HPUX :'scsimgr replace_wwid' command to validate the change

Disk are not visible while initial check, disk are not coming xpinfo and ioscan. Follow the below action to recommended.

 

Below are the error :

class : lunpath, instance 1215

Evpd inquiry page 83h/80h failed or the current page 83h/80h data do not match the previous known page 83h/80h data on LUN id 0x0 probed beneath the target path (class = tgtpath, instance = 20) The lun path is (class = lunpath, instance 1215).Run 'scsimgr replace_wwid' command to validate the change

 

class : lunpath, instance 1215

An attempt to probe existing LUN id 0x40b7000000000000 failed with errno of 14.

 

class : lunpath, instance 1192

Evpd inquiry page 83h/80h failed or the current page 83h/80h data do not match the previous known page 83h/80h data on LUN id 0x0 probed beneath the target path (class = tgtpath, instance = 20) The lun path is (class = lunpath, instance 1192).Run 'scsimgr replace_wwid' command to validate the change

 

class : lunpath, instance 1192

An attempt to probe existing LUN id 0x40cb000000000000 failed with errno of 14.

 

#scsimgr -f replace_wwid -C lunpath -I 1192

Binding of LUN path 0/0/0/7/0/0/0.0x50060e8016030d07.0x40ca000000000000 with new LUN validated successfully

 

#ioscan –fnNC disk

 

#insf -e

#dmesg

 

 

 

 

 

 

HPUX : reading the VGID

 

 

Both of the attachments you provided have no data – they only contain 1 line stating the following - Running install script : ./dump_lvmtab.install dump_lvmtab a 1

 

Let me again review the use of the dump_lvmtab tool, from the point where the tool is still in an encrypted state.

 

1)  decrypt the file – use this exact syntax

    #crypt ???04 <dump_lvmtab-28216.bin.crypt >dump_lvmtab.bin

 

3) make the archive executable

  #chmod +x ./dump_lvmtab.bin

 

2) open the archive

   #./dump_lvmtab.bin –s

 

   the following file will be seen after opening the archive

 

    dump_lvmtab                  dump_lvmtab.bin

    dump_lvmtab-28216.bin.crypt  dump_lvmtab.install

    dump_lvmtab-a-1

 

4) run the dump_lvmtab tool to collect data

 

   #./dump_lvmtab -s > dump_lvmtab.out

HPUX : inetd is not getting start (reserved in semaphore)

The error could occurs because there is an active semaphore attached  to the process id.  To show the active semaphores:

      # ipcs -s

      IPC status from /dev/kmem as of Fri Nov 21 16:11:53 2003
      T      ID     KEY                 MODE        OWNER     GROUP
      s       0    0x412007a0 --ra-ra-ra-      root      root
      s       1    0x4e0c0002 --ra-ra-ra-      root      root
      ...............

The resultant list will display information about active semaphores. To find the semaphore associated with the defunct inetd(1M) process,
look for KEY '0x01090522' in the output.  For example,

      # ipcs -s | grep 0x01090522
      s    6149 0x01090522 --ra-r--r--      root       sys

 Note:  0x01090522 represents the KEY identifier for inetd(1M) created semaphores.
To resolve the problem, remove the identified semaphore with following command

     # ipcrm -s <ID>                                                            //ipcrm –s 6149

Now you should be able to start inetd process.

If it does not solve the issue, then please install the attached tusc depot in your system and send us the trace file for sbin/init.d/inetd start command as follows

# ./tusc -o /tmp/inetd_tracefile -ccc -f -l -n -p -E -v -T%T  -b 80 -r all -w all /sbin/init.d/inetd start

example 

Run tusc as below :-

[root@mickey:/home/jkumar1]#
#-> ./tusc_ia -o /tmp/inetd_tracefile -ccc -f -l -n -p -E -v -T%T  -b 80 -r all -w all /sbin/init.d/inetd start
An inetd is already running
Usage:  inetd [-a] [-p proc_limit] [-r count [interval]] [-l | -s]
        inetd [-c]
        inetd [-k]
EXIT CODE: 255
Unable to start Internet Services
[root@mickey:/home/jkumar1]#

grep "semctl" from the output file “/tmp/inetd_tracefile” to find out the id for inetd service

