15
GSP External RS232 Ports (Console, Remote, UPS) 0/0/4/0 (new GSP 0/0/4/1)
GSP Internal Port 0/0/5/0 (old GSP only)
PCI Slot 1 (Short Twin Turbo) 0/2/0
PCI Slot 2 (Long Twin Turbo) 0/4/0
PCI Slot 3 (Long Shared) A500 only 0/6/2
PCI Slot 4 (Long Shared) A500 only 0/6/0
63
Servers
4 StormsPeak - rp3410 / rp3440 -- LongsPeak - rx2600
IO Paths /Locations
Slot Path RP Path RX

1 (top) 0/4/1 0/4/1 ACPI(HWP0002,400)/PCI(1|0)
2 0/3/1 0/3/1 ACPI(HWP0002,300)/PCI(1|0)
3 0/2/1 0/2/1 ACPI(HWP0002,200)/PCI(1|0)
4 (botom) 0/6/1 0/5/1 ACPI(HWP0002,600)/PCI(1|0)
Slot 3 and slot 4 are only available for RP3440.
Physical
Description Device
Path
10/100/1000BT Onboard Lan 0/1/2/0 ACPI(HWP0002,100)/PCI(2|0)
10/100 Onboard Lan 0/0/3/0 ACPI(HWP0002,0)/PCI(3|0)
External Ultra320 LVD SCSI VHDCI 0/1/1/1 ACPI(HWP0002,100)/PCI(1|1)
MP External RS232 0/6/1/1 ACPI(HWP0002,700)/PCI(1|1)
MP External RS232 Port UPS 0/6/1/0 ACPI(HWP0002,700)/PCI(1|0)
MP VGA 0/6/2/0 ACPI(HWP0002,700)/PCI(2|0)
IDE Controller for rx2600 0/0/2/0 ACPI(HWP0002,0)/PCI(2|0)/Ata
Onboard Serial A 250/1
Onboard Serial B 250/2
64
Servers
65
Servers
66
Servers
5 L-Class Rear View and Addressing (RP54xx)
The internal disk loading order is:

3 2
path: 0/0/2/0.0.0 path: 0/0/2/0.2.0
4 1
path: 0/0/1/1.0.0 path: 0/0/1/1.2.0
Physical Device Path Physical Device Path Physical Device Path

Slot
L1000 L2000 L3000
12 0/4/0/0.x (Turbo) 0/4/0/0.x (Turbo) 0/10/0/0.x (Twin Turbo)
11 0/7/0/0.x (Turbo) 0/7/0/0.x (Turbo) 0/12/0/0.x (Twin Turbo)
10 0/3/0/0.x (Turbo) 0/3/0/0.x (Turbo) 0/8/0/0.x (Turbo)
7 not available 0/5/0/0.x (Turbo) 0/1/0/0.x (Turbo)
6 not available 0/1/0/0.x (Shared) 0/5/0/0.x (Turbo)
5 not available 0/1/1/0.x (Shared) 0/2/0/0.x (Turbo)
4 not available 0/1/2/0.x (Shared) 0/4/0/0.x (Shared)
3 not available 0/1/3/0.x (Shared) 0/4/2/0.x (Shared)
2 Console Console Console

TOC TOC TOC
GSP Reset GSP Reset GSP Reset
1 LAN/SCSI LAN/SCSI LAN/SCSI

CORE I/O CORE I/O CORE I/O
67
Servers
L1000 – L2000
68
Servers
L1500 - L3000
69
Servers
6 Mt. Hamilton - rp4440 -- Mt. Diablo - rx4640
Slots Path RP Path RX EFI Paths Default Card

0/0/1/0 0/0/1/0 ACPI(HWP0002,0)/PCI(1|0) AUX (port 0) , UPS
0/0/1/1 0/0/1/1 ACPI(HWP0002,0)/PCI(1|1) Console (port 0) , Modem
(port 2)
0/0/2/0 0/0/2/0 ACPI(HWP0002,0)/PCI(2|0) USB Ports
0/0/2/1 0/0/2/1 ACPI(HWP0002,0)/PCI(2|1)
0/0/2/2
0/0/3/0 0/0/3/0 ACPI(HWP0002,0)/PCI(3|0)/Ata IDE Controller
NA 0/0/4/0 ACPI(HWP0002,0)/PCI(4|0) VGA
1 0/1/1 0/1/1 ACPI(HWP0002,100)/PCI(1|0) Ultra 3 SCSI for Internal &
External Connection
2 0/1/2 0/1/2 ACPI(HWP0002,100)/PCI(2|0) Gigabit Lan
3 0/4/1 0/3/1 ACPI(HWP0002,400)/PCI(1|0)
4 0/4/2 0/3/2 ACPI(HWP0002,400)/PCI(2|0)
5 0/5/1 0/4/1 ACPI(HWP0002,500)/PCI(1|0)
6 0/5/2 0/4/2 ACPI(HWP0002,500)/PCI(2|0)
7 0/2/1 0/2/1 ACPI(HWP0002,200)/PCI(1|0)
8 0/6/1 0/5/1 ACPI(HWP0002,600)/PCI(1|0)
70
Servers
Single Bus Config
Device Path EFI-Path

Disk 0 0/1/1/0.0.0 ACPI(HWP0002,100)/PCI(1|0)/SCSI(Pun0,Lun0)
DVD Drive 0/0/3/0.0.0 ACPI(HWP0002,0)/PCI(3:0)/ATA(Primary,Master)
Dual Bus Config
Device Path EFI-Path

DVD Drive 0/0/3/0.0.0 ACPI(HWP0002,0)/PCI(3:0)/ATA(Primary,Master)
71
Servers
7 N-Class Rear View and Addressing (RP7400)
Fru Location
72
Servers
Location/Description Device Type Address Path Note

PCI Slot 1, left I/O backplane PCI I/O card 0/5 -
PCI Slot 7, right I/O backplane PCI I/O card 1/12 -
Internal hot-swap disk (upper) FW-LVD SCSI-Disk 0/0/2/0.6 -
Internal hot-swap disk (lower) FW-LVD SCSI-Disk 0/0/2/1.6 -
FW-LVD SCSI-Adapter, ext. FW-LVD SCSI-compat. Devices 0/0/1/0.x, x=SCSI-Addr. -
10/100BT Interface (prod. LAN) Apropriate LAN cable 0/0/0/0 -
Console RS232 connector System or Web-console 0/0/4/0.0 1)
Console RS232 connector System or Web-console 0/0/4/1.0 2)
UPS/general purpose RS232 RS232 port 0/0/4/0.1 1)
Remote Console RS232 Apropriate Modem 0/0/4/0.2 1)
Remote Console RS232 Apropriate Modem 0/0/4/1.2 2)
Notes:
1) for old (Tosca) GSP
2) for new (Maestro) GSP, firmware < B.02.14
3) for new (Maestro) GSP, firmware >= B.02.14, insf + patches necessary!
73
Servers
74
Servers
8 Matterhorn/keystone MP, BCH commands
MP
Login: Admin
Passwd: Admin
Main menu (MP>):

CO Consoles - Access the consoles of the partitions
VFP Virtual Front Panel - Live display of the major state of partitions and their cells.
CM Command Menu - Change to MP command menu (MP:CM> prompt).
CL Console Logs - view the history of the Console output
SL Show chassis Logs - view the chassis codes in the log history
Activity All chassis codes of level 1 or greater
Error All chassis codes of level 2 or greater
Live Logs Tune into this channel for all chassis codes, live as they happen
HE Help
X Exit Connection
Command menu (MP:CM>):

CP Cells Assigned by Partition-The CP command will display a table of cells
assigned to partitions and arranged by cabinets.
DE Display Entity Status - This command displays the specified entity’s status: B-
BPS, U- CLU, A- CIO, GMPH- PDHC
DF Display FRUID - Displays the FRUID data of the specified FRU: S-
BACKPLANE, G- MP/CIO, P- PCI BACKPLANE, M- MASS STORAGE
BACKPLANE, I- PCI POWER MODULE, B- BPS, C- CELL, H- PDHC
FW Keystone Firmware Update - Use MP> he fw for details.
HE Display Available Commands
LC Configure LAN Connections - Modifies the LAN configuration for the MP LAN.
MA Return to Main menu - Change to MP main menu (MP> prompt).
OSP Online Scan Programming Utility - Use MP:CM> he osp for details
PE Power Entities On or Off - Turns power on/ off to the specified entity: T-CABINET, C-CELL,
P- PCI DOMAIN (0= Right, 1= Left)
PS Power Status- Displays hardware status: T- CABINET, S- SYSTEM
BACKPLANE, G- MP/CIO, P- PCI DOMAIN, C- CELL (cabinet option shows
present, available, and power status of all entities)
PWRGRD Keystone Power Grid - 1 = Grid A, 2 = Grid B, 3 = Grids A+ B Rear View: B1
A1 B0 A0
RR Reset a Partition for Reconfiguration - Resets the specified partition but does not
automatically boot it.
Sets BIB flag.
RS Reset a Partiton- The utility system resets each cell that is a member of the
specified partition.
SYSREV Keystone System Revisions - Report revs of programmable devices and F/ W.
TC Sends a TOC Signal to a Partition -Transfers the control (TOC) of the specified
partition.
XD MP Diagnostics and Reset- Diagnostics and/ or Reset of MP. 1= Parameter
Chksum, 2= Ping, 3= Soft Reset Master MP, 4= Soft Reset Slave MP, 5= Hard
Reset Slave MP
75
Servers
BCH
Main menu:
Boot Continues booting from the specified path
Path Displays or sets the PRI, HAA or ALT boot paths
SEArch Searches for boot devices
ScRoll Displays or sets the scroll mode- Pauses every 24 lines. Turn OFF before
using Error Log or PIM commands.
RECONFIGRESET Reset to Reconfig Complex Profile. Same as MP> RR. Sets “BIB”. Use
after CPU/Cell config!
Config menu:
CEllConfig Config/ Deconfig cell <cell> [ON] [OFF]
CPUconfig Config/ Deconfig cpu [<cell>[< cpu>]] [ON] [OFF]
PathFlags Display/ set Boot
PF <path> = [PRI | HAA | ALT] <value> = [0| 1| 2| 3] 0 = BCH, 1 = boot this
path -> BCH, 2 = boot this path -> next path, 3 = skip -> next path
Information menu:
ALL Display all of the information. [< cell>] Displays processor, PDC, Chip revision, Boot
information, I/O information and Memory.
FwrVersion Displays the firmware version of the various firmware modules in the partition.
Service menu:
Battery Displays the status of the battery on the cell board <cell> = [0 - 3]
DimmDealloc Display or deallocate a dimm on a cell - displays the configuration status of the memory
DIMMs on the cell or schedules them for configuration, or deconfiguration on the next
reboot. <cell> <[ 0A, 0B, 0C, 0D,..., 3A, 3B, 3C, 3D]> <[ ON | OFF]>
PDT Used to display and clear the contents of the page deallocation table. <cell>
CLEAR
PIM Displays the PIM logged on the first, most severe occurrence HPMC, LPMC or a
Transfer of Control (TOC) for any processor present in the partition. <cell>
<cpu> <[ HPMC | LPMC | TOC]>
SCSI Displays/ sets the SCSI controller parameters. SCSI [< path> [INIT| RATE|
TERM| WIDTH|DEFAULT [< id>]]] All parameters are kept in Cell NVRAM –
not complex profile
76
Servers
9 Matterhorn - rp74X0 -- Eiger – rx7620
HPUX Path Device Type Location ACPI Device Path

