Oracle Server Build - Global System Administration: Power 5 Virtualization

Current Location

css
GlobalSysAdmin
Power5

Digg this page

Post to del.icio.us

Slashdot it!

Business Web Site Hosting
$3.99 / month includes Tools,
Shopping Cart, Site Builder
www.siteox.com

FREE Domain Registration
included with Web Site Hosting
Tools, Social Networking, Blog
www.siteox.com

The following document describes the procedure to build and configure an existing AIX system as an Oracle database server. It is assumed the base AIX system is built according to the standard policies and procedures, and exists as a virtual LPAR client in a VIO environment.

Login to the LPAR and create ssh keys for password-less logins and command execution.
Oracle LPAR:
ssh-keygen -t rsa

Build a "dolvs" script for the LPAR and assign it a name using the following syntax: "dolvs.${LPAR_NAME}.sh". An example of this script is shown here which builds multiple resource groups with multiple volume groups. A unique resource group section should be configured in this script for each resource group that will be supported on the Oracle LPAR. Additionally, paging resource group is configured to contain the paging space LUN's.

Oracle LPAR: dolvs.${LPAR_NAME}.sh

#!/usr/bin/ksh93
################################################################
douser()
{
  UID=$( print "${U}" | sum -r )
  UID="$(( ${UID// /} + 0 ))"
  GID=$( print "${G}" | sum -r )
  GID="$(( ${GID// /} + 0 ))"
  rmuser ${U}
  rmgroup ${G}

  print "${G}=${GID}"
  mkgroup id=${GID} ${G}

  print "${U}=${UID}"
  mkuser id=${UID} pgrp=${G} ${U}

  lsuser ${U}
}
################################################################
dovg()
{
  umount ${MTPT}
  /usr/sbin/varyoffvg ${RG}${VGID}
  /usr/sbin/exportvg ${RG}${VGID}

  mkvg -f -y ${RG}${VGID} -V ${VGMJ} ${VGDISKS}
  chvg -a n ${RG}${VGID}
  chvg -c ${RG}${VGID}
  varyoffvg ${RG}${VGID}
  exportvg ${RG}${VGID}
  importvg -y ${RG}${VGID} -V ${VGMJ} ${VGDISKS%%[$' \t\n']*}
  varyonvg ${RG}${VGID}

  /usr/sbin/mklv -y ${RG}${LGID}lv -t jfs2log -a e "${RG}${VGID}" 1

  LVSIZE=$(( $( lsvg ${RG}${VGID} | grep "FREE PPs:" |
      sed -e "s/.*FREE PPs: *//g;s/ (.*//g" ) ))
  print "${LVSIZE}"

  /usr/sbin/mklv -y ${RG}${LVID}lv -t jfs2 -a e "${RG}${VGID}" ${LVSIZE%%.*}

  /usr/sbin/crfs -v jfs2 -d "${RG}${LVID}lv" -m "${MTPT}" -A y \
      -p rw -a agblksize=4096 -a logname="${RG}${LGID}lv"

  chown -R ${U}:${G} /${RG}
  ls -ld ${MTPT}
  ls -ld ${MTPT}
  mount ${MTPT}
  ls -ld ${MTPT}
  ls -ld ${MTPT}

  mkdir -p "${MTPT}/app/oracle/admin"
  chuser home="${MTPT}/app/oracle/admin" ${U}
  chown -R ${U}:${G} ${MTPT}

  lsuser -a home ${U}
  ls -ld ${MTPT}
  print
  lsvg
  print
  df
  print
}
################################################################

RG="eegadtu0"
U=oraeeg
G=dbaeeg
VGMJ="430"
# VGDISKS=$( lscfg -l 'hdisk*' | egrep -- '-C240-|-C245-' | awk '{ print $1 }' )
VGDISKS="hdisk2 hdisk3 hdisk4 hdisk5"
VGID="01vg"
LVID="ora1"
LGID="jfs1"
MTPT="/${RG}/u01"