[root@mickey:/home/jkumar1]#
#-> cat /tmp/inetd_tracefile |grep -i semctl
07:45:13 [/usr/sbin/in][4825]{2180184} <-0.000000> semctl(8198, 0, GETPID, 0) ......................................................... [entry]
07:45:13 [/usr/sbin/in][4825]{2180184} <0.000025> semctl(8198, 0, GETPID, 0) .......................................................... = 0
semctl                         0.00           1
[root@mickey:/home/jkumar1]#
                                ==> id = 8198

Verify the id i.e. root owned and nothing else is attached to this id as below :-

[root@mickey:/home/jkumar1]#
#-> ipcs |grep -i 8198
s       8198 0x01090522 --ra-r--r--      root      root
[root@mickey:/home/jkumar1]#
#-> ipcs -mas |grep -i 8198
s       8198 0x01090522 --ra-r--r--      root      root      root      root     1 no-entry 13:17:14
[root@mickey:/home/jkumar1]#

Now stop/kill the above id as below and start the inetd service :-

[root@mickey:/home/jkumar1]#
#-> ipcrm -s 8198
[root@mickey:/home/jkumar1]#
#-> /sbin/init.d/inetd start
Internet Services started
[root@mickey:/home/jkumar1]#
#-> ps -ef |grep -i inetd
    root 13744     1  0 07:54:24 ?         0:00 /usr/sbin/inetd
    root 13763 27302  0 07:54:36 pts/2     0:00 grep -i inetd
[root@mickey:/home/jkumar1]#
 

HPUX : How to check whether Ignite archive fit into a single Tape

Steps to determine if the data archived fits into a single tape:

To determine if a tape created by Ignite-UX versions C.7.11.439 or C.7.11.444 on an HP Integrity system is not valid, run the following commands:

# mt -f /dev/rmt/#mn rew
# mt -f /dev/rmt/#mn fsf 22
# pax -v -f /dev/rmt/#m

where # must be replaced by the number associated with the device that the tape has been loaded into. 

If the archive listing by pax completes without prompting for a new tape, the recovery tape has been created successfully. If you are prompted for a second tape, the recovery tape is not valid and should not be relied upon for recovery.

# pax -v -f /dev/rmt/#m

this will show long listing of content of the file in the tape.it should complete without the asking for the second tape. if it ask for second tape then, archive for recovery is incomplete.

HPUX : one line to Get license of server

one liner to find the Serial number and product of the server

echo "selclass qualifier system;info;wait;infolog" | cstm | grep -i -e "system serial" -e "Current Product"

HPUX :How to check the memory/processor at console for rp7410 server

How to check the memory/processor at console for rp7410 server

 

Main Menu: Enter command or menu > IN

 

---- Information Menu --------------------------------------------------------

 

     Command                           Description

     -------                           -----------

     ALL [<cell>]                      Display all of the information

     BootINfo                          Display boot-related information

     CAche [<cell>]                    Display cache information

     ChipRevisions [<cell>]            Display revisions of major VLSI

     ComplexID                         Display Complex information

     FRU [<cell>] [CPU|MEM]            Display FRU information

     FwrVersion [<cell>]               Display version for PDC, ICM, and Complex

     IO [<cell>]                       Display I/O interface information

     MEmory [<cell>]                   Display memory information

     PRocessor [<cell>]                Display processor information

 

     BOot [PRI|HAA|ALT|<path>]         Boot from specified path

     DIsplay                           Redisplay the current menu

     HElp [<command>]                  Display help for specified command

     REBOOT                            Restart Partition

     RECONFIGRESET                     Reset to allow Reconfig Complex Profile

     MAin                              Return to Main Menu

----

Information Menu: Enter command > Me

 

PARTITION MEMORY INFORMATION

 

Cell   DIMM Rank 0       DIMM Rank 1       DIMM Rank 2       DIMM Rank 3

      Size   Status     Size   Status     Size   Status     Size   Status

---- ------ ---------  ------ ---------  ------ ---------  ------ ---------

  1  2048MB Active     2048MB Active       ---               ---

 

       Partition Total Memory:     4096 MB

      Partition Active Memory:     4096 MB

Partition Deconfigured Memory:        0 MB

 

* status is scheduled to change on next boot.

