HP9000 Model 715 /100 Keyboard

[andybarrett1]

Hi
Thank you for reading.

I have a HP9000 Model 715/100 Workstation running HP-UX 9.05
I want to use a PS2 Keyboard but, even though I have told the boot administrator the path is PS2 and not HiL it still is not seeing my keyboard from the OS .

The Keyboard is good as I am using that keyboard to get into the Boot Admin.

Thank you for any help
Andy

[SanjeevGoyal]

Hello,

HP-UX workstations are finicky with keyboards. It is best to use the most basic versions, HP brand preferably. The recommended PS2 keyboard for the 715/100 715/80 715/64 is the C3757-60201.

If you feel this was helpful please click the KUDOS! thumb below!   
Regards,

[andybarrett1]

Hi 

Thank you 

That is the Keyboard we are using  :- C3757-60201 

We know keyboard ok as works in BOOT ADMIN to set up paths etc ??

The HP-UX is v9.05 if that is anything to do with issue ?

Thank you for "Any" Help 

Andy

[Jesse Dougherty]

Maybe its the moue thats throwing it off... you need to have one of those plugged in or it when it booting or the 715 will throw off errors. The recommended HP mouse is A2838A / A1658-62147. If I were you, I would.. turn it off, plug in the keyboard and mouse and re-try. If you dont have one of those HP A2838A / A1658-62147 mice, try the most basic mouse that you can find. Any newer optical / scrolling / fast keys will probably throw 9.05 into a panic because all that was developed after 9.05 was released.

Jesse

Cypress Technolofgy Inc

[andybarrett1]

Does it the need both the Mouse and Keyboard to be OK

Have just tried with no Mouse  (Just the Keyboard) No Joy

Andy

[Jesse Dougherty]

What mouse are you using, whats the part number? Try another type, if its not the recomended one, its going to be random if its compatable.... I don't know what to tell you.. there are a lot of variables that I don't know so I can only give you the basics. We build and sell these for the past 25 years, if my tech had it here, we could figure it out in 5 minutes but without knowing anything, its impossible to troubleshoot.

[SanjeevGoyal]

Hello,

I would suggest, you have to log a proper case with HPE, and share the appropriate logs for further analysis.
If you feel this was helpful please click the KUDOS! thumb below!   
Regards,

[Bertram Stoeckler]

The 715/100 workstation were introduced in 1994 !

HPUX 9.05 as well, is more than 25 years old.

So congratulation, if this box is still running !  ( maximum supported memory: 256MB !! )

If i remember correctly the 715/100, apart from a PS2, also came with an HIL-interface, which was a HP-proprietery computer-bus. ( https://en.wikipedia.org/wiki/HIL_bus )

So you may want to try to find a HIL-keyboard, if PS2 does not work.

Technically, if the keyboard works in the boot-menue, but not in the HPUX-shell, then you are most likly missing the PS2-driver, or keyboard driver. "ioscan -fnC keyboard" (i am not sure about the syntax, especially the class option "C keyboard" may not exist ) may give you some more information. It could also be that a HIL-keyboard-driver is blocking the other driver, and that you are simply missing that keyboard-type.

Unfortunately, to run a command on a shell, you will need a working keyboard. 

Are you able to connect to the workstation from another system via "telnet" or "rlogin" ( dont´even think about ssh ) ?

[andybarrett1]

Thank you  for you help ...

Yes we have 6 off these still aliive and well .... Except for the keyboards .... They are currently set up with custom made HIL Keyboards which are starting to show their age and failing  ... However the  workstations are  are fine.

I had a look at the commands you sent me to try ..... I can run them from the HIL Keyboard ok, But,  am I right in thinking that HIL will overrule the PS2 if the worstation finds that first .... Even with Ps2 Plugged in ? 

With just PS2 pluged in we get no joy at all ... HIL is locking us out ?

Any help very welcome

Andy

[Bertram Stoeckler]

I am throwing in some ideas, not sure how the keyboard stuff is done.

1.

I would not be surprised if there is NO automatic detection of the connected keyboard. The keyboard setup is most likly done manually in the startup-script. HPUX 10.x changed the system-file layout, and added configuration variables. I do not remember about 9.05. It is therefore possible that the configuration need to be set/removed in the startup script itself.

Unfortunatelly i do not have an user- or admin manual, so i can´t tell where the startup scripts are. 

However the whole startup should be triggered by the an entry located in the /etc/inittab file.

The /etc/inittab file has a line for each run-level, and if i am not wrong, should give you the name and location of the startup-script (could be something like "rc" )

It may also be possible that the hil-keyboard driver startup is triggered via an entry in /etc/inittab...

If you can find the startup file then read it and check if you can find a reference to the keyboard selection/driver setup.

The command to create a device file is "mknod". Adding hardware-pathes may also be done via "insf" (not sure if this exists in 9.05) . I would expect that the is some comment in the startup-file that points to HIL.

The startup file may also start other startup-file ...

2. The second idea that i have is about the device-driver mapping.

Any IO on a Unix machine is done via a device-driver.