[[ "_${VGDISKS}" == "_" ]] && exit 1

douser
dovg

################################################################

RG="eegadtu0"
U=oraeeg
G=dbaeeg
VGMJ="431"
# VGDISKS=$( lscfg -l 'hdisk*' | egrep -- '-C240-|-C245-' | awk '{ print $1 }' )
VGDISKS="hdisk6 hdisk7 hdisk8 hdisk9 hdisk10 hdisk11 hdisk12 hdisk13 hdisk14 hdisk15"
VGID="02vg"
LVID="ora2"
LGID="jfs2"
MTPT="/${RG}/u02"

[[ "_${VGDISKS}" == "_" ]] && exit 1

# douser
dovg

################################################################

RG="eegadtu0"
U=oraeeg
G=dbaeeg
VGMJ="432"
# VGDISKS=$( lscfg -l 'hdisk*' | egrep -- '-C240-|-C245-' | awk '{ print $1 }' )
VGDISKS="hdisk16 hdisk17 hdisk18 hdisk19 hdisk20 hdisk21 hdisk22 hdisk23 hdisk24 hdisk25"
VGID="03vg"
LVID="ora3"
LGID="jfs3"
MTPT="/${RG}/u03"

[[ "_${VGDISKS}" == "_" ]] && exit 1

# douser
dovg

################################################################

RG="sfaadtu0"
U=orasfa
G=dbasfa
VGMJ="433"
# VGDISKS=$( lscfg -l 'hdisk*' | egrep -- '-C250-|-C255-' | awk '{ print $1 }' )
VGDISKS="hdisk26 hdisk27 hdisk28"
VGID="01vg"
LVID="ora1"
LGID="jfs1"
MTPT="/${RG}/u01"

[[ "_${VGDISKS}" == "_" ]] && exit 1

douser
dovg

################################################################

RG="sfaadtu0"
U=orasfa
G=dbasfa
VGMJ="434"
# VGDISKS=$( lscfg -l 'hdisk*' | egrep -- '-C250-|-C255-' | awk '{ print $1 }' )
VGDISKS="hdisk29 hdisk30 hdisk31"
VGID="02vg"
LVID="ora2"
LGID="jfs2"
MTPT="/${RG}/u02"

[[ "_${VGDISKS}" == "_" ]] && exit 1

# douser
dovg

################################################################
################################################################
#### 
#### Paging Space
#### 

RG="pagadtu1"
PGMJ="911"
PGDISKS=$( lscfg -l 'hdisk*' | egrep -- '-C58-|-C59-' | awk '{ print $1 }' )

[[ "_${PGDISKS}" == "_" ]] && exit 1

/usr/sbin/varyoffvg ${RG}00vg
/usr/sbin/exportvg ${RG}00vg
/usr/sbin/mkvg -f -S -y ${RG}00vg -V ${PGMJ} ${PGDISKS}

LVSIZE=$(( $( lsvg ${RG}00vg | grep "FREE PPs:" |
    sed -e "s/.*FREE PPs: *//g;s/ (.*//g" ) + 0 ))
print "${LVSIZE}"

/usr/sbin/mklv -y ${RG}pag0lv -t paging -a e "${RG}00vg" ${LVSIZE%%.*}

chps -a y ${RG}pag0lv
lsps -a

Edit and transfer the "dolvs.${LPAR_NAME}.sh" script to the Oracle LPAR. Execute this script on the Oracle LPAR to build the resource groups, volume groups, logical volumes, filesystems and users.
Oracle LPAR
scp dolvs.${LPAR_NAME}.sh root@${LPAR_NAME}:. ssh root@${LPAR_NAME} dolvs.${LPAR_NAME}.sh

Remove any existing disks and virtual SCSI adapters on the Oracle LPAR, then reconfigure and rediscover the Virtual SCSI adapters and disks.

Oracle LPAR

for i in $( lsdev -Cc disk -F name )
do
    rmdev -Rdl ${i}
done

for i in $( lsdev -Cc adapter -F name | grep vscsi )
do
    rmdev -Rdl ${i}
done

cfgmgr

Make sure all disks have a PV ID number associated with it.

Oracle LPAR

for i in $( lsdev -Cc disk -F name )
do
  chdev -l ${i} -a pv=yes
done

Edit and transfer the "vscsiPriority.sh" script to the Oracle LPAR. Execute this script on the Oracle LPAR to prioritize the communication paths to each disk.