Cell
0/0/0/0/1 Serial Port MP Acpi(HWP0002,0)/Pci(0|1)
0
Cell
0/0/0/3/0.6.0 Internal disk Slot 0A Acpi(HWP0002,0)/Pci(3|0)/Scsi(Pun6,Lun0)
0
Cell
0/0/0/3/0.5.0 Internal disk Slot 0B Acpi(HWP0002,0)/Pci(3|0)/Scsi(Pun5,Lun0)
0
built in LAN LAN Cell
0/0/8/1/0/4/0 Acpi(HWP0002,40)/Pci(1|0)/Pci(4|0)
1000bt SCSI 0
External SCSI
LAN Cell
0/0/8/1/0/1/0 Fast Wide Acpi(HWP0002,40)/Pci(1|0)/Pci(1|0)
SCSI 0
LVD
Cell
0/0/1 PCI card Slot 8 Acpi(HWP0002,8)
0
Cell
0
Cell
0
Cell
0
Cell
0
Cell
0
Cell
0
Cell
0
Cell
1/0/0/0/1 Serial Port MP Acpi(HWP0002,100)/Pci(0|1)
1
Cell
1/0/1/1/0/1/1.6.0 Internal disk Slot 1B Acpi(HWP0002,108)/Pci(1|0)/PCI(1|1)/Scsi(Pun6,Lun0)
1
1/0/0/3/1.2.0 Internal DVD Cell Acpi(HWP0002,100)/Pci(3|1)/Scsi(Pun2,Lun0)
77
Servers
or DDS 1
Cell
1/0/0/3/0.6.0 Internal disk Slot 1A Acpi(HWP0002,100)/Pci(3|0)/Scsi(Pun6,Lun0)
1
built in LAN LAN Cell
1/0/1/1/0/4/0 Acpi(HWP0002,108)/Pci(1|0)/Pci(4|0)
1000bt SCSI 1
External SCSI
LAN Cell
1/0/1/1/0/1/0 Fast Wide Acpi(HWP0002,108)/Pci(1|0)
SCSI 1
LVD
Cell
1
Cell
1
Cell
1
Cell
1
Cell
1
Cell
1
Cell
1
Cell
1
78
Servers
79
Servers
10 Keystone - rp8400 -- Olympia - rx8620
80
Servers
I/O Path Path Description LBA or Rope # Card Slot #
x/0/0/0/1 Serial Port 0 Core I/O

x/0/0/1/0 10/100/1000BT LAN 0 Core I/O
x/0/0/2/0.6 Internal disk 0 Core I/O
x/0/0/2/1.x Internal DVD or DDS 0 Core I/O
x/0/0/3/0.6 Internal disk 0 Core I/O
x/0/0/3/1 external SCSI FW-LVD 0 Core I/O
x/0/1/x/x PCI Card 2x 33MHZ 1 8
This information is also visible in the blockdiagram.
0 / 0 / 0 / 0 / 0 . x <----- ID (See also figure in Superdome chapter.)

^ ^ ^ ^ ^
| | | | |__ Function
| | | |______ Device
| | | ___________ LBA (Rope)
| |__________________ SBA (always 0)
|______________________ CELL
81
Servers
Address Path Device Type Location Connected to
0/0/0/0/1 Serial (Console/UPS/Remote) Core I/0

0/0/0/1/0 10/100/1000Base-T Built-in LAN Core I/0
0/0/0/2/0.6.0 Internal disk Slot 0C (top middle)
0/0/0/2/1.X.0 Internal DVD (X=2) / DDS (X=3) Slot 0B (top left)
0/0/0/3/0.6.0 Internal disk Slot 0A (top right)
0/0/0/3/1 External SCSI Fast Wide LVD Core I/0
0/0/1/Y PCI (Y=0) / PCI-X (Y=1) [see note 1] Slot 8
Cell 0
0/0/2/Y PCI / PCI-X Slot 7
0/0/6/Y PCI / PCI-X Slot 5 I/O Chassis 0
0/0/8/Y PCI / PCI-X Slot 1 in Server

1/0/0/2/0.6.0 Internal disk Slot 1C (bottom middle)
1/0/0/2/1.X.0 Internal DVD (X=2) / DDS (X=3) Slot 1B (bottom left)
1/0/0/3/0.6.0 Internal disk Slot 1A (bottom right)
1/0/1/Y PCI (Y=0) / PCI-X (Y=1) [see note 1] Slot 8
Cell 1
1/0/6/Y PCI / PCI-X Slot 5 I/O Chassis 1
1/0/8/Y PCI / PCI-X Slot 1 in Server
Note 1: The I/O slot HW pathes change with the new PCI-X I/O chassis from cell#/0/LBA#/0 to
cell#/0/LBA#/1.
Olympia, Rainier and SEU have always PCI-X chassis. Keystone may have PCI or PCI-X.
Note 2: These slots are available only if the optional A6434A Server Expansion Unit (SEU) is
installed. The SEU provides an additional 16 PCI-X I/O slots for hp9000 rp84xx servers or hp
Integrity rx8620 servers. Each cell in the hp9000 rp84xx server or hp Integrity rx8620 server
enables eight IO slots. The server must contain four cells to enable all 16 of the SEU PCI-X IO
slots.
82
Servers

2/0/0/2/0.6.0 Internal disk Slot 2C (top middle)
2/0/0/2/1.X.0 Internal DVD (X=2) / DDS (X=3) Slot 2B (top left)
2/0/0/3/0.6.0 Internal disk Slot 2A (top right)
2/0/1/1 PCI-X Slot 8 Cell 2
2/0/2/1 PCI-X Slot 7 [see note 2]
2/0/4/1 PCI-X Slot 6
I/O Chassis 2
in SEU
2/0/10/1 PCI-X Slot 2
2/0/12/1 PCI-X Slot 3
2/0/14/1 PCI-X Slot 4

3/0/0/2/0.6.0 Internal disk Slot 3C (bottom middle)
3/0/0/2/1.X.0 Internal DVD (X=2) / DDS (X=3) Slot 3B (bottom left)
3/0/0/3/0.6.0 Internal disk Slot 3A (bottom right)
3/0/1/1 PCI-X Slot 8 Cell 3
I/O Chassis 3
in SEU
3/0/10/1 PCI-X Slot 2
3/0/12/1 PCI-X Slot 3
3/0/14/1 PCI-X Slot 4
83
Servers
84
Servers
11 V-CLASS TROUBLESHOOTING
Flashing Light Bar - indicates the presence of an error code and an environmental error detected at power
on. Start with "dcm 0” or "pce_util".
Using dcm 0
At one of the bottom (KSH, lower 2) windows at the test station type:
dcm 0
For bad memory, look for a bank that has been deconfigured, to find dimm ex:
EMB0
======
B0S2 - 64H00 / 64MB the "H" signifies it being "hardware disabled" because of a bad dimm
in the bank
B1S2 - 64S00 / 64MB look for dimm that has been software (64S00) deconfigured, this is
the bad dimm, 128 mb.
B2S6 - 64H00 / 64MB
B3S6 - 64H00 / 64MB
Bad processor ex:

PB5L - SW_DOWN (unknown, unknown) Processor might not show SW_DOWN and
still be bad, check for HPMC.
Using pce_util
pce_util generates an error code "LED" hex code. 00-67 shuts down the system, 68-73 leaves the system
up.
At one of the lower 2 windows type:
pce_util for 2200's
pce_util -n node_number for 2500/2600
Look under the LEDS heading to find the hex error code. ex:
Node Clocks LEDS @C U SHPT Supply 1 Supply 2 Supply 3 Supply 4

------- -------- ------------- ---- ----------- ---------- ---------- ---------- -----------
0 Normal 0xXX 28 1 0000 Nominal Nominal Nominal Nominal
Type: man leds

Look for meaning of displayed error code by scrolling down the page or call backline.
Using EST (Make sure the OS is not running!!!!)