There should be device-files under /dev, with names like "disk", and also maybe "keyboard", or "HIL"

Each device-file, when listed with "ls -l" will show a Major and a Minor number.

The Major-number (the first one) is the Index in the device-driver table, and the other one, if i remember correctly is the hardware-path, which raises the question about on which path the keyboard connector is located. (hardware-admin manual would show this)

If the HIL-driver overwrites the PS2-driver than removing it from the kernel may be the only solution.

I am speculating as i really don´t know for sure how it is/were coded. Removing would require re-building the kernel and if you do it wrong, probably no keyboard will work.

So good luck trying to find your way to the startup. It is not that easy.

Do NOT expect that you can use both, HIL and PS2 at the same time.

Btw. There used to be HIL-to-PS2, or PS2-HIL converters around. Do you have one, or does the workstation itself has both connectors?

[andybarrett1]

Hi Thank you for your help 

I think HIL has taken over ....

I have telnet from outside.... Running ioscan with PS2 Key and mouse plugged in still not showing up on lists.

Chasing the below trail I can see the interesting things seem to happen in the /usr/lib/X11  directory 

Around a "XHPKeymaps" file ...... Is this where the configuration will live .... 

Reading the manual it sugests "keymap_ed" will let me edit that but it is not on any of our machines ?

Looking for any more help ??

etc/rc

map_keyboard()

{

#

# Load the appropriate keymap.  If the interface_type is HIL, query the

# keyboard for its language.  If the interface type is PS2, install

# the mapping specified by "-l <map_name>".

#

# This is done here as well as in /etc/bcheckrc.

itemap_option=""

if [ -f /etc/kbdlang ]

then

        read MAP_NAME filler < /etc/kbdlang

        if [ $MAP_NAME ]

        then

                itemap_option="-l $MAP_NAME"

        fi

fi

if [ -x /etc/itemap ]

then

        itemap -i -L $itemap_option -w /etc/kbdlang

fi

}

etc/bcheckrc

# Load the appropriate keymap.  If the interface_type is HIL, query the

# keyboard for its language.  If the interface type is PS2, install

# the mapping specified by "-l <map_name>".

#

itemap_option=""

if [ -f /etc/kbdlang ]

then

        read MAP_NAME filler < /etc/kbdlang

        if [ $MAP_NAME ]

        then

                itemap_option="-l $MAP_NAME"

        fi

fi

if [ -x /etc/itemap ] && [ -f /usr/lib/X11/XHPKeymaps ]

then

        itemap -i -L $itemap_option -w /etc/kbdlang

fi

[Bertram Stoeckler]

Yes, X11 could be an issue here. I forgot about that.

Let´s try to confirm that this is "just" an X11 problem...

Try to boot into single-user-mode and check if the PS2 keyboard still works.

If yes, then you would know that it is not a hardware connectivity issue, nor a HPUX-driver issue, but purely an X11 one.

To boot into single user, interrupt the boot process, then type "hpux -is".

After that just test the keyboard. The majority of commands will not be available in single-user mode.

Either reboot, or run "init 5" to change to another run-level and get X11 started. (i do not remember which run level that was. Maybe 4 or even less the /etc/inittab can tell you)

If the PS2 keyboard works in single-user mode, and stops working in when you run X11, then you can start playing around with the X11 Keyboard setting. I will try to find some X11 documentation....

Can you paste the /etc/inittab file ?

[andybarrett1]

Hi Bertram

I cant get into single user..... It just auto starts and I cant break in .... 

May below files can help ????????

Thank you 

Andy

Inittab

---

 more inittab
init:3:initdefault:
stty::sysinit:stty 9600 clocal icanon echo opost onlcr ienqak ixon icrnl ignpar </dev/systty
brc1::bootwait:/etc/bcheckrc </dev/console >/dev/console 2>&1 # fsck, etc.
slib::bootwait:/etc/recoversl </dev/console >/dev/console 2>&1 #shared libs
brc2::bootwait:/etc/brc >/dev/console 2>&1          # boottime commands
link::wait:/bin/sh -c "rm -f /dev/syscon; \
                       ln /dev/systty /dev/syscon" >/dev/console 2>&1
rc  ::wait:/etc/rc </dev/console >/dev/console 2>&1 # system initialization
powf::powerwait:/etc/powerfail >/dev/console 2>&1   # power fail routines
lp  ::off:nohup sleep 999999999 </dev/lp & stty 9600 </dev/lp
halt:6:wait:/usr/lib/X11/iiapps/haltsys.sh \
  # NOTE: run-level 6 is reserved for system shutdown.
cons:012356:respawn:/etc/getty -h console console           # system console
OS:3:respawn:/products/components/OS/bin/dcStartup
vue :4:respawn:/etc/vuerc                       # VUE validation and invocation​

rc

# more rc
#! /bin/sh
# @(#) $Revision: 1.5 $

#
# Definitions of functions used within this script
#
# The initialize() and localrc() functions below contain
# most of what must be customized in this script.
# This structure has been adopted to minimize the number
# and difficulty of changes required to adopt new functionality
# in future releases of /etc/rc for HP-UX.
#
# Other portions of this script may be customized also, but HP
# recommends that changes be minimized to simplify future updates.
#

initialize()
{
        # The following parameters may be modified for the specific
        # needs of your local system.