Oracle LPAR

ssh  root@${LPAR_NAME}     mkdir -p /usr/local/scripts   
scp  vscsiPriority.ksh     root@${LPAR_NAME}:/usr/local/scripts
ssh  root@${LPAR_NAME}     chmod 755 /usr/local/scripts/vscsiPriority.ksh
ssh  root@${LPAR_NAME}     /usr/local/scripts/vscsiPriority.ksh -vori 20

Set the SMTP mail relay host to "mail-relay.txu.com"

Oracle LPAR

sed -e "s/^DS.*/DSmail-relay.txu.com/g" /etc/sendmail/sendmail.cf > /tmp/tmp${$}.out
sed -e "s/^#DS.*/DSmail-relay.txu.com/g" /tmp/tmp${$}.out > /etc/sendmail/sendmail.cf
rm -f /tmp/tmp${$}.out

refresh -s sendmail

Configure the network time protocol daemon and refresh

Oracle LPAR

print "server 146.61.47.25 prefer
server tick.cs.unlv.edu
server wuarchive.wustl.edu
tracefile /etc/ntp.trace" > /etc/ntp.conf

refresh -s xntpd

Set the "ulimits" for each oracle user

Oracle LPAR

for UNAME in $( grep "^ora" /etc/passwd | awk -F: '{ print $1 }' )
do
    chuser cpu=-1       ${UNAME}
    chuser fsize=-1     ${UNAME}
    chuser data=-1      ${UNAME}
    chuser stack=32768  ${UNAME}
    chuser rss=-1       ${UNAME}
    chuser core=2097151 ${UNAME}
    chuser nofiles=2000 ${UNAME}
    grep -p "${UNAME}:" /etc/security/limits
done

Manually edit the "/etc/passwd" file to change the home directories to their appropriate location for each oracle user:
Oracle LPAR
# "/mtpt/u01/app/oracle/admin" vi /etc/passwd

To prepare for Oracle install, create the "hagsuser" group
Oracle LPAR
mkgroup hagsuser

Allow the Oracle users to schedule cron jobs by adding their user names to the "cron.allow" file.

Oracle LPAR

print "root" > /var/adm/cron/cron.allow
for UNAME in $( grep "^ora" /etc/passwd | awk -F: '{ print $1 }' )
do
  print "${UNAME}" >> /var/adm/cron/cron.allow
done

Configure WLM to run in passive mode and configure it to be started at boot time from the /etc/inittab
Oracle LPAR
nohup wlmcntrl -p > /dev/console 2>&1 & rmitab wlm mkitab "wlm:2:once:wlmcntrl -p > /dev/console 2>&1"

Configure system and Asynchronous I/O parameters into the kernel

Oracle LPAR

# Maximum number of processes per user = 3000
chdev -l sys0 -a maxuproc=3000 -P

# Maximum Async I/O requests = 16386
chdev -l aio0 -a maxreqs=16386 -P

# Maximum number of Async I/O servers = 300
chdev -l aio0 -a maxservers=300 -P

# Minimum number of Async I/O servers = 150
chdev -l aio0 -a minservers=150 -P

chdev -l aio0 -a autoconfig='available'
mkdev -l aio0

Configure performance monitoring on LPAR's with storage on the NIM server. The specific NIM server will depend upon in which datacenter the LPAR resides.

Oracle LPAR


# Begin by starting a Korn Shell 93 session
/usr/bin/ksh93

# Set a shell variable to contain the hostname of the Oracle LPAR
AIXSYS="$( hostname )"
AIXSYS="${AIXSYS%%.*}"

# Set a shell variable to contain the hostname of the NIM Server.  
# In this example "ddcapnim01" is used, this should be changed to 
# the actual NIM server desired.
NFSSRV="ddcapnim01"

print "AIXSYS=${AIXSYS}    NFSSRV=${NFSSRV}"

# Set an array of values to contain the performance data mount 
# points for each NIM Server
typeset -A MTPT
MTPT[mdcapnim01]="/prfpmce0"
MTPT[ddcapnim01]="/prfdmce0"

Create an NFS mount for the performance data on the Oracle LPAR and mount the filesystem.