Lower 2 (KSH) windows
cd /spp/est
Type:
est -x 0 Brings up the est GUI. This is for a V2200/2250
est -x node_number for V2500/2600 multinode system
On the GUI interface there are 4 tests that can be run by clicking on them
Ring Test - verifies scan rings are functional
DC Connectivity - checks connectivity between parts
AC Connectivity - confirms that signals move from one part to another in time
Gate Array - tests internal gate array logic, this can take quite awhile.
Checking the hard_hist file for HPMC's
For a v2200/2250
cd /spp/data directory
85
Servers
/spp/data$ ll | more hit the space bar to scroll down to" hard_hist" file to check for last
date and time of HPMC. No current entry could indicate a
software problem.
more or tail -xxx hard_hist file, call backline to translate if nothing obvious is found
V2500/2600 (Multinode)
cd /spp/data directory
/spp/data$ ll | more to find the complex_name (node name)
cd /spp/data/complex_name
more or tail -xxx hard_hist hit the space bar to scroll down to" hard_hist" file to check for last date
and time of HPMC. No current entry could indicate a software problem.
Call backline to translate if nothing obvious is found.
86
Servers
12 Superdome
Blower
Cell with
CPU and
memory
BPS
IO bay
To verify which partition has which cell and IO-bay.

At CM> prompt (GSP>cm)
CP: Display partition cell assignments
IO: Show cell to IO chassis relationship to determine the physical location of IO chassis
87
Servers
88
Servers
I/O Path Path Description LBA or Rope # Card Slot #
x/0/1/x/x Core IO 0 0
x/0/1/x/x 2x PCI card 1 1
This information is also visible in the blockdiagram.
89
Servers
90
Procedures
Chapter 5: Procedures
1 SCSI bus error troubleshooting and log decoding

Identify the suspect SCSI bus in the /var/adm directory or by typing ‘dmseg’.
In syslog.log, OLDsyslog.log and/or dmesg, look for one or more of the following entries:
dev_t=0x1f014000
dev: 1f014000
SCSI: Reset detected – lbolt:8386160, bus: 1 lbolts – bus errors
bp->b_dev: 1f014000 scsi resets – disk errors
Identify the suspect bus by decoding the following information:
0 x 1 f 0 1 4 0 0 0
major# bus# target lun flags
c1 t4 d0
Execute the ioscan –fn command.
From ioscan –fn output, identify the suspect hardware path:
ext_bus 1 8/12 c720 CLAIMED INTERFACE GSC add-

Bus# Fast/Wide SCSI Interface
target 12 8/12.4 tgt CLAIMED DEVICE
disk 12 8/12.4.0 sdisk NO_HW DEVICE SEAGATE
/dev/dsk/c1t4d0 /dev/rdsk/c1t4d0
Record the bus# and the hardware path of the suspect bus:
Then go to EXERCISE A DISK section run diskinfo & dd command
91
Procedures
2 Interpreting the FC hardware path
Fabric Arbitrated loop (FC_AL)
High scalability All nodes share total bandwidth

Switch If one component fails a new initialization is needed
Multiple devices communicate simultaneously Private en public loop possible
Full bandwidth for each switch port - Private loop: all nodes in the loop
can talk to each other but not to
nodes outside the loop
- Public loop: nodes are register with
there WWN. And can talk with
nodes outside the loop
LoopID versus AL_PA

The Arbitrated Loop Physical Address (AL_PA) is used for addressing within a loop. The equivalent to the
AL_PA (10 bit) is the decimal LoopID (8 bit).
The table below shows the relation between LoopID and AL_PA.
The priority grows with increasing LoopID and decreasing AL_PA respectively.
Loop ID +0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +A +B +C +D +E +F
(hex)
00 EF E8 E4 E2 E1 E0 DC DA D9 D6 D5 D4 D3 D2 D1 CE
10 CD CC CB CA C9 C7 C6 C5 C3 BC BA B9 B6 B5 B4 B3
20 B2 B1 AE AD AC AB AA A9 A7 A6 A5 A3 9F 9E 9D 9B
30 98 97 90 8F 88 84 82 81 80 7C 7A 79 76 75 74 73
40 72 71 6E 6D 6C 6B 6A 69 67 66 65 63 5C 5A 59 56
50 55 54 53 52 51 4E 4D 4C 4B 4A 49 47 46 45 43 3C
60 3A 39 36 35 34 33 32 31 2E 2D 2C 2B 2A 29 27 26
70 25 23 1F 1E 1D 1B 18 17 10 0F 08 04 02 01 00 --
Example: determine the AL_PA for LoopID 22 (decimal):

0d22 = 0x16
LoopID can be found in 2nd row (0x10–0x1F)
7th column holds the appropriate AL_PA: 0xC6
92
Procedures
Determine which topology: using fcmsutil /dev/tdX or fcmsutil /dev/fcmsX :

Loop Fabric Topology Fcmsutil output
Yes No Private (arbitrated) loop PRIVATE_LOOP/IN_LOOP
Yes Yes Public loop PUBLIC_LOOP/IN_LOOP_FL
No Yes Switched Point-to-point IN_PTTOPT_FABRIC
Overview of fcmsutil options
T means: option available with Tachyon

TL means: option available with Tachyon TL/TS/XL2.
red font means: destructive task, i.e the communication gets interrupted.
tdutil, tdlist, tddiag

In the directory /opt/fcms/bin you can find additional utilities that might be helpful:
tdutil This is fcmsutil but for td driver only.

tdlist This is a shell script that uses ioscan and tdutil in order to list all devices that are handled by the
td driver, i.e. all devices attached to the Tachyon TL/TS/XL2 adapters on the system. The script
contains a nice function that translates LoopID to AL_PA.
tddiag This is a shell script that gathers the following fibre channel related information
of the system:
system name, system model, system uptime, memory information, mounted
file systems, TachLite version in kernel, system file, patches installed, device
special files for TachLite, ioscan output, tdlist output, running processes,
infos about each /dev/td#:
device info, device vpd info, device topology sensing mode, chip registers,
device statistics, Name Server device statistics, CT Server device statistics, all
remote statistics, name server port info (from kernel), name server port info
(from switch), device statistics on all targets.
93
Procedures
Types used for FC-addressing
PDA: Peripheral Device Addressing

- Less than 8 luns and directly seen (attached)
- FC controllers and Bridges
- Emulation Lun presented by bridge to HPUX (New bridges)
Æ FC-controllers, Galactica, FC10, …
LUA: Logical Unit Addressing

Less then 8 luns and addressed through a legacy FC-SCSI Bridge (4/2)
Æ everything after a legacy bridge
VSA: Volume Set Addressing

Arrays that hve LUN numbers larger than 8
Æ XP, VA, FC60, EMC, EVA, …
Private Loop (no switches) Public/Switched (Fabric)

PDA HBA.8.0.255.Loop.IDhex.L HBA.sw#.port#.255.x.y.L
LUA HBA.8.0.LoopIDdec.B.T.L HBA.sw#.port#.AL_PAdec.B.T.L
VSA HBA.8.0.LoopIDdec.L.U.N HBA.sw#.port#.z.L.U.N
B: Bus
T: Target
L: Lun
L.U.N: lun addressing Æ up to 16k luns possible

N=3bit (4 2 1)
U=4bit (64 32 16 8)
L=7bit (…. 256 128)
Lun number: 128*L + 8*U + U
Sw#: Switch domain number of switch device (not the host) is connected to.
Port#: port number, with some brocade switches port#=port+16
x.y: if loop Æ AL.PAhex

if DFA Æ 0.0
z: if loop Æ AL_PAdec
if DFA Æ 0
Note: Quick Loop (configured at FC switch; QL at switch and NL at target device)

= an emulated private loop configuration that allows private hosts on fabric-attached loops to access
private target devices across the switch. This configuration can consist of multiple private arbitrated loops
that are interconnected and may be isolated by the switch. Typical used as migration path to a fabric
configuration with devices that are not fabric capable.
94
Procedures
3 Fibre card troubleshooting

1. ioscan -kfnd fcT1_cntl Gives hardware path, device file (fcmsX), for each fibre
channel adapter. This for the old fibre channel card A3740-69101
ioscan -kfnd | grep td Gives the hardware path, device file ( tdX ) for
machines with the new fibre channel cards A5158- 69001
2. fcmsutil /dev/fcmsX Shows the status of the of the fibre channel card,
/tdX topology, drive state etc Look for:
Topology = IN_LOOP
Driver state = READY
3. fcmsutil /dev/fcmsX stat gives error statistics on the fibre channel card
/tdX Scroll down and look for these headings. Run this
command multiple times to see if the Counts are
increasing:
Bad Tx Char Count: 100 to 300 is normal
Link Fail Count:
Loss of signal Count:
Loss of Sync Count:
4. fcmsutil /dev/fcmsX reset Resets the fibre channel card. Don’t do a reset on a
/tdX running system.
5. var/adm/syslog/syslog.log File can also be checked. May not show an error

outside of the server
95
Procedures
4 Exercise a disk
dd if =/dev/rdsk/cxtxdx of=/dev/null bs=1024k
or
dd if =/dev/rdsk/cxtxdx of=/dev/null bs=1024k & & causes this command to run in background
tail –f syslog.log to receive ongoing errors from the disk
Ctrl + C to quit this process
Find Error info on a disk using MSTM if no errors are seen the problem could be a patch issue,
controller problem, disk header problem etc.
type mstm
highlight the disk you want to run the test on
tools, expert, run
or
tools, information, run, log tab log, then view.
To obtain disk information

diskinfo -v /dev/rdsk/cxtxdx info includes firmware rev., product ID., disk size, & verifies
disk controller is working
Find a hardware path using the device file