        #
        # Set the device file(s) used by /etc/rbootd
        # If no device is specified, /etc/rbootd will
        # use the device corresponding to the ethernet
        # address of the machine.
        #
        RBOOTD_DEVICES=""

        # Set the system's network name:
        # This is done automattically at the first bootup
        # by the /etc/set_parms script.  The system name is
        # written to the /etc/src.sh file for subsequent bootups.
        # The /etc/src.sh file is sourced by this script to set
        # the SYSTEM_NAME variable.

        if [ "$SYSTEM_NAME" = "" ]
        then
           SYSTEM_NAME=unknown
           export SYSTEM_NAME
        fi

        # set the timeout length for date setting:
        # TIMEOUT=0  # skips date setting
        TIMEOUT=20

        # setup for the optional vt gateway, see vtdaemon(1m)
        vtgateway=""            # name of system acting as the gateway
        vtgopts=""              # vtdaemon options
        vtginterfaces=""        # gateway devices
}

localrc()
{
        # This function is intended for adding local initialization
        # functions to rc.  This function is called after all other
        # system initialization is completed.

        # The following line is required for function syntax.
        : # do nothing instruction (a function must contain some command)

        # For example:
        # For HP-IB printers:
        #       Uncomment the 'slp' line below to
        #       set indentation to 0 for /dev/lp.
        #       Similar lines should be added for additional printers.
        #               /usr/bin/slp -i0 > /dev/lp &
        #
        # ABB additions
        # =============
        # Oracle database startup (M Oijwall 920611)
        su oracle -c /products/components/oracle/bin/dbstart
}

# The following functions should require no additional customization:

set_date()
{
  if [ $SET_PARMS_RUN -eq 0 ] ; then
        if [ $TIMEOUT -ne 0 ] ; then
                # This section confirms that the date and time are
                # correct.
                # Systems with battery-backed real-time clock will
                # be correct.  Therefore, the default answer is yes.
                # The question will timeout in $TIMEOUT seconds.  If the
                # question is not answered within the specified timeout,
                # the default answer will be returned.  To increase the
                # timeout, change the value assigned to TIMEOUT (above).
                # TIMEOUT of 0 will skip this question.

                echo "\007Is the date `date` correct? (y or n, default: y) \c"
                reply=`line -t $TIMEOUT`
                echo ""

                if [ "$reply" = y -o "$reply" = "" -o "$reply" = Y ]
                then
                        return
                else
                        if [ -x /etc/set_parms ]; then
                                /etc/set_parms time_only
                        fi
	fi
        fi

   fi   # if SET_PARMS_RUN
}

hfsmount()
{
        # create /etc/mnttab with valid root entry
        /etc/mount -u >/dev/null

        # enable quotas on the root file system
        # (others are enabled by mount)
        [ -f /quotas -a -x /etc/quotaon ] && /etc/quotaon -v /

        # Mount the HFS volumes listed in /etc/checklist:
        /etc/mount -a -t hfs -v
        # (NFS volumes are mounted via net_start() function)

        # Uncomment the following mount command to mount CDFS's
        /etc/mount -a -t cdfs -v

        # Preen quota statistics
        [ -x /etc/quotacheck ] && echo checking quotas && /etc/quotacheck -aP
}

map_keyboard()
{
#
# Load the appropriate keymap.  If the interface_type is HIL, query the
# keyboard for its language.  If the interface type is PS2, install
# the mapping specified by "-l <map_name>".
#
# This is done here as well as in /etc/bcheckrc.

itemap_option=""
if [ -f /etc/kbdlang ]
then
        read MAP_NAME filler < /etc/kbdlang
        if [ $MAP_NAME ]
        then
                itemap_option="-l $MAP_NAME"
        fi
fi

if [ -x /etc/itemap ]
then
        itemap -i -L $itemap_option -w /etc/kbdlang
fi
}

syncer_start()
{
        # Syncer helps minimize file system damage in the event
        # of a power failure or other system crash.
        # run at rtprio to avoid being swapped out
        if /usr/bin/rtprio 127 /etc/syncer
        then
                echo syncer started
        fi
}

lp_start()
{
        #
        # Start lp printer scheduler, if configured.
        #
        # NOTE:
        # For RS-232 printers:
        #     If your line printer interface is RS232 and not set
        #     to 300 baud, then change the 'lp' line in
        #     /etc/inittab from 'off' to 'once' and make sure the
        #     baud rate set there is correct for your printer.
        #
        if [ -s /usr/spool/lp/pstatus ]
        then
                lpshut > /dev/null 2>&1
                rm -f /usr/spool/lp/SCHEDLOCK
                lpsched
                echo line printer scheduler started
        fi