 

Information Menu: Enter command > Pr

 

PROCESSOR INFORMATION

 

        Cab/                                                     Processor

Cell   Slot   CPU    Speed     HVERSION   SVERSION   CVERSION     State

----   ----   ---   --------   --------   --------   --------  --------------

   1    0/1     0     750 MHz    0x5e40     0x0491     0x0202    Active

                1     750 MHz    0x5e40     0x0491     0x0202    Idle

 

             Partition Total Cells: 1

        Partition Total Processors: 2

       Partition Active Processors: 2

Partition Deconfigured Processors: 0

 

Information Menu: Enter command >

HPUX : How to Rename a partition with parmodfiy

How to Rename a partition with parmodify

root@Jupiter#parstatus -P

[Partition]

Par              # of  # of I/O

Num Status       Cells Chassis  Core cell  Partition Name (first 30 chars)

=== ============ ===== ======== ========== ===============================

 0  active         1      1     cab0,cell2 Partition 0

root@Jupiter#parmodify -p 0 -P Jupiter


Command succeeded.

root@Jupiter#parstatus -P

[Partition]

Par              # of  # of I/O

Num Status       Cells Chassis  Core cell  Partition Name (first 30 chars)

=== ============ ===== ======== ========== ===============================

 0  active         1      1     cab0,cell2  Jupiter

root@Jupiter#

HPUX : Unable to allocate MI shared memory (midaemon )

Unable to allocate MI shared memory (midaemon )

 

more /var/opt/perf/status.mi

 

Unable to allocate MI shared memory. The current MI shared memory size is

too small. Possible MI subclass overflow (too  many processes, threads,

transactions, etc.) Further allocation errors will be suppressed. Terminate

the midaemon and restart using the -smdvss option. For more details see the

midaemon man page.

Add_to_freelist fails: ENOMEM

 

#-> cp -p /etc/rc.config.d/ovpa  /etc/rc.config.d/ovpa.backup

#-> vi /etc/rc.config.d/ovpa

"/etc/rc.config.d/ovpa" 25 lines, 1037 characters

#!/sbin/sh

# @(#)ovpa config     11.00.020       25FEB2010

#

# OpenView Performance Agent (ovpa) startup control

#

# To execute the scopeux daemon automatically at boot time, ensure the

# MWA_START variable below is set to 1.  Any other value will disable

# scopeux execution at bootup.  For values other than 1, scope can be

# started after bootup is complete with the /opt/perf/bin/mwa script.

#

# To start up OpenView Performance Agent with special options, change the

# MWA_START_COMMAND from its default value of "/opt/perf/bin/mwa start"

# to the desired value.  For example, to start up scope but not the

# servers change the value to "/opt/perf/bin/mwa start scope".

#

# WARNING: BE SURE TO MAINTAIN "" AROUND THE START COMMAND STRING

#          OR THE SCRIPT WILL NOT WORK!!!

#

# This file originally has MWA_START set to 1 because it is assumed

# that the user will edit the /var/opt/perf/parm file to suit their

# environment prior to system boot.

#

MWA_START=1

MWA_START_COMMAND="/opt/perf/bin/mwa start"

MWA_PROTOCOL=http

export MIPARMS

MIPARMS="-p -pids 18000 -kths 18000 -bufsets 16 -skipbuf 8 -smdvss 160M"

~

~

"/etc/rc.config.d/ovpa" 27 lines, 1125 characters

 

#-> mwa stop all
#-> mwa start all

 

#-> mwa status

Perf Agent status:

    Running scopeux               (Perf Agent data collector) pid 12565

    Running midaemon              (Measurement Interface daemon) pid 12554

    Running ttd                   (ARM registration daemon) pid 16780

 

Perf Agent Server status:

 

    Running ovcd                  (OV control component) pid 16838

    Running ovbbccb               (BBC5 communication broker) pid 16867

    Running coda                  (perf component) pid(s) 16935

    Running perfalarm             (alarm generator) pid(s) 12584

 

#-> ps -ef |grep -i midaemon

    root 12554     1 65 08:39:35 ?         0:02 /opt/perf/bin/midaemon -p -pids 18000 -kths 18000 -bufsets 16 -skipbuf 8 -smdvss 160M

    root 12908 24122  1 08:39:54 pts/2     0:00 grep -i midaemon

 

================

 

HPUX :How to Activate Root logging script on the server.

How to Activate Root logging script on the server.