Oracle LPAR

umount ${MTPT[${NFSSRV}]}
/usr/sbin/rmnfsmnt -f ${MTPT[${NFSSRV}]} -B
/usr/sbin/mknfsmnt -f ${MTPT[${NFSSRV}]} -d ${MTPT[${NFSSRV}]} -h ${NFSSRV} -M sys -t rw -B -A -w bg -Y -X -H -j -q -g
mount ${MTPT[${NFSSRV}]}
grep -p "${MTPT[${NFSSRV}]}:" /etc/filesystems
df

Start xmwlm and xmtrend in init level "a" with HACMP after NFS starts

Oracle LPAR

rmitab ptxwlm
rmitab ptxtrend
mkitab "ptxwlm:a:respawn:/usr/bin/xmwlm -d ${MTPT[${NFSSRV}]}/${AIXSYS}/wlm -n xmwlm > /dev/null 2>&1 # Start xmwlm"
mkitab "ptxtrend:a:respawn:/usr/bin/xmtrend -f /etc/perf/jtopas.cf -d ${MTPT[${NFSSRV}]}/${AIXSYS}/Top -n jtopas > /dev/null 2>&1 # Start trend"
egrep "^ptxwlm:|^ptxtrend:" /etc/inittab

Kill any existing xmtrend or xmwlm processes and restart them

Oracle LPAR


for PID in  $( ps -ef -F pid,args | egrep 'xmwlm|xmtrend' | grep -v grep | awk '{ print $1 }' )
do
    print kill ${PID}
    kill ${PID}
    sleep 1
    print kill -9 ${PID}
    kill -9 ${PID}
done
ps -ef -F pid,args | egrep 'xmwlm|xmtrend'

mount ${MTPT[${NFSSRV}]}
mkdir -p ${MTPT[${NFSSRV}]}/${AIXSYS}/wlm
mkdir -p ${MTPT[${NFSSRV}]}/${AIXSYS}/Top
nohup /usr/bin/xmwlm -d ${MTPT[1]}/${AIXSYS}/wlm -n xmwlm > /dev/null 2>&1 &
sleep 4
nohup /usr/bin/xmtrend -f /etc/perf/jtopas.cf -d ${MTPT[${NFSSRV}]}/${AIXSYS}/Top -n jtopas > /dev/null 2>&1 &
sleep 4
find  ${MTPT[${NFSSRV}]}/${AIXSYS} -exec ls -ld {} \;
jobs

Reboot the LPAR to ensure the AIO, kernel, and /etc/inittab settings take effect and are restarted on system startup.
Oracle LPAR
shutdown -Fr

Create Start scripts for each Oracle database. An example start script is shown for the resource group "etapdtu0", database "EEXTPRD", oracle user "oraeta", and oracle listener port "1571".

Oracle LPAR: /usr/local/hascripts/etapdtu0_oradb01_start.sh

#!/usr/bin/ksh93
################################################################

print "START SCRIPT BEGIN: ${0##*/}"

RG="etapdtu0"
DBNAME1="EEXTPRD"
UNAME="oraeta"
LSNRPORT="1571"
typeset -l WLMTAG="${DBNAME1}"

print "/${RG}/u01/app/oracle/product/9.2.0/bin/dbstart ${DBNAME1}" > /tmp/dbstart.${DBNAME1}.${$}.sh
chmod 755 /tmp/dbstart.${DBNAME1}.${$}.sh

su - ${UNAME} -c "/usr/local/bin/wlmsettag ${WLMTAG} /tmp/dbstart.${DBNAME1}.${$}.sh" &
su - ${UNAME} -c "/${RG}/u01/app/oracle/product/9.2.0/bin/lsnrctl start lsnr_${UNAME}_${LSNRPORT}" &

/usr/local/hascripts/rg_wlmassign.sh

print "START SCRIPT END:  ${0##*/}"

exit 0

Create Stop scripts for each Resource Group of Oracle databases. An example stop script is shown for the resource group "etapdtu0", databases "EEXTPRD" and "ETSSPRD", oracle user "oraeta", and oracle listener port "1571".