}

clean_ex()
{
        if [ -x /usr/bin/ex ]
        then
                echo "preserving editor files (if any)"
                ( cd /tmp; expreserve -a )
        fi
}

clean_uucp()
{
        if [ -x /usr/lib/uucp/uuclean ]
        then
                echo "cleaning up uucp"
                /usr/lib/uucp/uuclean -pSTST -pLCK -n0
        fi
}

net_start()6%)@
{
        if [ -x /etc/netlinkrc ] && /etc/netlinkrc
        then
                echo NETWORKING started.
        fi
}

swap_start()
{
        # Turn on swapping on alternate swap devices.
        # /etc/checklist "swap" entries configured in the kernel are used.
        if /etc/swapon -a
        then
                echo 'swap device(s) active'
        fi
}

newdisk_run()
{
        #
        # Scan for new disks that need disktab entries
        #
        if [ -x /etc/newdisk -a -x /etc/newdisk_scan ]
        then
           echo "Checking for new unrecognized disks."
           /etc/newdisk_scan -v
        fi
}
cron_start()
{
        if [ -x /etc/cron ]
        then
                if [ -f /usr/lib/cron/log ]
                then
                        mv /usr/lib/cron/log /usr/lib/cron/OLDlog
                fi
                /etc/cron && echo cron started
        fi
}

pty_start()
{
        # Not supported on all systems
        # ptydaemon allocates pty's to various processes
        if [ -x /etc/ptydaemon ]
        then
                echo "starting the ptydaemon"
                /etc/ptydaemon
        fi
}

vt_start()
{
        # Not supported on all systems
        # vtdaemon responds to vt requests from other systems
        #  See vtdaemon(1m) for more information about the vtdaemon.

        if [ -x /etc/vtdaemon ] && [ -c /dev/ieee ]
        then
                case `hostname` in
                    $vtgateway) echo "starting the gateway vtdaemon"
                                /etc/vtdaemon $vtgopts $vtginterfaces
                    ;;
                    *)          echo "starting the vtdaemon"
                                /etc/vtdaemon
                    ;;
                esac}

list_tmps()
{
        for dir in /tmp /usr/tmp /lost+found
        do
                if [ "`ls -A $dir`" ]
                then
                        echo "NOTE:  Files in $dir:"
                        ls -lA $dir
                fi
        done
}

clean_adm()
{
        mask=`umask`
        umask 022
        for LOG in sulog diaglog messages syslog
        do
                if [ -f /usr/adm/$LOG ]
                then
                        mv /usr/adm/$LOG /usr/adm/OLD$LOG
                        if [ $LOG != sulog ]
                        then
                                > /usr/adm/$LOG
                        fi
                fi
        done
        umask $mask
}

switch_over()
{if [ -x /etc/switch/switchsetlan ] && ([ ! -x /bin/getcontext ] ||
            (/bin/getcontext | /bin/egrep "localroot|standalone" > /dev/null))
        then
            . /etc/switch/switchrc
            SWITCH_INFO="${SWITCH_INFO:-/etc/switch/Switchinfo}"
            if [ $SYSTEM_NAME != "unknown" ]; then
                if [ -f $SWITCH_INFO ]; then
                    /etc/switch/switchsetlan -f $SWITCH_INFO $SYSTEM_NAME
                else
                    echo "SwitchOver/UX: $SWITCH_INFO not found"
                fi
            else
                echo "SwitchOver/UX: SYSTEM_NAME unknown"
            fi
        fi
}

envd_start()
{
        if [ -x /etc/envd ] && [ -f /etc/envd.conf ]
        then
                set -- `ps -e | grep envd` ||
                (/etc/envd && echo "Environmental daemon started")
        fi
}

set_state()
{
        # Determine what kind of system this is
        # (standalone, cluster server or client)
        # Set hostname to cnode name if diskless,
        # otherwise use the value from initialize()
        if [ -x /bin/getcontext ] &&  set -- `getcontext` && cnodename=$1 &&
[ "$cnodename" != standalone ]
        then
                cnodes -s || /etc/cluster
                SYSTEM_NAME=$cnodename
                rootname=`cnodes -r`
                if [ ! "$rootname" ]
                then
                        # something is wrong, emit a warning and come up standalone
                        echo "\007ERROR: cannot determine name of DISKLESS ROOTSERVER"
                        echo "\tCorrect, then reboot."
                        echo "\tBringing system up STANDALONE."
                        state=standalone
                elif [ "$SYSTEM_NAME" = "$rootname" ]
                then
                        state=localroot
                else
                        state=remoteroot
                fi
        else
                state=standalone
        fi
}

csp_start()
{
        if ncsp=`/etc/csp`
        then
                echo "$ncsp cluster server process(es) started"
                return 0
        else
                return 1
        fi
}