# cd to /usr/sbin
#lssf /dev/rdsk cxtcxdx
96
Procedures
5 Boot machine to single user or maintenance mode

1. Stop machine at Boot Console Handler (PDC)
2. From main menu type in ‘ bo pri ’

and
‘ Y ’ to interact with IPL
ISL>hpux –is For single user mode. Only root and /stand directories
are mounted. Only vg00 is activated. To run commands
such as (ioscan, diskinfo, etc.)
type: mount /usr
ISL> hpux –is (;0)/stand/vmunix.prev To boot to previous kernel in single user mode
ISL> hpux (;0)/stand/vmunix.prev To boot to previous kernel
ISL> hpux –lm Boots to maintenance mode. Root directory is the only
one mounted. No volume group is activated
97
Procedures
6 EMS errors
Errors can be found from the root prompt at : EMS Commands

cd /var/opt/resmon/log From the root prompt type:
tail -xxx event.log file set_fixed -L gives a list of resources currently in
the DOWN state.
check for: example follows
Event data from monitor: To bring a process up type: example
Event Time ……: Wed Mar 21 12:51:40 2001 set_fixed -v storage/status/disks/default/24.2.0
Severity………...:SERIOUS
Monitor…………:dm_core_hw /storage/status…. was taken from the list created by
Event # ………...:51 the set_fixed -L command
System………….:stshpux1
Summary:
I/O power supply has failed
Desciption of Error:
Probable Cause/Recommended Action:
If problem is not obvious call backline to have Event # translated or for event code lookup go to:
http://wojo.rose.hp.com/hp ux/onlinedocs/diag/ems/eme_summ.htm
EMS configuration can be changed at: /etc/opt/resmon/lbin/monconfig
98
Procedures
7 Checking for HPMC’s

Using Boot Admin.
Stop machine at the Boot Admin prompt (PDC).
At prompt, type ‘ser pim’
Record chassis codes for each processor.
D, R, & K can be translated from manual
call back line for A, L, & N translation.
Using MSTM
Type ‘mstm’ at the root prompt.
Move the cursor to the processor and highlight it by pushing the spacebar.
Tab over to ‘Tool’ on the toolbar, push Enter (or Return)
Move the cursor to ‘Information’ and Enter, highlight ‘Run’ and Enter.
Record the chassis codes. Do this for each processor.
You can also find HPMC’s in ‘var/tombstone/ts99’ file if the machine has diagnostics loaded.
99
Procedures
8 Using MSTM diagnostics

1. Type mstm and when information screen comes up press F2 key (OK) to continue.
2. To check Error logs in Logtool Utility;

Hit Tab key and it will highlight File. Type T for Tools, then U for Utility and R for Run.
Use arrow keys and arrow down to logtool. Press spacebar to highlight logtool then F2 for OK.
Hit Tab key then F for File. Type S for Select Raw. Middle of screen shows raw logs. If a log has
.cur at the end (ex. Log1.raw.cur) then this log is currently running. Suggest you should close this
log before viewing by hitting F8 for Cancel, Tab to File, hit F for file, then w for Switch log and
F2. Redo step 2c and continue with step e. Arrow down to the last log or the highest number log.
Press spacebar to highlight that log then hit F2 for OK. It will create a Raw log summary. It will
give you the number of LPMC, LVM, LOGGER event and I/O error entries. Arrow down or hit
Next key for next page. It will show the number of errors on a specific piece of equipment with
the scsi address. After viewing press F8 for done. To get a more detailed explanation of the
errors, press Tab then F and F again for Format Raw. Press F2 and it will then give you a
Formatted Summary log. This should be exactly what you saw in the Raw Summary log. Press F8
for Done. Press Tab then V for View then o for Formatted Log. It will give you precise
information on each error that occurred. This can be very long depending on the number of errors
in this log. The Raw Summary log is usually enough. Press F8 for Done when finished viewing
log.
3. To check for Memory errors in Logtool Utility;

If you are not already in Logtool Utility do steps 2a and 2b. Press Tab then M for Memory then R
for Report Log. Press I to view detail summary or V to view deallocated memory. This will give
the specific bank or pair of memory that has errors. It will give you the date and the number of
errors on that specific memory. It should also tell you what kind of errors occurred on them. If
there are no errors it will inform you of that. When done viewing press F8 for Done.
4. To exit Logtool Utility press F8 then F2 for OK to exit. This puts you back to the main menu in
mstm.
5. To view any information off of a specific scsi address or memory or cpu information; Arrow down
to the desired path, press the spacebar to highlight that address, press Tab then T for Tools then
I for Information and R for Run. This is a great help with a memory problem, cause you can view
the memory size in each pair bank of memory. You will know what size memory we will need to
bring out with us. Be aware that when looking at the memory in each pair bank that is the total
memory and not the actual memory simm size. Example – Memory information says that in 1A/B
the total physical size in (MB) is 16. That means the actual memory simms in each bank are 8mb
simms. Both banks total are 16 but you have two memory simms making up that total. So a rule
of thumb is take half the total and that is the size of the actual memory simm in each bank. Press
F8 for Done to get back to main menu screen in mstm
6. To exit mstm press F8 then F2 for OK to exit Support Tool Manager.
100
Procedures
9 MSTM in single user
1. Boot into single user mode:

main menu>bo
interact with IPL: Y
ISL> hpux –is
2. mount /usr and /var

#/usr/sbin/mount /usr
#/usr/sbin/mount /var
3. type --> hostname <name of system>

#/usr/bin/hostname quality (valid hostname of system)
To find the hostname, type the following before shutting down the system:
just ask the customer
or #/usr/bin/hostname
or #/usr/bin/more /etc/hosts (usually first entry)
4. Change directories to the rc2.d directory:

#cd /sbin/rc2.d
5. Start the following scripts:

#./S008net.init start
#./S320hpether start
#./S340net start
#./S370named start
#./S500inetd start
#./S742diagnostic start
6. Initiate the STM diagnostics by typing the following command:

#/usr/sbin/cstm (Command Line Mode)
or #/usr/sbin/mstm (Menu Mode)
10 Stopping mstm if wrong terminal-emulation
Try before starting mstm if clear command works, if not you have the wrong terminal emulation.
Wait for a moment, type o (not zero)

Wait for a moment, hit tab and type f and x
Wait for a moment, type o (not zero)
Enter and you got the prompt back
101
Procedures
11 Updating firmware from ECU cdrom
See booklet of ECU-cd’s !!!
PDC upgrade
1. Stop machine at Boot Admin (PDC).

2. At Main Menu type sea ipl
3. Find “PX” number of CDROM, then type bo pX
4. Interact with IPL ( Y or N) > y
5. ISL> ls check for ODE
6. ISL > ode
7. ODE > readme
8. README > index
9. Enter \, <index>, h, q > l gives an overview of all PDC files, hardware dependencies,
ODE> (f) update commands
10. Find the update command for the box you are updating ex:
ODE> update ; image cult36342 ; run
11. Write it down. (push the spacebar to scroll to next screen)
12. Enter -, \, q, h, I > q to quit
13. README > exit
14. ODE> update ; image cult36342 ; run <enter> example for Dbox
15. Continue ( [y] /n ) y
16. Box will reboot, PDC firmware updated.
Updating Autoraid Controller Firmware
1. Boot to the ECU CD

2. At the ISL prompt type ODE ARDIAG
3. ARDIAG> isopause off
4. ARDIAG> errpause off
5. ARDIAG> testdisk <primary controller> (either X or Y is the primary controller, can be found
on the front panel)
6. ARDIAG> testlevel <primary controller> ( either X or Y, the primary controller)
7. ARDIAG> download
8. Select a download option ( usually 1)
9. Enter the filename
i. For HP62 the filename is C5447AHP62 ( requires a patch when upgrading )
ii. For HP56 the filename is C5447AHP56
10. Do you want to do the update ? y then enter
102
Procedures
Updating GSP Firmware Flowtchart
Console path for the A/L/N

For revision A GSP, the console path must be assigned to path 0/0/4/0.
For revision B/C GSP, the console path must be assigned to path 0/0/4/1.
103
Procedures
Updating GSP Firmware A-revision (Tosca A6696A)
1. Before you start with the GSP upgrade make sure that you have your laptop available via ftp. IP-addressed in the same
SUBNET
And the right directory were the files are located
example : /tmp/GSP/APPLICATION.ROM
/tmp/GSP/BOOTSTRAP.ROM
/tmp/GSP/upgrade.cnf
2. Login at the console on the system.

NOTE: It's not possible to do the firmware update from the LAN
console port and also not if you are connected to the remote modem port
3. Ping the LAN where the GSP files are located via the XD command item #3 (LAN access (PING)
GSP> XD
Non destructive tests:
1. Parameters checksum
2. I2C access (get Power Monitor status)
3. LAN access (PING)
4. Modem selftests
Type R to reset the GSP or Q to quit the diagnostic menu.

-> Choice: 3
Enter IP Address:
This test must pass to continue.
4. Reset the GSP via the XD command (use the R option) This will free up any unallocated memory for the GSP update.
GSP> XD
Non destructive tests:
1. Parameters checksum
2. I2C access (get Power Monitor status)
3. LAN access (PING)
4. Modem selftests
Type R to reset the GSP or Q to quit the diagnostic menu.