Oracle LPAR: /usr/local/hascripts/etapdtu0_oradb01_stop.sh


#!/usr/bin/ksh93
################################################################

print "STOP SCRIPT BEGIN: ${0##*/}"

# All databases running under the same user ID should be shutdown
# together in a single "dbshut" command, so modify this section
# to include all database names.  Shown here are two databases.

RG="etapdtu0"
DBNAME1="EEXTPRD"
DBNAME2="ETSSPRD"
UNAME="oraeta"
LSNRPORT="1571"

su - ${UNAME} -c "/${RG}/u01/app/oracle/product/9.2.0/bin/lsnrctl stop lsnr_${UNAME}_${LSNRPORT}" &
su - ${UNAME} -c "/${RG}/u01/app/oracle/product/9.2.0/bin/dbshut ${DBNAME1} ${DBNAME2}" &

CNT="0"
while (( $( jobs | wc -l ) > 0 ))
do
  print "STOP SCRIPT KILL TIMER: $( date ): ${0##*/}"
  if (( CNT >= 300 ))
  then

    for SIGNAL in TERM HUP INT ABRT KILL
    do
      ps -fu ${UNAME}
      for PIDNUM in $( ps -fu ${UNAME} -F pid | grep -v PID )
      do
          print kill -${SIGNAL} ${PIDNUM}
          kill -${SIGNAL} ${PIDNUM}
      done
      sleep 15
    done
    break
  fi
  sleep 15
  (( CNT = CNT + 15 ))
done

while (( $( ps -fu ${UNAME} -F pid | grep -v PID | wc -l ) > 0 ))
do
  print "STOP SCRIPT WAIT TIMER: $( date ): ${0##*/}"
  sleep 15
  (( CNT = CNT + 15 ))
  (( CNT >= 300 )) && break
done

ps -fu ${UNAME}
if (( $( ps -fu ${UNAME} -F pid | grep -v PID | wc -l ) > 0 ))
then
  for SIGNAL in TERM HUP INT ABRT KILL
  do
    print "STOP SCRIPT FINAL KILL: $( date ): ${0##*/}"
    for PIDNUM in $( ps -fu ${UNAME} -F pid | grep -v PID )
    do
        print kill -${SIGNAL} ${PIDNUM}
        kill -${SIGNAL} ${PIDNUM}
    done
    sleep 15
  done
fi

sync
sync
sync
sync

print "STOP SCRIPT END: ${0##*/}"

exit 0

Create an HACMP build/configuration script and customize it for the specific cluster.

Oracle LPAR


#!/usr/bin/ksh93
#################################################################
NODE1="ddcaaora01"
RG1="eegadtu0"
RG1APPLS="${RG1}_oradb01 ${RG1}_oradb02 ${RG1}_oradb03"
RG1VGS="${RG1}01vg ${RG1}02vg ${RG1}03vg"

NODE2="ddcaaora02"
RG2="sfaadtu0"
RG2APPLS="${RG2}_oradb01 ${RG2}_oradb02"
RG2VGS="${RG2}01vg ${RG2}02vg"

################################################################
# Remove Existing Cluster and create a new one:
/usr/es/sbin/cluster/utilities/clrmclstr
/usr/es/sbin/cluster/utilities/clmodclstr -n ${NODE1%??}tu01 -p "${NODE1}-boot.tu.com ${NODE2}-boot.tu.com"
  
################################################################
# Create Service Addresses
/usr/es/sbin/cluster/utilities/clrmnode -a ${NODE1}.tu.com
/usr/es/sbin/cluster/utilities/clrmnode -a ${NODE2}.tu.com
/usr/es/sbin/cluster/utilities/clrmnode -a ${RG1}.tu.com
/usr/es/sbin/cluster/utilities/clrmnode -a ${RG2}.tu.com
/usr/es/sbin/cluster/utilities/claddnode -T service -B ${NODE1}.tu.com -w net_ether_01
/usr/es/sbin/cluster/utilities/claddnode -T service -B ${NODE2}.tu.com -w net_ether_01
/usr/es/sbin/cluster/utilities/claddnode -T service -B ${RG1}.tu.com -w net_ether_01
/usr/es/sbin/cluster/utilities/claddnode -T service -B ${RG2}.tu.com -w net_ether_01