rbootd_start()@
{
    if [ -x /etc/rbootd -a -n "$RBOOTD_DEVICES" ] || \
       [ -x /etc/rbootd -a -s /etc/clusterconf -a -x /bin/cnodes ]; then
        #
        # rbootd is started if we are the rootserver.
        #
        # rbootd is also started if we are a cnode with local swap
        # and >0 csp's running (this is so that cnodes who swap to
        # this cnode can boot).
        #
        case $state in
        localroot|standalone)
            if /usr/bin/rtprio 64 /etc/rbootd $RBOOTD_DEVICES; then
                echo "remote boot daemon started"
            else
                echo "Could not start remote boot daemon"
            fi
        ;;
        remoteroot)
            ncsps=`/usr/bin/awk -F: '$3 == name && $1 !~ /#/ \
                { i = $NF+0; print i }' name=$SYSTEM_NAME /etc/clusterconf`
            set -- `cnodes -l -m`
            if [ "`cnodes -m`" = "$3" -a "$ncsps" -gt 0 ]; then
                if /usr/bin/rtprio 64 /etc/rbootd; then
                    echo "remote boot daemon started"
                else
                    echo "Could not start remote boot daemon"
                fi
            fi
        ;;
        esac
    fi
}

save_core()
{
        # Not supported on all systems
        # save old kernel core dumps
        if [ -x /etc/savecore ] && [ -d /tmp/syscore ]
        then
        ###     /etc/savecore /tmp/syscore
        ### ABB can have limited disk space, allow partial dumps
                /etc/savecore -p /tmp/syscore
        fi
}

diag_start()
{
        if [ -x /etc/diag.rc ] && /etc/diag.rc
        then
                echo "Diagnostic System Started"
                return 0
        else
                echo "The HP-UX 9.05 Support Tool Manager (Diagnostics) is not installed."
                echo "Installation instructions are in the HP-UX 9.05 document entitled"
                echo "\"README before Installing or Updating HP-UX 9.05\"."
                echo ""
                echo "Note: The Support Tool Manager is not supported on HP-UX Desktop"
                echo "configurations.  It is supported on HP-UX Runtime configurations."
                return 1
        fi
}

audit_start()
{
        # Start up the auditing subsystem
        if [ -x /etc/auditrc ] && /etc/auditrc
then%)@
                echo "Audit subsystem started"
        fi
}

audio_start ()
{
        # Start up the audio server
        if [ -x /etc/audiorc ] && /etc/audiorc
        then
                echo "Audio server started"
        fi
}

syslogd_start()
{
    #
    # Start up the system message logger, see syslogd(1M).
    #
    # The system logger is only started here if networking is not
    # installed.  If networking is installed, syslogd is started by
    # /etc/netlinkrc
    #
    if [ -x /etc/syslogd -a -f /etc/syslog.conf ]
    then
        #
        # If syslogd is already running, we do nothing
        #
        [ -s /etc/syslog.pid ] &&
        kill -0 "`cat /etc/syslog.pid`" 2>/dev/null &&
            return

        if [ -f /usr/adm/syslog ]
        then
   mv /usr/adm/syslog /usr/adm/OLDsyslog
           mask=`umask`
           umask 022
           > /usr/adm/syslog
           umask $mask
        fi
        /etc/syslogd && echo "System message logger started"
    fi
}

set_privgrp()
{
    if [ -f /etc/privgroup ]
    then
        /etc/setprivgrp -f /etc/privgroup
    fi
}

setparms()
{
        #
        # Set system configuration values
        #
        if [ ! -f /etc/src.sh -a -x /etc/set_parms ]
        then    # set the system name, IP addr., TZ, time/date.
           /etc/set_parms
           SET_PARMS_RUN=1
        else
           SET_PARMS_RUN=0
        fi

        if [ -r /etc/src.sh ]
        then
           . /etc/src.sh
		else%)@
           echo "\nWARNING: /etc/src.sh not created by /etc/set_parms."
           echo "Time zone and system name not set.\n"
        fi
}

#
# Here is the heart of the rc script:
#

# Where to find commands:
PATH=/bin:/usr/bin:/usr/lib:/etc

# Set termio configuration for output device.
stty clocal icanon echo opost onlcr ixon icrnl ignpar

# Announce start of rc script
echo "\n/etc/rc:\n"

if [ ! -f /etc/rcflag ]         # Boot time invocation only
then
        # /etc/rcflag is removed by /etc/brc at boot and by shutdown
        touch /etc/rcflag

        hfsmount
        map_keyboard
        setparms
        initialize
        switch_over
        set_state       # determine if standalone, diskless server or
                        # client.  Also sets SYSTEM_NAME for diskless

        uname -S $SYSTEM_NAME
        hostname $SYSTEM_NAME

        # Actions based on system type:
        case $state in
        standalone)     # Not a member of a diskless cluster
                echo "Starting up standalone system"
                set_privgrp
                set_date
                save_core
                swap_start
                syncer_start
                lp_start
                clean_ex
                clean_uucp
                net_start
                rbootd_start
                cron_start
                pty_start
                vt_start
                list_tmps
                clean_adm
                diag_start
                syslogd_start     # must be invoked after net_start
                envd_start        # must be invoked after syslogd_start
                audit_start
                audio_start
                ;;