-> Choice: R
GSP> XD
The GSP is now being reset...
5. Login again on the GSP
6. Check the GSP firmware revision.
GSP>he
Hardware Revision 8 Firmware Revision A.xx.xx May 10 1999,12:05:16
……
7. Start the upgrade
GSP>xu
XU
This command activates the upgrade mode. All connections will be closed, the network will be stopped, the session will be
aborted and the modem connection will be dropped.
Please, confirm your intention to activate the upgrade mode (Y/[N]) : y
Enter source system IP address : xxx.xxx.xxx.xxx
104
Procedures
Enter file path :

Do you wish to use the default login : anonymous / [email protected] (Y/[N]):n Enter login :
Enter password :
-> GSP firmware upgrade in progress....

-> GSP firmware upgrade complete - GSP will now reset....
8. Now you will be disconnected from the GSP
9. Connect again and verify the revision
GSP>he
Hardware Revision 8 Firmware Revision A.xx.xx May 10 1999,12:05:16

……
Updating GSP Firmware B/C-revision (Maestro)
If the firmware is B.02.14 (C.02.09) or below, you will have to update to B.02.15 (C.02.10) prior to go
to the latest.
1. Just prior to update, insure that the following operating system patches are installed to change the device drivers. In
addition to the firmware update, these patches are required to resolve the UPS HPMC issue.
11.00
Console Driver Patch PHNE_27788
INSF Patch PHCO_27370
11.11
Console Driver Patch PHNE_27760
INSF Patch PHCO_28169
2. If B.02.14 (C.02.09) or below, you will first need to update to B.02.15 (C.02.10).
The B.02.15 (C.02.10) version requires two firmware update steps. You should reset the GSP with GSP>xd command
after the first updateonly image download. Not completing this step may cause the update in the next step to hang.
Make sure the GSP files are in a known directory, check file sizes too, make sure the names are correct. Also check the
name and extension. You need to specify the file path like as follows.
For Updateonly image -> /tmp/gsp/UpdateOnly

For new firmware image -> /tmp/gsp/B.02.15 (C.02.10)
3. First upgrade FW at directory UpdateOnly like a GSP with A-revision.
4. Reset GSP
5. Upgrade FW at directory B.02.15 (C.02.10) like a GSP with A-revision
6. Upgrade FW at directory B.02.xx (latest) like a GSP with A-revision
105
Procedures
12 SS_CONFIG from Tape or ECU cdrom

1. Insert tape or EURO cdrom and stop machine at Boot Admin.
2. Perform a “sea” at prompt and find “Px” number for tape drive (or EURO cdrom).
3. Boot to the alternate path and type “y” to interact with ISL.
4. List “ls” and type “ss_config” and enter. SSKEY required
5. A, B, & C can be done in any order.
a. “swcap” enter (0 to 15) Usually 0
b. “modelstr” ex. 9000/800/K580
c. “swid” Product and Serial Number
d. type “save” when done
e. type “display” to check entries
13 Diagnostic (STM) install

1. Type swremove scroll down the list to see if there are any bundles associated with previous
versions of diagnostics. New diag install will not work if there are any pointers to previous
versions.
2. Make a directory (mount_directory)
3. Type /sbin/init.d/mwa stop stops measure ware
4. mount –r /dev/dsk/cXtXdX /mount_directory cd /mount_directory
5. swinstall –i –s /mount_directory/DIAGNOSTICS/B.10.20 (or current OS ex.B.11.00)
a. highlight bundle you wish to install
b. Tab to Actions and push Enter
c. Pick ‘Mark For Install’ and Enter
d. Tab to Actions again
e. Pick ‘Install (analysis)’ and Enter
f. If the machine has ‘ready with errors’, Tab to OK and Enter, Let the machine run the
analysis, then do the install. Then highlight the ‘logfile’ and check errors. Write down any
“skipped” patch names.
g. Type /sbin/init.d/mwa start starts measure ware
14 Tar to or untar from a tape

1. ioscan –fnC tape to find the device address
2. tar –cvf /dev/rmt/xm /.profile to write to tape
3. tar -tvf /dev/rmt/xm to read from tape
or
tar -xvf /dev/rmt/xm /filename to extract files from tape and save to a file
If you want to extract files with a fixed path to a path of your choice use the following command:
pax -r -s![path on tape]!.[wanted path]!gp -t -f [tape device]
106
Procedures
15 Service guard
1. /var/adm/cmcluster
default directory for the cluster files
2. View info of cluster

To view information about the current high availability cluster use cmviewcl. It displays the current
status information of a cluster. Output can be displayed for the whole cluster or it may be limited to
particular nodes or packages.
# cmviewcl –v
3. Start cluster
# cmruncl
To run a cluster on a subset of nodes you need to use the –n option of cmruncl. In this case you
have to explicitely confirm that you intent to override Serviceguard’s internal integrity protection.
4. Halt cluster
To halt a high availability cluster use the command cmhaltcl. This causes all nodes in a configured
cluster to stop their cluster daemons, optionally halting all packages or applications in the process.
This command will halt all the daemons on all currently running systems. The -f option causes all
packages to be halted first automatically.
# cmhaltcl –f
5. Start node
Please note that cmrunnode only joins to an existing cluster!
# cmrunnode –n node
6. Halt node
# cmhaltnode –n node
This causes a node to halt its cluster daemon and remove itself from the existing cluster. When
cmhaltnode is run on a node, the cluster daemon is halted and, optionally, all packages that were
running on that node are moved to other nodes if possible.
7. Start package
To run a high availability package use the command cmrunpkg.
# cmrunpkg pkg1
8. Halt package
To halt a high availability package use the command cmhaltpkg.
# cmhaltpkg pkg1
9. Modify a package
# cmmodpkg –e pkg1 Globally enable switching for a package.
# cmmodpkg –d pkg1 Globally disable switching for a package.
# cmmodpkg –n node1 –e pkg1 Enable a node to run a package.
# cmmodpkg –n node1 –d pkg1 Disable a node to run a package.
10. The cmviewconf comand can be used to retrieve the current cluster lock configuration:
# cmviewconf | grep –e "Node name" –e lock
The cluster lock information is backed up and restored with LVM’s vgcfgbackup and vgcfgrestore
commands. If no vgcfgbackup was done after cluster lock initialization then a new cmapplyconf
needs to be done to get this fixed.
107
Procedures
16 Mount a cd
1. ioscan –fn to find device file
2. ls from root to check for SD_CDROM dir
3. mkdir SD_CDROM will create directory if not present
4. mount –r /dev/dsk/cXtXdX /SD_CDROM
5. cd SD_CDROM
6. ls will check files on CDROM
7. umount /dev/dsk/cXtXdX
8. bdf to check to that CD is not mounted
17 Changing instance numbers
Use the following procedure (along with the infile) to change

the instance numbers of device files:
1. Execute the following command:

# ioscan -kf | grep -e INTERFACE -e DEVICE | \
grep -v target | \
awk '{printf "%s %s %s\n",$3,$1,$2}' > infile
2. Change the ext_bus instances numbers as desired.
3. Execute the following commands:
# mv infile /stand/infile
# mv /stand/ioconfig /stand/ioconfig.sav
# mv /etc/ioconfig /etc/ioconfig.sav
# shutdown -ry 0
4. Interrupt the boot process to interact with ISL: boot pri
5. Execute the following command: ISL> hpux –is
6. The system will come to an ioinitrc prompt.
Recreate ioconfig:
# cd /stand
# /sbin/ioinit –c
7. Exit, and continue the boot process by pressing Ctrl-D.
8. Verify that the system is in single-user mode:

# init s
# /sbin/ioinit -f infile –r
9. Once the system reboots, verify that all the instance numbers are correct. It may be necessary to re-import all volume
groups to ensure that /etc/lvmtab contains the correct entries.
108
Procedures
18 Cisco routers
After hooking OMNI book to router then choose icon VT100 SERIAL
User name: cisco
You are now in user mode.
To access privilege mode at the router prompt:
Router> en
Password: san-fran or customer supplied password
Useful commands
• sh int show interface look for: serialx is up line protocol up

operational
serialx is up line protocol down
connection problem
serialx is down line protocol down
interface problem, try the loopback conn.
serialx is administratively down line protocol
down disabled
look for enthernetx also. sh int shows if physical
layer is up or down.
• sh ip route displays routing information
• trace ip looks at route packets take to destination
• ping <IP number> sends a packet to destination and waits for a reply layer 3
• config t accesses global config mode from privilege mode
• sh run looks at the running (or active) file in RAM. If there is a clockrate entered
the interface configured as a DCE. No clockrate is a DTE
• sh start Looks at the backup (or startup) file in NVRAM.
• sh ip protocol shows the status of the protocols. If they are RIP or IGRP
• Telneting to the box verifies the upper 4 layers IOS stack.