 

mkdir /.root/.histdir/

 

Enter below code in /.root/.profile

cp –p /.root/.profile /.root/.profile.$(date +%Y%m%d-%H%M%S)

 

#--------------------------------------------

EDITOR=vi

HISTSIZE=1000

USER=$(whoami)

PARID=$(ps -f -p $$ | tail -1 | awk '{print $3}')

RUSER=$(ps -f -p $PARID | tail -1 | awk '{print $1}')

HISTFILE=$HOME/.histdir/$USER.$RUSER.$(date +%Y%m%d-%H%M%S)

HOST=$(hostname)

PS1='[$USER@$HOST:$PWD]#

#-> '

 

export USER HISTFILE RUSER HOST HISTFILE HISTSIZE EDITOR PATH PS1

 

#----------------------------------------------

HPUX : USAGE of vgchgid (man vgchgid)

DESCRIPTION

      The vgchgid command is designed to change the LVM Volume Group ID

      (VGID) on a supplied set of disks.  vgchgid will work with any type of

      storage, but it is primarily targeted at disk arrays that are able to

      create "snapshots" or "clones" of mirrored LUNs.  vgchgid accepts a

      set of raw physical devices and ensures that they all belong to the

      same volume group, before altering the VGID (see WARNINGS section).

 

      The same VGID is set on all the disks and it should be noted that in

      cases of multi-PV volume groups, all the physical volumes should be

      supplied in a single invocation of the vgchgid command.

 

    Background

      Some storage subsystems have a feature which allows a user to split

      off a set of mirror copies of physical storage (termed BCVs, BCs, or

      Snapshots) just as LVM splits off logical volumes with the lvsplit

      command.  As the result of the "split," the split-off devices will

      have the same VGID as the original disks.  vgchgid is needed to modify

      the VGID on the BCV devices.  Once the VGID has been altered, the BCV

      disks can be imported into a new volume group by using vgimport.

 

      Once the VGID has been changed, the original VGID is lost until a disk

      device is re-mirrored with the original devices.  If vgchgid is used

      on a subset of disk devices (for example, two out of four disk

      devices), the two groups of disk devices would not be able to be

      imported into the same volume group since they have different VGIDs on

      them.  The solution is to re-mirror all four of the disk devices and

      re-run vgchgid on all four BCV devices at the same time, and then use

      vgimport to import them into the same new volume group.

 

      If a disk is newly added to an existing volume group and no subsequent

      LVM operations has been performed to alter the structures (in other

      words, operations which perform an automated vgcfgbackup(1M)); then it

      is possible a subsequent vgchgid will fail.  It will report that the

      disk does not belong to the volume group. This may be overcome by

      performing a structure changing operation on the volume group (for

      example, using lvcreate).

 

EXAMPLES

      An example showing how vgchgid might be used:

 

      1. The system administrator uses the following commands to create the

         Business Continuity (BCV or BC) copy:

 

         1)   For EMC Symmetrix disks, the commands are BCV establish and

              BCV split.

 

         2)   For XP disk array, the commands are paircreate and pairsplit.

 

         Three BCV disks are created.

 

      2. Change the VGID on the BCV disks.

 

         vgchgid /dev/rdsk/c0t0d0 /dev/rdsk/c0t0d1 /dev/rdsk/c0t0d2

 

      3. Make a new volume group using the BCV disks.

 

         mkdir /dev/vgbcv

         mknod /dev/vgbcv/group c 64 0x040000

 

         NOTE: This step can be skipped as the group file will be created

         automatically.  If the group file is manually created it will have

         different major and minor numbers (see lvm(7)).

 

      4. Import the BCV disks into the new volume group.

 

         vgimport /dev/vgbcv /dev/dsk/c0t0d0 /dev/dsk/c0t0d1 /dev/dsk/c0t0d2

 

      5. Activate the new volume group.

 

         vgchange -a y /dev/vgbcv

 

      6. Backup the new volume group's LVM data structure.

 

         vgcfgbackup /dev/vgbcv

 

      7. Mount the associated logical volumes.

 

         mkdir /bcv/lvol1 /bcv/lvol2

         mount /dev/vgbcv/lvol1 /bcv/lvol1

         mount /dev/vgbcv/lvol2 /bcv/lvol2

HPUX : Businees copy disk check for XP array disks

Prior to adding a disk to an existing volume group please make sure that the disk is not a Business Copy SVOL volume.  

To verify if device is an SVOL do:

# /usr/local/CPR/bin/xpinfo –c

This will list you all the devices on the system with a lable indicating if it’s a PVOL or an SVOL.

Please note that on large systems this can take more time than a regular xpinfo –i.

You can also use the –d option which will give you a long detail report of all the luns. 

This will give you BC/CA  Status, as well as the VG name.