        localroot)      # This is a root server in a Diskless system
                echo "Starting up CLUSTER SERVER: $rootname"
                set_privgrp
                set_date
                save_core
                swap_start
                syncer_start
			lp_start
                clean_ex
                clean_uucp
                net_start
                csp_start
                rbootd_start
                cron_start
                pty_start
                vt_start
                list_tmps
                clean_adm
                diag_start
                syslogd_start     # must be invoked after net_start
                envd_start        # must be invoked after syslogd_start
                audit_start
                audio_start
                ;;

        remoteroot)     # This is a client in a Diskless system
                SWAP_SITE=`awk -F: '{if (substr($1,1,1)!="#" && $3==sn) print $5}' sn=$SYSTEM_NAME /etc/clusterconf`
                SWAP_SERVER=`awk -F: '{if (substr($1,1,1)!="#" && $2==ss) print $3}' ss=$SWAP_SITE /etc/clusterconf`

                echo "Starting up CLUSTER CLIENT: $SYSTEM_NAME"
                echo "\troot server: $rootname"
                echo "\tswap server: $SWAP_SERVER"
                set_privgrp
                save_core
                swap_start
                net_start
                csp_start
                rbootd_start
                cron_start
                pty_start
                vt_start
			list_tmps
                clean_adm
                diag_start
                syslogd_start     # must be invoked after net_start
                envd_start        # must be invoked after syslogd_start
                audio_start
                ;;
        esac
        newdisk_run
        localrc
fi
date

bcheckrc

# more bcheckrc
#!/bin/sh
# @(#) $Revision: 70.11 $
#
# This file has those commands necessary to check the file
# system, and anything else that should be done before mounting
# the file systems.
#
# NOTE: This script is not configurable!  Any changes made to this
#       scipt will be overwritten when you upgrade to the next release
#       of HP-UX

trap "" 2

#
# The variable DOECHO controls the echoing of NON-error messages
# from bcheckrc.  To turn off these "progress" messages set DOECHO=FALSE.
#
DOECHO=TRUE


if [ "$DOECHO" = "TRUE" ]
then
   echo "/etc/bcheckrc:\n"
fi

# Disk Mirroring...
#
# If /etc/mirrorrc exists and /etc/bcheckrc is invoked without
# any argument (i.e. when run by init), execute /etc/mirrorrc.
# Tasks performed by /etc/mirrorrc include: configuring mirrors,
# running fsck, invoking mirrorlog and then reimaging mirrors.
#
if hp9000s800 && [ -x /etc/mirrorrc -a -z "$1" ]; then
        exec /etc/mirrorrc fsck
exit 1 #should never get here
fi


#
# Activate LVM volume groups.
# Each logical volume will be activated according to its mirror
# consistency recovery policy.
#
if hp9000s800 && [ -x /etc/lvmrc ]
then
        if [ "$DOECHO" = "TRUE" ]
        then
           echo "Checking for LVM volume groups and Activating (if any exist)"
        fi
        /etc/lvmrc
fi

#
# Run eisa_config in automatic mode on the series 700s (if appropriate).
#
machine=`/bin/uname -m`
if hp9000s700 && [ "$machine" != 9000/705 -a "$machine" != 9000/710 ]; then

    #
    # Figure out if we should run eisa_config. There are three cases:
    #    o  the root file system is already clean -- run eisa_config
    #    o  /dev/rroot is not there (this is a client) -- run eisa_config
    #    o  the root file system is not clean -- don't run eisa_config
    #       (we will reboot again after the root file system is fixed)
    #
    if [ -c /dev/rroot ]; then
        /etc/fsclean /dev/rroot
        status=$?
    else-(31%)@
        status=0
    fi

    if  [ $status = 0 ]; then

        # Set the timezone (if the script is there)
        if [ -r /etc/src.sh ]; then
           . /etc/src.sh
        fi

        # Run the eisa_config program if it is present
        if [ -x /etc/eisa_config ]; then
            /etc/eisa_config -a
            # Now handle the return code from eisa_config. Most codes mean
            # go on with the boot. A few of the codes, however, mean reboot
            # the system or halt the system. All error messages are
            # displayed by the eisa_config program.
            case $? in
                1 | 10)         /etc/reboot ;;
                11 | 12)        /etc/reboot -h ;;
            esac
        else
            echo "eisa_config is not present -- cannot check eisa configuration"
        fi

    fi

fi


#
# Load the appropriate keymap.  If the interface_type is HIL, query the
# keyboard for its language.  If the interface type is PS2, install
# the mapping specified by "-l <map_name>".
#
itemap_option=""
if [ -f /etc/kbdlang ]
then
        read MAP_NAME filler < /etc/kbdlang
        if [ $MAP_NAME ]
        then
                itemap_option="-l $MAP_NAME"
        fi
fi

if [ -x /etc/itemap ] && [ -f /usr/lib/X11/XHPKeymaps ]
then
        itemap -i -L $itemap_option -w /etc/kbdlang
fi