• Gateway - address of NIC card that leads to the other subnetworks on the router
• If sh int shows an interface down try resetting the interface.
o Make sure you are in privilege mode.
o Type config t
o Type int sX X is the interface number. Also eX
o Type shutdown
o Type no shutdown
o Try sh int to see if the interface has come up
• All NICS by default are MID. All ports on a HUB are MID-X
109
Procedures
19 DTC16
(underline = flashing)
Normal startup
8. Display being tested
0 Self-test in progress
0 20 sec. To press any key for diagnostics screen via diag. port
0. Download requested
1 Download in progress
1 Download completed OK. Server operational
1. Download completed OK. Server operational and active
Errors
2 Download interrupted. No response from network manager. Auto retry in 30 sec.
2. Download interrupted. Error in data being download. Auto retry in 30 sec.
2. Download interrupted. Error in server. Auto retry in 30 sec.
3 Upload requested, but interrupted. No response from netw. manager. Auto retry in 30 sec
3 Upload in progress
3. Upload requested. Error in protocol. No response from netw manager. Auto retry in 30 s
3. Upload requested. Error in server. No response from netw. manager. Auto retry in 30 sec
5 Fatal error in communication. Access diagnostics screen.
6 Reprogramming in progress
6 Fatal error in reprogramming. Server is dead
6. Non-fatal error in programming. You must retry reprogramming from start.
7 Error on port 0-7

7. Error on port 8-15
7. Error on port 16-23
9 Fatal error on board. If lan board or DTC , server is dead

9. Error on LAN connection. Auto restart from self-tests after 30 sec.
9. Error in network address. Access diag. Port
Enter the exchange DTC16 machine type

* connect a terminal to the exchange DTC16 offline diagnostic port using the offline diagnostic cable
* Switch off the DTC16
* Press a key when prompted to do so on the terminal screen
* Press "P" to enter in the machine type program screen.
* Press "TN" if you are replacing a faulty DTC 16TN or press "iX" if you are replacing a faulty DTC 16iX.
NOTE: Once the machine type has been confirmed, you cannot change it.
If you made an error and selected a wrong machine type, you will have to use a second
exchange DTC16.
Change the DTC16 LAN address

* Reprogram the LAN address of the exchanged DTC 16TN with the original LAN
address using the offline diagnostic screens.
* Note the original LAN address on the exchanged DTC16 in the free reserved space
under the factory LAN address.
DEFAULT PASSWORD: AVESTA
110
Procedures
20 4 Channel ULTRA 2 SCSI RAID 4SI
Information
# irdisplay
Configuration
WARNING:
#irdiag –v is used to get configuration information but in case of hard problem on disks if irdiag –v is
executed a 2nd time, configuration can be lost and must be recreated.
To avoid this problem, enter

#irm (instead of #irdiag –v) This command is menu driver.
KNOW ISSUE
After a complete system shutdown and power OFF the physical drives connected to the RAID CONTROLLER
could be in status 'FAIL' or 'OFFLINE'.
Start #irm (Press Any Key To Continue...)
--Management Menu-- Choose the item with explaination:

| Configure | Configure Objects like Adapter Parameters etc...
| Objects |
| Rebuild |
| Check Consistency |
| Advanced Menu |
|___________________|
----Objects----- Choose the item with explaination:

--Managemen| Adapter | Change Physical Drive Parameters
| Configure | Logical Drive |
| Objects | Physical Drive |
| Rebuild | SCSI Channel |
| Check Cons| Battery Backup |
| Advanced M|________________|
|___________________|
Select the disk who is offline and press enter. And make it online.
------------------Objects - PHYSICAL DRIVE SELECTION MENU-------------------

| |
| Channel-0 Channel-1 Channel-2 Channel-3 |
|ID.============. x ID.============. x ID.============. x ID.============. x |
||0 :* ONL A00-00: + 0 :* ONL A00-01: 0 :* ONL A00-02: 0 :* ONL A00-03: |
|| :============: :============: :============: :============: |
||1 :* ONL A01-00: 1 :* ONL A01-01: 1 :* ONL A01-02: 1 :* ONL A01-03: |
|| :=== --Channel-0, Target-0-- ===: :============: :============: |
||2 :* O| Rebuild |-01: 2 :* ONL A02-02: 2 :* ONL A02-03: |
|| :===| Make Online |===: :============: :============: |
|3 :* | Fail Drive | : 3 :* : 3 :* : |
| :===| Make HotSpare |===: :============: :============: |
|4 :* | Properties | : 4 :* : 4 :* : |
| :===| View Rebuild Progress |===: :============: :============: |
|5 :* |_______________________| : 5 :* : 5 :* : |
| :============: :============: :============: :============: |
|6 :* : 6 :* : 6 :* : 6 :* : |
| :============: v :============: v :============: v :============: v |
|____________________________________________________________________________|
111
Procedures
21 Optical Boxes With Numbered Config and Test Choices

Config Menu
Config 16 - restores default configuration,
** Write down the CU’s drive and AC controller SCSI ID’s before performing this restore, and reenter
them after the test.**
200T, 120T
After replacing the AC controller (on power up it will give a "fail 1") go into, Options→ Config 16→
enter→ when it says "save", hit next until it says "clear"→ enter, power machine off, then back on, then
run test 10 (initial element status). Reenter scsi ID's.
Test Menu
2 Wellness - one loaded cartridge, drive and mailslot empty, one empty slot
5 Init Mech
10 Initialize Element status
11 Mechanism Exercise - One loaded cartridge, drive and mailslot empty, one empty slot
14 Flip Test
15 Storage Slot - one loaded cartridge, one empty slot
16 Drive I/O – One loaded cartridge, drive must be empty
17 Mailslot I/O – One loaded cartridge, mailslot must be empty
25 Empty picker - ** Caution CU will have to perform a reinventory **
43 Mailslot sensor
44 Clear path
112
Procedures
22 ICOD
iCOD Commands
At the root prompt type

icod_modify -c updates customer information
icod_modify -m turn the customers processors on or off
icod_stat shows icod information, processors on or off etc.
icod_ notify sends test message to HP's iCOD node outside cu's domain
iCOD Install ( you must use the DEC. 2000 or later Support CD )
1. Download latest PDC

2. Download Diags
3. Download the iCOD software
4. icod_modify -c input the cu info
5. Go to this website http://newshad.rsn.hp.com/icoda/support/index.html to obtain the customer number of the iCOD
processors
Logon: fsuser password: icod4.u (this is not the iCOD password)
6. icod_modify -m input the processor information
a. you need the iCOD password to make changes
b. input your information: name, phone number etc.
7. The system must be rebooted for the processor change to take affect. Do a icod_stat to insure
that the changes took effect.
It's the CU's responsibility to turn the processors on, CE's to turn them off.
113
Procedures
23 Partitioning
vPar Commands
The monitor is called vpmon and is stored in /stand/vpmon on each partition.
You can use Ctrl-A to cycle between all live vPars
The daemon vpard is started with the script /sbin/init.d/vpard. It will synchronize the master database and
the local disk database for all “up” state vPars,
The daemon vphbd is started with the script /sbin/init.d/vparhb. It provides a heartbeat status, written to
the local disk.
Once the monitor is booted, the database selected (/stand/vpdb is default) is copied into memory and
becomes the master copy.
ISL> hpux /stand/vpmon

ISL> hpux /stand/vpmon vparload -p vpar1
Commands at MON> prompt
vparload boots or loads a vPar from the monitor.

MON> vparload -all
MON> vparload -auto
MON> vparload -p partition_name [-b kernelpath] [-o boot_options] [-B hardware_path]
Boots the virtual partition partition_name;
this command is similar to the vPars Unix shell command vparboot.
-all boots all virtual partitions, regardless of the autoboot attribute
-auto boots all virtual partitions that have their autoboot attribute flag set to
AUTO
-b kernelpath boots the virtual partition using the kernel kernelpath, such
as /stand/vmunix.prev, instead of the default kernel /stand/vmunix
-o boot_options boots the virtual partition using the options boot_options,
such as -is for single-user mode or -lm for LVM maintenance mode
-B hardware_path boots the virtual partition using the disk device at the
hardware_path
114
Procedures
bootpath Shows the bootpath from which vpmon was loaded

reboot reboots the monitor and the whole server like CTRL-B, RS !!!
help or ? displays the list of monitor commands
getauto returns the contents of the AUTO file within the LIF volume
log displays the contents of the monitor log
ls will list the files in a directory. The default directory is “/stand”
scan displays the hardware found by the monitor
vparinfo displays information about a vPar, or, about unassigned hardware
Commands at HP-UX prompt
vparcreate Create a new virtual partition

vparboot Boot a virtual partition same options as vparload at MON>
vparmodify Modify the attributes of a virtual partition
vpardump Manages monitor dump files
vparextract Extracts memory images from a running virtual partition system
vparreloc Relocates the load address of a vmunix file, determines if a vmunix file is relocatable
and promotes the scope of symbols in a relocatable vmunix file
vparremove Remove a virtual partition
vparreset Reset a virtual partition
vparstatus Display information about one or more virtual partitions on a system. If Monitor is not
loaded the vparstatus shows this with an message. The output shows also the state of a
vPar
State Description
Load A vPar is loading/booting
Boot A vPar is loaded into memory
Up A vPar is up and running.
Shut A vPar is shutting down.
Down A vPar is down.
Crash A vPar is crashing (panic - HPMC, TOC)
Hung A vPar is not responding
vparutil Gets and sets SCSI
Changing a bootdisk in a vPar
On a non-vPars computer, the setboot command allows you to read from and write
to the system-wide stable storage of non-volatile memory.
However, on a vPars computer, the setboot command does not affect the stable storage. Instead, it reads
from and writes to only the partition database.
The mkboot -a command will no longer need to be used to set boot options for the HP-UX kernel. These will
now need to be specified as part of the configuration of the virtual partition. Instead, mkboot -a needs to
be used to change the AUTO LIF file to get /stand/vpmon loaded instead of the regular HP-UX kernel.
The system’s boot command needs to be changed to boot “/stand/vpmon -a”, rather than the normal HP-
UX kernel.
The “mkboot -a” command is used to change the boot command in the LIF AUTO file.
mkboot -a “/stand/vpmon -a” /dev/rdsk/cXtYdZ (the device file of your boot disk)
The “-a” option to vpmon tells it to start all the partitions with the auto boot flag set, normally all virtual
partitions.
115
Procedures
nPar Commands (Hardpartitioning)
parstatus displays information about the nPartitions or hardware within a server complex
-s Indicate whether the system is an HP server that supports nPartitions. Not supported with the
-g... -h... options.
-w Display the nPartition number for the local nPartition. Not supported with the -g... -h...
options.
-X Display the server complex attributes.
-A Only display the available resources in the complex.
-V Increase the amount of information displayed.
-M Produce output suitable for machine parsing.
-C Show information for all the cells in the complex.
-I Show information for all I/O chassis in the complex.
-B Show information for all cabinets in the complex.
-P Show information for all nPartitions in the complex.
-c cell Show information about the specified cell.
-i IOchassis Show information about the specified I/O chassis.
-b cabinet Show information about the specified cabinet.
-p partition Show information about the specified nPartition.
parcreate This command assigns the specified cells (and any attached I/O chassis) to an nPartition,
assigns a number to the new nPartition, and returns the partition number of the newly
created nPartition.
parmodify You can use the parmodify command to modify the following attributes of an existing
nPartition:
- Partition name
- Cell assignments (add cells or remove cells)
- Attributes of existing cells: Cell type, Use-on-next-boot value, Failure usage, HP-UX B.11.23
and Windows SMS only: Cell local memory (CLM) value
- Core cell choices
- Boot paths (the primary, alternate, and HA alternate paths)
- IP address (if set, this value should be consistent with the IP address assigned to the
nPartition when HP-UX is booted)
parremove This command removes an existing nPartition. This removes all cells from the nPartition and
destroys the nPartition definition.
parunlock The parunlock command unlocks the Stable Complex Configuration Data or Partition
Configuration Data.
fruled The fruled command blinks hardware attention indicators (LEDs) or turns them off.
frupower This command turns on, turns off, or displays the current status of power for cells and I/O
chassis in nPartition servers.
cplxmodify This command modifies attributes of an nPartition-capable server complex, such as the
complex name. Only HP-UX B.11.23 and Windows SMS.
To verify which partition has which cell and IO-bay.