# /usr/local/CPR/bin/xpinfo -d

#-> /usr/local/CPR/bin/xpinfo -c

                                   CT       CA        BC Volume

Device File              Subsys    Group    Vol     MU#0 MU#1 MU#2

==================================================================

/dev/rdsk/c70t0d0         004d     ---      SMPL    PVOL SMPL SMPL

/dev/rdsk/c70t0d1         004d     ---      SMPL    PVOL SMPL SMPL

/dev/rdsk/c70t0d2         004d     ---      SMPL    PVOL SMPL SMPL

/dev/rdsk/c70t0d6         004d     ---      SMPL    PVOL SMPL SMPL

/dev/rdsk/c70t0d7         004d     ---      SMPL    PVOL SMPL SMPL

/dev/rdsk/c70t1d0         0004     ---      SMPL    SMPL SMPL SMPL

/dev/rdsk/c70t2d1         0059     ---      SMPL    SMPL SMPL SMPL

/dev/rdsk/c70t2d2         0059     ---      SMPL    SMPL SMPL SMPL

/dev/rdsk/c70t2d4         005e     ---      SMPL    SMPL SMPL SMPL

/dev/rdsk/c70t2d6         0060     ---      SMPL    PVOL SMPL SMPL

/dev/rdsk/c70t2d7         0060     ---      SMPL    PVOL SMPL SMPL

/dev/rdsk/c70t3d0         0060     ---      SMPL    PVOL SMPL SMPL

/dev/rdsk/c70t3d1         0060     ---      SMPL    PVOL SMPL SMPL

/dev/rdsk/c70t3d3         0060     ---      SMPL    SMPL SMPL SMPL

/dev/rdsk/c70t3d4         0060     ---      SMPL    SMPL SMPL SMPL

/dev/rdsk/c70t3d5         0060     ---      SMPL    SMPL SMPL SMPL

/dev/rdsk/c70t5d3         0061     ---      SMPL    PVOL SMPL SMPL

/dev/rdsk/c70t6d0         0062     ---      SMPL    PVOL SMPL SMPL

/dev/rdsk/c74t0d0         004d     ---      SMPL    SVOL SMPL SMPL

/dev/rdsk/c74t0d1         004d     ---      SMPL    SVOL SMPL SMPL

/dev/rdsk/c74t0d2         004d     ---      SMPL    SVOL SMPL SMPL

HPUX : Issue with Syslog

problems with syslog
The configuration file /etc/syslog.conf should NOT contain spaces as separation between two parameters (of course, in comments lines). Make sure that they are replaced by (one or more) TAB characters.
Do this by means of:
Lazy-cat #/etc/syslog.conf

If the string here somewhere '^ I' is for it means that rather than a TAB character.
Initialize the syslog daemon by:
Lazy-cat # Kill-HUP `cat/var/run/syslog.pid`
Of the restart through:
Lazy-cat # Kill `cat/var/run/syslog.pid`
Lazy-cat #Syslogd -D


 debugging syslog
Start the syslog daemon in debugging mode:
# Syslogd –D –d -m 1

HPUX :"m": Illegal option while extending logical volume with lvextend in PVG

While lvextend we are getting below error on the server, even mirrordisk/UX software is

present on the system

 

$ swlist -l fileset |grep -i mirr

  LVM.LVM-MIRROR-RUN                                    B.11.23        LVM Mirror

  PHCO_40920.LVM-MIRROR-RUN                             1.0            LVM.LVM-MIRROR-RUN

 

 

$ lvdisplay /dev/vgdbRPM1/lvdata_new

--- Logical volumes ---

LV Name                     /dev/vgdbRPM1/lvdata_new

VG Name                     /dev/vgdbRPM1

LV Permission               read/write

LV Status                   available/syncd

Mirror copies               0

Consistency Recovery        MWC

Schedule                    parallel

LV Size (Mbytes)            81920

Current LE                  2560

Allocated PE                2560

Stripes                     0

Stripe Size (Kbytes)        0

Bad block                   on

Allocation                  PVG-strict/distributed

IO Timeout (Seconds)        default

 

[root@server_B:/.root]

$ fstyp /dev/vgdbRPM1/lvdata_new

unknown_fstyp (no matches)

[root@server_B:/.root]

$ newfs -F vxfs -o largefiles -b 8192 /dev/vgdbRPM1/rlvdata_new

    version 5 layout

    83886080 sectors, 10485760 blocks of size 8192, log size 2048 blocks

    unlimited inodes, largefiles supported

    10485760 data blocks, 10483288 free data blocks

    320 allocation units of 32768 blocks, 32768 data blocks

[root@server_B:/.root]

$ lvextend -m 1 /dev/vgdbRPM1/lvdata_new beerse2xp0x_new

Usage: lvextend

        [-A Autobackup]

        {-l LogicalExtentsNumber |

         -L LogicalVolumeSize}

        LogicalVolumePath [ PhysicalVolumePath... | PhysicalVolumeGroupName... ]

"m": Illegal option.