################################################################
# Create Application Servers
for APPLS in ${RG1APPLS} ${RG2APPLS}
do
/usr/es/sbin/cluster/utilities/clrmserv ${APPLS}
/usr/es/sbin/cluster/utilities/claddserv -s ${APPLS} -b "/usr/local/hascripts/${APPLS}_start.sh" -e "/usr/local/hascripts/${APPLS}_stop.sh"
done

################################################################
# Create persistant addresses
/usr/es/sbin/cluster/utilities/clrmnode -a ${NODE1}-pers
/usr/es/sbin/cluster/utilities/clrmnode -a ${NODE2}-pers
/usr/es/sbin/cluster/utilities/claddnode -a ${NODE1}-pers.tu.com :ether :net_ether_01 : :persistent : : -n ${NODE1}
/usr/es/sbin/cluster/utilities/claddnode -a ${NODE2}-pers.tu.com :ether :net_ether_01 : :persistent : : -n ${NODE2}

################################################################
# Create resource groups
/usr/es/sbin/cluster/utilities/clrmgrp -g ${NODE1} 
/usr/es/sbin/cluster/utilities/clrmgrp -g ${NODE2} 
/usr/es/sbin/cluster/utilities/clrmgrp -g ${RG1}
/usr/es/sbin/cluster/utilities/clrmgrp -g ${RG2}

/usr/es/sbin/cluster/utilities/claddgrp -g ${NODE1} -n "${NODE1} ${NODE2}" -S OHN -O FNPN -B FBHPN
/usr/es/sbin/cluster/utilities/claddgrp -g ${NODE2} -n "${NODE2} ${NODE1}" -S OHN -O FNPN -B FBHPN
/usr/es/sbin/cluster/utilities/claddgrp -g ${RG1} -n "${NODE1} ${NODE2}" -S OHN -O FNPN -B NFB
/usr/es/sbin/cluster/utilities/claddgrp -g ${RG2} -n "${NODE2} ${NODE1}" -S OHN -O FNPN -B NFB

################################################################
# Add the Service address to the resource group
/usr/es/sbin/cluster/utilities/claddres -g "${NODE1}" \
  FALLBACK_AT= \
  SERVICE_LABEL="${NODE1}" \
  APPLICATIONS= \
  VOLUME_GROUP= \
  FORCED_VARYON="false" \
  VG_AUTO_IMPORT="false" \
  FILESYSTEM= \
  FSCHECK_TOOL="fsck" \
  RECOVERY_METHOD="sequential" \
  FS_BEFORE_IPADDR="false" \
  EXPORT_FILESYSTEM= \
  MOUNT_FILESYSTEM= \
  NFS_NETWORK= \
  SHARED_TAPE_RESOURCES= \
  DISK= \
  AIX_FAST_CONNECT_SERVICES= \
  COMMUNICATION_LINKS= \
  MISC_DATA=

/usr/es/sbin/cluster/utilities/claddres -g "${NODE2}" \
  FALLBACK_AT= \
  SERVICE_LABEL="${NODE2}" \
  APPLICATIONS= \
  VOLUME_GROUP= \
  FORCED_VARYON="false" \
  VG_AUTO_IMPORT="false" \
  FILESYSTEM= \
  FSCHECK_TOOL="fsck" \
  RECOVERY_METHOD="sequential" \
  FS_BEFORE_IPADDR="false" \
  EXPORT_FILESYSTEM= \
  MOUNT_FILESYSTEM= \
  NFS_NETWORK= \
  SHARED_TAPE_RESOURCES= \
  DISK= \
  AIX_FAST_CONNECT_SERVICES= \
  COMMUNICATION_LINKS= \
  MISC_DATA=