At CM> prompt (GSP>cm)
CP: Display partition cell assignments
IO: Show cell to IO chassis relationship to determine the physical location of IO chassis
Only for superdome. On other systems this is fixed.
116
Procedures
Chapter 6: HPUX Command Summary for CE’s
The following is a summary of the most useful HPUX commands and file locations for the typical HP CE.
1 General HPUX Commands
man tar - Man page listing for the “tar” command

catman - Catalogue the man pages so the man –k will work.
model - Returns the full model string. i.e. 9000/879/K260
uname -a - Display information about the computer system.
hostname - Displays the Hostname of the system. also can use uname -a
ioscan -fn - Lists all I/O devices.
ioscan -fnC disk - Lists only I/O devices with a class of “disk”
lsdev - List the device drivers in the system
diskinfo -v /dev/rdsk/c0t6d0 - Displays disk information. An invalid disk size indicates a bad disk.
lssf /dev/dsk/c0t6d0 - List Special file. Displays physical path info of device file.
strings /etc/lvmtab - Extracts ASCII characters only from a binary file.
more /etc/fstab - Lists an ASCII file one screen at a time. cntl f =forward, cntl b = back
cat /etc/fstab - Lists an ASCII file.
tail -n 25 syslog.log - Display the last 25 lines of a file (i.e. syslog.log)
grep pattern file1 - Display only the lines which match ‘pattern’ in file1
pvdisplay -v /dev/dsk/c0t0d0 - Lists Physical Volume information about a disk.

vgdisplay -v /dev/vg00 - Lists Volume Group information.
lvdisplay -v /dev/vg00/lvol1 - Lists Logical Volume information.
shutdown -hy 0 - Shutdown and halt the system

shutdown -ry 0 - Shutdown and reboot the system
shutdown - Bring the System into single user mode
init 4 - Change the current run level to 4
dmesg - Display the diagnostic message log
ps –ef - Display All processes.
kill –9 13252 - Kill the process with the process ID of 13252.
fuser –u /dev/vg00/lvol6 - Show processes ID’s and user names using the device lvol6.
fuser –u /etc/passwd - Show processes ID’s and user names using the /etc/passwd file.
uptime - System uptime since it was last booted
bdf - Displays free space on all mounted filesystems
mount - Mount a file system
mount -r /dev/dsk/c6t2d0 /cdrom -Mounts a cdrom to the /cdrom directory
swapinfo –m - Display system swap information
who - Display all users on the system
whoami - Display your session information
who -r - Display the run level
find / -name syslog.log - Find a file. (i.e. find / -name syslog.log)
whereis psconfig - Lists locations of psconfig executable. (only works on executable programs defined in the current
PATH variable)
pwd - Display the present working directory
cd /etc - Change directory to /etc
l,ll ls,lsf - List files
file - Determine the file type
mv file1 file2 - Rename file1 to file2
cp file1 file2 - Copy file1 to file2. file1 is not changed.
chmod 777 file1 - Change file permissions of file1 to 777
chown root:bin file1 - Change the owner and group of file1
rm file1 - Remove or Delete a file from the system
rm –r file1 - Remove a file or subdirectory and all the files underneath.
(BE VERY CAREFUL as this can be DANGEROUS! )
lifls -l /dev/rdsk/c0t6d0 - List Contents of a lif directory
lvlnboot –v - List the boot definitions for vg00. (shows disk, HPUX was booted from)
setboot - Display the current Boot and Alternate paths for your System.
setboot -a 0/0/2/0.2.0 - Set the Alternate Boot path to 0/0/2/0.2.0 without shutting down system.
-p Sets the Primary, -s on | off Sets autosearch, -b on | off Sets autoboot
mkboot -a "hpux -lq (;0)/stand/vmunix" /dev/rdsk/c2t2d0 -Make the specified disk bootable
117
HPUX
fcmsutil /dev/fcms0 - Fiber channel diagnostic and status

top - Performance tool. Cpu Utilization, etc…
su - Become Superuser
cu –ltty2a6 - HPUX terminal emulation. Connect to port tty2a6. Type ~. to exit
/usr/sbin/diag/contrib/pdcinfo –nobanner - Reads the pim dump info on the system bd.
Normally runs at system startup and sends output to ts99 file.
/etc/opt/resmon/lbin/monconfig - Run (EMS) Event Monitoring Request Manager

/sbin/init.d/ems start - To Start the EMS daemon
/sbin/init.d/diagnostic start - Start the Diagnostics daemon (diagmond)

/sbin/init.d/predictive start - Start the Predictive Support daemon (psmon)
/opt/pred/bin/psconfig - Run Predictive Support
/opt/ignite/bin/print_manifest - Displays system information (must have ignite UX installed!)
dd if=/dev/rdsk/c0t6d0 of=/dev/null bs=512k - Light the LED on a Disk drive.

disk test…ctl c will stop. dead disk won't work, hang or io error disk def.
dd if=/dev/rdsk/c0t6d0 of=/dev/rdsk/c0t5d0 bs=2048 - Copy entire disk to a second disk.

Disks must be same size. Copy from c0t6d0 to c0t5d0.
tar -cvf /dev/rmt/0m /etc/hosts - Copy the file /etc/hosts to tape

tar -xvf /dev/rmt/0m /etc/hosts - eXtract a file /etc/hosts from tape back to disk
tar –tvf /dev/rmt/0m - List the files on the tape
mt –f /dev/rmt/0m status - Show model/description of Tape device (i.e. HP C1537A)
mt –f /dev/rmt/0m offl - Rewind a tape
swlist –l bundle - show bundles only

swlist –l fileset - show products, subproducts, and filesets
swlist –l patch - shows all applied patches
swlist – l fileset –a state - shows the state of the fileset
Important files:
/etc/passwd - Password file

/etc/inittab - boot init script.
/var/adm/syslog/syslog.log - System log file. error messages, etc…
/var/tombstones/ts99 - tombstone log file. ts99 most recent, ts98 older
/var/adm/shutdownlog - Log of System shutdowns
/etc/rc.log - System Startup log file
How to Change your terminal emulation
set - Display the users enviroment variables. i.e. PATH, terminal

emulation (vt100 or hp)
TERM=hp or TERM=vt100 - Terminal you want to emulate (hp or vt100)
export TERM - Export the TERM variable
tset - Initialize the terminal settings
Now change the 700/96 or the webconsole settings to match the emulation you just set.
TERM=xterm - Set the TERM variable (terminal emulation)

DISPLAY=15.32.149.186:0.0 - Set the DISPLAY variable for your local workstation
export DISPLAY - export the DISPLAY variable
xhosts + - Allow access to your system.
/usr/bin/X11/xterm –fn 6x13 –sb –sl 10000 –ls –display %IP#% -name %T% &
- Set the DISPLAY variable when running Reflection-X
118
Procedures
HPUX Boot Commands
hpux (10/0.6;0) /stand/vmunix - Boot from vmunix kernel on disk at 10/0.6.0

hpux /stand/vmunix.prev - Boot from vmunix.prev kernel
hpux –is - boot to single user mode
hpux –lm - boot to single user LVM Maintenance mode
hpux ls - Lists potential kernel files to boot from
hpux –lq - boot ignoring quorum requirement
2 LVM Commands
pvdisplay -v /dev/dsk/c0t0d0 - Lists Physical Volume information about a disk.

vgdisplay -v /dev/vg00 - Lists Volume Group information.
lvdisplay -v /dev/vg00/lvol1 - Lists Logical Volume information.
LVM Commands are found in the /etc directory.
pvchange - allow/disable disk allocation for LV's.