[root@server_B:/.root]

$ whereis lvextend

lvextend: /usr/sbin/lvextend /sbin/lvextend /usr/share/man/man1m.Z/lvextend.1m

 

Solution :

 

For tweak we copied the /usr/sbin/lvextend from 11.23 version os and then try to extend it and its work.

 

[root@server_B:/home/user_one]

$ ll

total 2960

-r--r--r--   1 user_one   users          831 Dec 13  2010 .cshrc

-r--r--r--   1 user_one   users          347 Dec 13  2010 .exrc

-rw-------   1 user_one   users           40 Dec 13  2010 .k5login

-r--r--r--   1 user_one   users          333 Dec 13  2010 .login

-r--r--r--   1 user_one   users          438 Dec 13  2010 .profile

-rw-------   1 user_one   users          650 Feb  1 20:25 .sh_history

-rwxrwxr--   1 root       sys           7647 May 19  2012 SMScheck.sh

-rw-r-----   1 user_one   users         1956 Dec 11  2011 adviser.out

-r-sr-xr-x   1 user_one   users      1371012 Sep  4  2011 lvextend

-rwxrwxr--   1 root       sys            854 Feb 27  2012 nags_check.sh

-rw-r-----   1 root       sys            280 Dec 10  2011 package

drwxr-x---   2 root       sys           8192 Dec 11  2011 todayaction

[root@server_B:/home/user_one]

$ ./lvextend -m 1 /dev/vgdbRPM1/lvdata_new beerse2xp0x_new

lvextend: "/dev/vgdbRPM1/lvdata_new": Permission denied

Usage: lvextend

        [-A Autobackup]

        {-l LogicalExtentsNumber |

         -L LogicalVolumeSize |

         -m MirrorCopies [-s] }

        LogicalVolumePath [ PhysicalVolumePath... | PhysicalVolumeGroupName... ]

 

[root@server_B:/home/user_one]

$ chown root:sys lvextend

[root@server_B:/home/user_one]

$ ./lvextend -m 1 /dev/vgdbRPM1/lvdata_new beerse2xp0x_new

The newly allocated mirrors are now being synchronized.This operation will

take some time. Please wait ....

 

 

 

 

HPUX : local VG activation in Cluster Env

The parameter AUTO_VG_ACTIVATE in file /etc/lvmrc should be set to 0.

Include all the volume groups that are not cluster-related   in the custom_vg_activation function, e.g. VG01 and VG02.

/sbin/vgchange -a y -s /dev/vg01

/sbin/vgchange -a y -s /dev/vg02

parallel_vg_sync "/dev/vg01 /dev/vg02"

Note: Volume groups that will be used by packages should not be included anywhere in the file, since they will be activated and deactivated by control scripts for the package.

The root volume group does not need to be included in the custom_vg_activation function, because it is automatically activated before the /etc/lvmrc file is used at boot time.

HPUX : Adding CLuster node in Quorum Server.

If you’re using a quorum server, authorize the nodes in the cluster to the quorum server.

Login to the quorum server and update the file /etc/cmcluster/qs_authfile

Save a copy and add the FQDN of main interface and the management interface of every node in the cluster to the qs_authfile file.

 

After modifying this file, you must force the quorum server to re-read it:

# /usr/lbin/qs  -update

 

Create the cluster in a semi-automated fashion. The command arguments are different if we’re using a cluster lock device or a quorum server. Execute it first in test mode to verify if you got the arguments correct:

·         Cluster lock device usage:
# cmdeploycl -t -c <cluster>  -N $SGCONF/<cluster>_network -n «PrimaryNode» -n «FailoverNode» -b –L /dev/<vglock>:</dev/cdisk/lockdisk>

·         Quorum server usage:

# cmdeploycl -t -c <cluster>  -N $SGCONF/<cluster>_network -n «PrimaryNode» -n «FailoverNode» –q < FQDN main quorum server> <FQDN management interface quorum server>

When the cmdeploycl command finishes you should see a message like:

Preview complete. To apply changes, run cmdeploycl  without -t option.