#
# Check the file systems, if necessary.
# The kernel will mark the file systems clean if the disk was unmounted
# before system shutdown or the system was shutdown cleanly.  If the
# file system marked clean, we don't need to run fsck.
#
status=0
stty clocal icanon echo opost onlcr ienqak ixon icrnl ignpar erase "^h"

trap "echo Interrupt" 2

if [ "$DOECHO" = "TRUE" ]
then
   /etc/fsclean -v
else
   /etc/fsclean
fi
clean_status=$?
case $clean_status in
0)
    if [ "$DOECHO" = "TRUE" ]; then
       echo "File system is OK, not running fsck"
    fi
    ;;
1|3)
    if [ $clean_status = 3 ]; then
        echo "\007fsclean RETURNED AN ERROR, CHECK /etc/checklist FILE"
        status=3
    fi
    echo "\007\007FILE SYSTEM(S) NOT PROPERLY SHUTDOWN, BEGINNING FILE SYSTEM REPAIR"
    /etc/fsck -P -F
    case $? in
    0)
        echo "FILE SYSTEM IS FIXED"
        ;;
    4)
        echo "\007\n\n"
        echo "ROOT FILE SYSTEM MODIFIED"
        echo "\007\007REBOOTING SYSTEM TO UPDATE KERNEL DATA STRUCTURES"
        exec /etc/reboot -n -q
        ;;
    8)
        echo "\007\n\n"
        echo "COULD NOT FIX FILE SYSTEM WITH  fsck -P, RUN fsck INTERACTIVELY!"
        echo "STARTING A SHELL FOR MANUAL fsck, ^D WHEN FILE SYSTEM FIXED"
        if [ $clean_status = 3 ]; then
            echo "THIS COULD BE DUE TO A BAD /etc/checklist FILE"
            echo "IF SO, FIX /etc/checklist AND ^D WHEN FINISHED"
        fi

PS1="(in bcheckrc)# "
        export PS1
        /bin/sh
        echo "CONTINUING bcheckrc"
        ;;
    12)
        echo "\007\n\n"
        echo "FSCK INTERRUPTED"
        echo "STARTING A SHELL FOR MANUAL fsck, ^D WHEN FILE SYSTEM FIXED"
        PS1="(in bcheckrc)# "
        export PS1
        /bin/sh
        echo "CONTINUING bcheckrc"
        ;;
    *)
        echo "\007\n\n"
        echo "UNEXPECTED ERROR DURING fsck -P, RUN fsck INTERACTIVELY!"
        echo "STARTING A SHELL FOR MANUAL fsck, ^D WHEN FILE SYSTEM FIXED"
        PS1="(in bcheckrc)# "
        export PS1
        /bin/sh
        echo "CONTINUING bcheckrc"
        ;;
    esac
    ;;
*)
    echo "/etc/bcheckrc:\n"
    echo "\007fsclean RETURNED AN ERROR, CONTINUING, BUT CHECK /etc/checklist FILE"
    status=2
    ;;
esac

#
# Make sure /lib is not mounted.  /lib must not be a mounted volume
# or improper system operation may be observed.
#
if [ -s /etc/checklist ]; then
    error=0
    case "`/bin/getcontext`" in
    *localroot*|*)
        if [ -h /lib -o ! -d /lib ]; then
            error=1
        else
            error=`while read device dir type junk; do
                if [ "$dir" = /lib -a "$type" != ignore -a \
                      "$type" != swap -a "$type" != swapfs ]; then
                    echo 1
                fi
            done < /etc/checklist`
        fi
        ;;
    esac

    case "$error" in
    1*)
        echo "\007ERROR:   /lib must be a regular directory on the root partition." >&2
        echo "\t Your system has an unsupported configuration for /lib." >&2
        echo "\t Improper system operation may be observed when /lib is" >&2
        echo "\t not a regular directory on the root partition.\007" >&2
        echo >&2
        echo "\007WARNING: mounting all local file systems so that /lib is available." >&2
        /etc/mount -a -t hfs && echo "local file systems mounted"
        ;;
    esac
fi

exit $status

---

[Bertram Stoeckler]

Hello Andy,

At the beginning of the discussion you wrote:

"The Keyboard is good as I am using that keyboard to get into the Boot Admin."

Are you able to enter into boot-admin mode ( isl ) ?

If yes, then you may find a menue item that says "enter isl" or so...

If yes, can you try this syntax ( i think that in 9.X you had to specify the full hardware path

hpux -is (;0)/hp-ux

BTW.: Is this a diskless workstation running in a workstation cluster ?

Can you paste the contents of "/etc/kbdlang"

The inittab shows that run-level 4 is the VUE startup, which starts the graphical-window system.

So maybe the /etc/vue script has some settings. 

[andybarrett1]

Hi Bertram

It is stand alone with its internal SCSI Drive With V9.05 

We today managed to get to the isl prompt as you can see in pics:--

"Boot pri isl" got us in ..... Keyboard works fine ... 