pvcreate - load a disk with LVM structures, preparing for use.
pvdisplay - describe disks use with respect to LVM.
pvmove - move physical extents to other disk(s)
Related to VG's - volume groups.
vgcfgbackup - SAVE the LVM structures for a specific VG into a file.
vgcfgrestore - RESTORE the LVM structures previously saved by the vgcfgbackup command to a specific disk in
trouble.
vgchange - change or activate the characteristics of a VG.
vgcreate - create a new volume group.
vgdisplay - describe a VG's LVM characteristics.
vgexport - disassociate a VG from the system - can be associated to this or another system later.
vgextend - adopt a disk into a VG. Can also create alternate links to disk(PVLINKS)
vgimport - adopt a in existing orphaned volume group into the system.
vgreduce - divorce a disk from a volume group.
vgremove - remove a volume group from the system.
vgscan - Useful command - looks at all disks and lists their LVM associations, rebuilds /etc/lvmtab if desired.
vgsync - resyncs stale mirrored extents . Related to LV's - logical volumes.
lvchange - change or activate an LV's characteristics.
lvcreate - create an LV.
lvdisplay - describe the LVM characteristics about an LV.
lvextend (followed by extendfs) - add mirrors or increase an LV's size.
lvmerge - merge two LV's of same size and resync extents
lvreduce - shrink an LV - BACKUP data 1st and newfs after
lvremove - eliminate a Logical volume.
lvmmigrate - used orig. to migrate to LVM – recommend reinstall.
lvrmboot - remove root/swap/dumps definitions from BDRA.
lvsplit - split a logical volume mirror into distinct LV's.
lvsync - synchronize mirrors of a mirrored lvol.
lvlnboot - prepare a LV to be a root,swap, or dump device.

lvlnboot –v Display current boot partitions for Boot, Root, Swap, Dump.
setboot - List the current Boot and Alternate paths for your System.
lifls -l /dev/rdsk/c0t6d0 - List Contents of a lif directory at the specified disk.
mkboot -a "hpux -lq (;0)/stand/vmunix" /dev/rdsk/c2t2d0 -Make the specified disk bootable
setboot -a 0/0/2/0.2.0 - Set the Alternate Boot path to 0/0/2/0.2.0 without shutting down system. -p Sets the
Primary, -s on | off Sets autosearch, -b on | off Sets autoboot
/dev/vg00 - This, as well as other directories naming volume

groups you may have created, contain special files
pointing to each lvol in the volume group. You CAN
rename an lvol in these directories, so long as you
reference the new name in /etc/checklist.
119
HPUX
3 LVM FILES and DIRECTORIES
LVM related information is kept in various places on a system.

In the file system:
/etc/lvmtab - This binary file is used to activate volume

groups. It contains the association of VG's to
PV's. That is, which volume groups own which
physical volumes. If this file is out of date (from
a restore etc), then activation of volume groups at
boot or other time could error out.
List the contents of the lvmtab file using the strings command.
i.e. strings /etc/lvmtab
/etc/lvmconf - This directory contains an lvm_lock flag file

used by lvm. More important to you, it is also
where your systems' LVM configuration files are
placed when you do vgcfgbackup's.
/etc/lvmrc - This script is used at bootup to activate all

volume groups. You normally don't use it.
/etc/lvmpvg - This ACSII file (if existing) would contain a list

of disk associations you have created. See
lvmpvg(4). PVG: physical volume group - like disks
/etc/fstab – Filesystems are mounted at boot time from entries in this file.
/etc/mnttab – File contains currently mounted filesystems.
120
Procedures
4 Networking Commands
lanscan - Display LAN device configuration and Status.

ifconfig lan0 (/etc) - Display lan card IP Address or Configure Network interface.
ifconfig lan0 up - Changes the status of a down lan card to ‘up”.
linkloop - Tests LAN connectivity at the link level. Uses MAC hardware address.
i.e. /etc/linkloop -n5 -f /dev/lan0 -v 0x080009123456
ping 15.32.122.12 - Tests LAN connectivity at the IP level.
telnet 15.32.122.12 - Connect to a networked system using telnet protocol.
arp -a - Displays the IP to MAC translation table in cache.
lanadmin - Local area network administration program.
- Display LAN card statistics. Reset the LAN interface.
netstat -i - Shows the state of network interfaces.
netstat -s - Shows statistics for all protocols.
Route (/usr/sbin) - Make entries to the route tables
/sbin/set_parms - Initial networking and host name setup
Important files:
/etc/rc.config.d/netconf - Network configuration file

/etc/hosts - Hosts file
/etc/resolv.conf - defines which domain to use when short host names are used.
(i.e. hostname is mapped to hostname.esr.hp.com)
5 Software/Patches
sh filename - Unshar (uncompress) a file.

swremove - Remove software/patches installed with swinstall.
swinstall - Install software/patches.
swlist -l product - Lists installed software products. (Displays installed patches)
Important files:
/var/adm/sw/swagent.log - software installation log file. Check after swinstall for errors!
/var/adm/sw/swremove.log - command log file for swremove
/var/adm/sw/swinstall.log - command log file for swinstall
PHCO - general patches

PHKL - kernal
PHSS - all other
PHNE – network
121
HPUX
6 FTP commands
ascii to set the mode of file transfer to ASCII

binary to set the mode of file transfer to binary
bye to exit the FTP environment (same as quit)
cd to change directory on the remote machine
close to terminate a connection with another computer
delete to delete (remove) a file in the current remote directory (same as rm in UNIX)
get to copy one file from the remote machine to the local machine
help to request a list of all available FTP commands
lcd to change directory on your local machine (same as UNIX cd)
ls to list the names of the files in the current remote directory
mkdir to make a new directory within the current remote directory
mget to copy multiple files from the remote machine to the local machine
mput to copy multiple files from the local machine to the remote machine
open to open a connection with another computer
put to copy one file from the local machine to the remote machine
pwd to find out the pathname of the current directory on the remote machine
quit to exit the FTP environment (same as bye)
rmdir to to remove (delete) a directory in the current remote directory
7 Vi Commands
vi file Starting Vi Session
Cursor control commands in vi

h Move left one character
j Move down one line
k Move up one line
i Move right one character
Adding text in vi
a append new text after the cursor
i insert new text before the cursor
o open a line below the current line and insert
esc Get back to command mode
Deleting text in vi
x Delete the character at the cursor
dd Delete the entire line
Changing Text in vi
rchar Replace the current character with char
Copying in vi
yy Yank the current line
Undo in vi
u Undo the last change
U Undo all changes to the current line
122

Ce 6 Cookbook

Uploaded by

Copyright:

Available Formats

Ce 6 Cookbook

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ce 6 Cookbook

Uploaded by

Copyright:

Available Formats

CE cookbook

Table of contents ....................................................................................2

Chapter 1: LVM and disk replacement ..................................................6

Chapter 3: Patches .............................................................................56

Chapter 4: Servers .............................................................................58

VSA: Volume Set Addressing.................................................................................................94

Chapter 6: HPUX Command Summary for CE’s .................................117

Chapter 1: LVM and disk replacement

1 Gathering information about your LVM configuration.

The hardware path and special device file.

The disktype, example. C2490WD, ST32550W

The volume group the disk belongs to.

Also, check if the mirror has offline diagnostics.

Check how the boot disk booted in normal way.

2 Main troubleshooting flowchart.

Information from Software Recovery Handbook, chapter 16 LVM. http://bali.grc.hp.com/scunix/sw_recovery/

3 Replacing FC disk + FC-controllers

BUT When you replace a FC disk or FC controller please be aware of…

THIS IS NEEDED IN CASE IOSCAN GIVES NO_HW.

# fcmsutil /dev/tdX replace_dsk <remote_N-port_ID>

1. Take out defective disk, if you know N-port_ID or Loop_ID go to step 3

2. Lookup the N-port_ID or Loop_ID. (also check FC-addressing or storage section)

a. # fcmsutil /dev/tdX devstat all | grep loop

b. Try to login into each device

3. Run replace_dsk using the remote-N-port_ID or Loop_ID

4. Insert new disk

4 Hot-Swap Procedure for Attached Physical Volumes

# lvreduce -m 0 -A n /dev/vgXX/lvolX <list of PVs> (for 1 way mirroring)

To check the progress of the synchronization you could use:

5 Hot-Swap Procedure for Unattached Physical Volumes

To check the progress of the synchronization you could use:

6 LIF/BDRA Configuration Procedure

2. Write content of AUTO File: (if autoboot is desired)

3. Install ODE files (may be skipped):

7 Itanium mirror bootdisk: Procedure for Attached Physical Volumes

2. Replace the faulty disk. Verify with an ioscan.

Create a partition description file:

at UX 11.22 at UX 11.23 (cold-installed)

4. Clear the previously partition configuration

5. Create the disk partitions

Use idisk command to partition the disk according to this file:

7. Verify the partitiontable

8. Use efi_fsinit to initialize the FAT filesystem on the EFI partition:

# efi_cp -d /dev/rdsk/cXtYdZs1 -u /EFI/HPUX/AUTO /tmp/prim

# efi_cp -d /dev/rdsk/cXtYdZs1 -u /EFI/HPUX/AUTO /tmp/mirror

# dd if=/dev/rdsk/cAtBdCs3 of=/dev/rdsk/cXtYdZs3 bs=1024k

13. Now the disk is configured as boot devices.

16. Extend the mirrors back onto the replaced disk.

8 Itanium mirror bootdisk: Procedure for UnAttached Physical Volumes

1. Replace the faulty disk. Verify with an ioscan.

Create a partition description file:

3. at UX 11.22 4. at UX 11.23 (cold-installed)

7. Clear the previously partition configuration

8. Create the disk partitions

Use idisk command to partition the disk according to this file:

10. Verify the partitiontable

# efi_ls -d /dev/rdsk/cXtYdZs1 (to check EFI)

# efi_cp -d /dev/rdsk/cXtYdZs1 -u /EFI/HPUX/AUTO /tmp/prim

# efi_cp -d /dev/rdsk/cXtYdZs1 -u /EFI/HPUX/AUTO /tmp/mirror

# dd if=/dev/rdsk/cAtBdCs3 of=/dev/rdsk/cXtYdZs3 bs=1024k

Logical Volume Creation

Create a RAW partition:

**Create a MIRRORED set*