System Director 6.3 in Cluster Environment IBM System Mirror 7.1.3

think, today you as i, wanted  to put ur System Director into  fresh cluster.

some easy preparation steps:

0. Install ISDirector on first node

1. propagate between nodes this directories:

/etc/ibm/director/*

          /etc/tivoli/cce/*
and file:
          /etc/rc.d/init.d/smserver
   

check the file  /etc/tivoli/cce/*/cce.ini   for CCE_Config_HomeDirectory var that should be point to  /opt/ibm/director/data/tivoli/cce/*/config
   
2. all nodes must have this users:

dirinst1:!:232:101:IBM Systems Director Database User:/home/dirinst1:/usr/bin/ksh

and groups:

db2iadm1:!:101:dirinst1
smadmin:!:205:root
smmon:!:206:
smmgr:!:207:
smuser:!:208:

3. Check db2 autostart with command:

/opt/ibm/director/db2/instance/db2iauto -on dirinst1

4. Define LIBPATH in the Systems Director startup script /etc/rc.d/init.d/smserver, like this:

# For Systems Director server during system startup
LIBPATH=/home/dirinst1/sqllib/lib64/:$LIBPATH
export LIBPATH

5. add all cluster nodes to file /home/dirinst1/sqllib/db2nodes.cfg

0 sysdir01 0
1 sysdir02 0

===>> AND FAIL!! after this step i cant run db2 << need fix this issue

sync between nodes file /etc/services or add on every node this lines to /etc/services :

DB2_dirinst1    60010/tcp
DB2_dirinst1_1  60011/tcp
DB2_dirinst1_2  60012/tcp
DB2_dirinst1_END        60013/tcp
db2c_dirinst1   50010/tcp

6. Cluster START and STOP scripts:

START:

#!/usr/bin/bash -x

#hello, ima start scipt for IBM System Director 6 in Cluster environment
#exit 0
log=/tmp/ha.log
hn=`/usr/bin/hostname`

date | tee -a $log

#cat /home/dirinst1/sqllib/db2nodes.cfg
#0 sysdir01 0

echo $hn | tee -a $log

echo "0 sysdir01 0" > /home/dirinst1/sqllib/db2nodes.cfg

if [ $hn == "sysdir02" ] ; then
        echo "0 sysdir02 0" > /home/dirinst1/sqllib/db2nodes.cfg;
fi

          su - dirinst1 -c "db2start" | tee -a $log
          su - dirinst1 -c "db2 activate database DIRDB" | tee -a $log

/opt/ibm/director/bin/smstart

SMSTAT=`/opt/ibm/director/bin/smstatus`
echo System director is $SMSTAT >> $log
#Inactive


while [ "$SMSTAT" != "Active" ]; do

        date | tee -a  $log
        SMSTAT=`/opt/ibm/director/bin/smstatus`
        echo System director is $SMSTAT | tee -a $log
        su - dirinst1 -c "db2 list active databases" | tee -a $log
        su - dirinst1 -c "db2 list applications" | tee -a  $log
        sleep 10

done

su - dirinst1 -c "db2 list active databases"  | tee -a $log

#su - dirinst1 -c "db2start"  >> $log
#su - dirinst1 -c "db2 activate database DIRDB"  >> $log

STOP:

#!/usr/bin/bash -x
#hello, ima stop scipt for IBM System Director 6 in Cluster environment
#exit 0

log=/tmp/ha.log
date | tee -a  $log
/opt/ibm/director/bin/smstop

SMSTAT=`/opt/ibm/director/bin/smstatus`
echo System director is $SMSTAT >> $log
#Inactive
su - dirinst1 -c "db2 list active databases" | tee -a $log
su - dirinst1 -c "db2 list applications" | tee -a  $log
su - dirinst1 -c "db2 terminate" | tee -a  $log
su - dirinst1 -c "db2 deactivate database DIRDB" | tee -a $log
su - dirinst1 -c "db2stop" | tee -a $log

IPCSJUNK=`/usr/bin/ipcs | grep -i dirinst1 | awk '{print $2}' | xargs`
for i in $IPCSJUNK;do
        echo remove IPC $i | tee -a $log;
        /usr/bin/ipcrm  -q $i
done
ps -ef | grep -i db2 | tee -a $log
/usr/sbin/slibclean

OS AIX 7.1 of course