Rebooting results in no keyboard or mouse being available anymore 

/etc/kbdlang and /etc/vue dont exist  on our system

This is our X11 Directory if it helps any  :-

# cd X11
# ls
SharedX            XHPKeymaps.old     fonts              rgb.txt
X0devices          XHPmodmap          if                 sys.Xdefaults
X0devices.orig     XKeysymDB          itemap             sys.x11start
X0pointerkeys      XPCmodmap          itemap_option=-l   system.mwmrc
X0screens          app-defaults       nls                system.mwmrc.os
XDINmodmap         bitmaps            read               then
XErrorDB           extensions         rgb.dir            vue
XHPKeymaps         fi                 rgb.pag

[Bertram Stoeckler]

Thanks for the pictures.

Regarding the the single-user boot.

The command "hpux -is" or "hpux (;0)/hp-ux" need to be run from the isl-prompt and not from the BOOT_ADMIN prompt.

In one of my previous post i wrote "So maybe the /etc/vue script has some settings."

The name of the file is wrong. Sorry. It must read "/etc/vuerc" (see /etc/inittab contents. I is started from there)

[andybarrett1]

Hi Bertram

Thank you for ALL your help so far...>>

The command "hpux -is" or "hpux (;0)/hp-ux" need to be run from the isl-prompt and not from the BOOT_ADMIN prompt.

>> We are at the ISL prompt hpux just does a reeboot 

>> When we type "hpux (;0)/hp-ux" or "hpux-is" we get not an isl command or utility .....See enclosed pics

Also find our vuerc script 

---

# more vuerc
#!/bin/ksh
#################################################################
#
#  vuerc
#
#  @(#)HP Visual User Environment 3.0          $Revision: 2.7 $
#
# Shell script for initialization of the HP Visual User Environment
#
#################################################################




#       Default (hp-ux) version
        VUELOGIN=/usr/vue/bin/vuelogin
        VL_ARGS=""


        if [ -r /etc/src.sh ]
        then
            . /etc/src.sh
        fi

        if [ -x "$VUELOGIN" ]
        then
            exec $VUELOGIN $VL_ARGS </dev/null >/dev/null 2>&1
        else

            #
            # Note: This path occurs if vuelogin can not be found or
B--More-    #       is not executable...
            #


            #
            # Wait for the console getty (if any) to start.
            #

            for COUNT in 1 2 3 4 5
            do
                TMP=`/bin/ps -tconsole | /bin/fgrep getty`
                if [ $? -eq 0 ]; then
                    GETTY=true
                    break
                fi
                sleep 1
            done


            #
            # Print a message to the console.
            #

            echo "" > /dev/console
            echo "HP VUE can not be started. (See /etc/vuerc)" \
              > /dev/console


            #
            # If a getty is not already running on the console, start one.
            #
            if [ -z "$GETTY" ]; then
                exec /etc/getty console console
            fi

        fi

###################             eof         #####################
​

****** One more thing ******

If ok to ask .... We are looking to do backups to a tape streamer for transfer to another disk.

What would be the best method ....We are looking at Backup / Recovery in SAM

But is there a mirror command to mirror disk directly 

THank you again for all your help 

Andy

[Bertram Stoeckler]

Hello Andy,

1. About the single-user mode. Please try with a "-" removed, from the ISL:

hpux -is(;0)/hpux

2. The vuerc-file simply sources the /etc/src file and then starts "vuelogin".

You may want to check if you find any reference/tuneable inside the /etc/src file for the keyboard setting.

As a last resource, what you can try to do is to run/use "sam". Its a command to configure the system. Not sure if it was already available in 9.05 and especially not sure if you can use it to configure the keyboard.

I would focus on trying to get into single-user to check if the PS2 keyboard still works. 

3. About the backup question. Here are some options. I suggest that you google and/or search HPE-community for more information. If you have further questions about this, open a new topic.

a). Check if the logical volume manager "LVM" is used/configured. I do not think it is on a 9.05 workstation, but you can simply run "vgdisplay", "pvdisplay" or "lvdisplay". If they return some information, then LVM is in use. Search the internet or HPE community for "backup" or "mirror" with HPUX-LVM. You could use the LVM mirroring to copy the raw-LV-volumes to a different disk. There will be several steps needed and your target system also will need to have/support LVM.

2.  You can use "dd" to copy raw-devices. Search the HPE-community or the internet.

3. On filesystem level you can use the following (check the man-page of the commands)

fbackup and frestore, cpio and tar. 

If you have remote-login (remsh) configured, then you can also copy files with "cp", "tar" or "cpio" over the network. Note that as far as i know, "ssh" does not exist. "rlogin and remsh" are the ones that you can use, instead.

Hope that helps.

[andybarrett1]

Hi Bertram

An update.....  We are still struggling with the keyboard issue .... Thank you for all your help so far , we will come back to that 

At the moment the bigger focus is on getting backups from the machines .

This is also giving us issues .... As we have long lost the install tapes 

We are attempting to backup using DD to new media.

This is not easy either 

Andy