[udayperumal]

Hi,

I am facing a problem in one of my aix machine running aix 5.3 ML04 SP03 with oracle database 9i and configured in HACMP V5.3 with latest patches.
Cluster is configured in Active/passive mode and there is problem with passive server.

When i am checking ps aux command there more than 6000 defunt process running and owner of the process are root and taoracle.

After reboot the no. of defunt process getting reduce to less than 30, but owner of those process is only taoracle. Because of this preformance problem and system is behaving abnormally.

example:
root 4871128 1 8 0:00 <defunct>
taoracle 4879216 1 3 0:00 <defunct>
root 4883454 1 8 0:00 <defunct>

Need to know why this defunt process are generating and what is the RCA for finding the process which is generating this.

I have pasted the inittab output below for reference. If any output is required pls do let me know...

#/etc/inittab.
==============
naudio::boot:/usr/sbin/naudio > /dev/null
logsymp:2:once:/usr/lib/ras/logsymptom # for system dumps
perfstat:2:once:/usr/lib/perf/libperfstat_updt_dictionary >/dev/console 2>&1
diagd:2:once:/usr/lpp/diagnostics/bin/diagd >/dev/console 2>&1
ctrmc:2:once:/usr/bin/startsrc -s ctrmc > /dev/console 2>&1
ha_star:h2:once:/etc/rc.ha_star >/dev/console 2>&1
clcomdES:2:once:startsrc -s clcomdES >/dev/console 2>&1
lft0:2:respawn:/usr/sbin/getty /dev/lft0
ntbl_reset:2:once:/usr/bin/ntbl_reset_datafiles
rcml:2:once:/usr/sni/aix53/rc.ml > /dev/console 2>&1
naviagent:2:wait:/etc/rc.agent > /dev/console 2>&1
rcemcpower:2:wait:/etc/rc.emcpower set_ipldevice > /dev/console 2>&1
cons:0123456789:respawn:/usr/sbin/getty /dev/console
clinit:a:wait:/bin/touch /usr/es/sbin/cluster/.telinit # HACMP for AIX These must be the last entries of run level a in initt
ab!
pst_clinit:a:wait:/bin/echo Created /usr/es/sbin/cluster/.telinit > /dev/console # HACMP for AIX These must be the last entri
es of run level a in inittab!
tty0:2:off:/usr/sbin/getty /dev/tty0
sshtadb:2:respawn:su taoracle "-c /uhome/taoracle/dba/scripts/start_ssh.tadb -9 -u"
sshoasis:2:respawn:su taoracle "-c /uhome/taoracle/dba/scripts/start_ssh.oasis -9 -u"
altiris:235:once:/etc/rc.altiris start
xmdaily:2:once:/usr/bin/xmwlm -L 2>&1 >/dev/null # Start xmwlm daily recording

[mshulman1980]

Good morning... I've seen this before with Oracle 9i as well. We were able to kill the defunct processes without any apparent impact on the server. The problem as far as I could tell was that we had a lot of poorly written applications that were not closing their db connections properly.
Here's the script that we used to kill off the dedunct processes:
ps -ef | grep defunct| cut -c10-16 > /tmp/a
for i in `cat /tmp/a`
do
kill $i
done
rm /tmp/a