################################################################
# Add application servers and volumes to resource groups
/usr/es/sbin/cluster/utilities/claddres -g "${RG1}" \
  FALLBACK_AT= \
  SERVICE_LABEL="${RG1}" \
  APPLICATIONS="${RG1APPLS}"  \
  VOLUME_GROUP="${RG1VGS}" \
  FORCED_VARYON="false"  \
  VG_AUTO_IMPORT="false"  \
  FILESYSTEM=  \
  FSCHECK_TOOL="fsck" \
  RECOVERY_METHOD="sequential"  \
  FS_BEFORE_IPADDR="false"  \
  EXPORT_FILESYSTEM= \
  MOUNT_FILESYSTEM=  \
  NFS_NETWORK=  \
  SHARED_TAPE_RESOURCES=  \
  DISK= \
  AIX_FAST_CONNECT_SERVICES=  \
  COMMUNICATION_LINKS=  \
  MISC_DATA= 

/usr/es/sbin/cluster/utilities/claddres -g "${RG2}" \
  FALLBACK_AT= \
  SERVICE_LABEL="${RG2}" \
  APPLICATIONS="${RG2APPLS}" \
  VOLUME_GROUP="${RG2VGS}" \
  FORCED_VARYON="false" \
  VG_AUTO_IMPORT="false" \
  FILESYSTEM= \
  FSCHECK_TOOL="fsck" \
  RECOVERY_METHOD="sequential" \
  FS_BEFORE_IPADDR="false" \
  EXPORT_FILESYSTEM= \
  MOUNT_FILESYSTEM= \
  NFS_NETWORK= \
  SHARED_TAPE_RESOURCES= \
  DISK= \
  AIX_FAST_CONNECT_SERVICES= \
  COMMUNICATION_LINKS= \
  MISC_DATA=

Manually select and configure the concurrent disk heartbeat, then synchronize and start the cluster.
Oracle LPAR
smitty cm_config_hacmp_communication_interfaces_devices_menu_dmn smitty cm_ver_and_sync.select

Mount the Oracle CD's from the MDC NIM server

Oracle LPAR

mkdir /Oracle9201_Disk1
mkdir /Oracle9201_Disk2
mkdir /Oracle9201_Disk3
mkdir /Oracle9201_Disk4
mkdir /Oracle9206_Disk1
mkdir /Oracle9207_Disk1

mount mdcapnim01:/export/cdimages/Oracle9201_Disk1 /Oracle9201_Disk1
mount mdcapnim01:/export/cdimages/Oracle9201_Disk2 /Oracle9201_Disk2
mount mdcapnim01:/export/cdimages/Oracle9201_Disk3 /Oracle9201_Disk3
mount mdcapnim01:/export/cdimages/Oracle9201_Disk4 /Oracle9201_Disk4
mount mdcapnim01:/export/cdimages/Oracle9206_Disk1 /Oracle9206_Disk1
mount mdcapnim01:/export/cdimages/Oracle9207_Disk1 /Oracle9207_Disk1

Oracle Server Build

FREE Domain Registration
included with Web Site Hosting
Tools, Social Networking, Blog
www.siteox.com

Business Web Site Hosting
$3.99 / month includes Tools,
Shopping Cart, Site Builder
www.siteox.com

*AIX^tm^mtXIA*	Site Guide \| Contact Us Mt Xia: *Technical Consulting Group*
	Business Continuity / Disaster Recovery / High Availability Data Center Automation / Audit Response / Audit Compliance


	Home About Mt Xia Mt Xia Officers Privacy Policies FAQs Contact Us Events Press Releases Business Continuity Business Continuity Planning BC Auditing BC Policies BC Guidelines BC Standards BC Procedures BC Automated Deployment Disaster Recovery Disaster Recovery Planning DR Plan Template DR Project Implementation DR Auditing DR Automated Deployment High Availability High Availability Planning HA Policies HA Guidelines HA Standards HA Procedures HA Automated Deployment Tech Support/Training AIX 4.3 Tech Support AIX 5.1 Tech Support AIX 5.2 Tech Support AIX 5.3 Tech Support HACMP 4.4.1 Tech Support HACMP 5.2 Tech Support HACMP 5.3 Tech Support Disaster Recovery Unix / Linux Shell Programming E-Business Downloads Scripts Editors Documentation Standards Project Template E-Business Methodology E-Business Model Random Fortune Cookie