Apache Debug

You might be knowing it, but I would like to summarize the steps again to take Debug for Apache, believe me set this and you can easily resolve the errors:

1. Shut the HTTP Server (Apache/iAS) down. - You can grep for the 'httpd' process to verify it is down
2. Rename (or delete) the following files so we get a fresh copy of them:
/Apache/Jserv/logs/jserv.log
/Apache/Jserv/logs/mod_jserv.log
/Apache/Jserv/logs/jvm (the whole directory if it exists)/Apache/Apache/logs/access_log*
/Apache/Apache/logs/error_log
/Apache/Apache/logs/error_log_pls (if it exists)

Now we will turn on debugging in the log files:

3. Modify the $IAS_ORACLE_HOME/Apache/Jserv/etc/jserv.conf file. Search for the parameter:
ApJServLogLevel notice
Change the 'notice' to 'debug'

4. Modify the $IAS_ORACLE_HOME/Apache/Jserv/etc/jserv.properties file. Search for the following section:

log=false
Change this to be log=true
and Changelog.channel=false
To be: log.channel=true

5. Modify the $IAS_ORACLE_HOME/Apache/Apache/conf/httpd.conf file. Search for:
LogLevel
Set the LogLevel to 'debug' from its current value.

6. Modify the $IAS_ORACLE_HOME/Apache/Apache/conf/httpd_pls.conf file (if it exists). Search for the following parameter
LogLevelSet the LogLevel to 'debug' from its current value.

7. In the zone.properties...- servlet.framework.initArgs=debuglevel=1

8. Start the HTTP Server

Happy Troubleshooting !!!

Finding Huge files

Yes its a big tasks for DBA's, when you do df -k in solaris and see your mount points are reaching 100%. It difficult to find what to delete for what to compress. I usually try some commands mentioned below :

This command will list top 10 directory from parent directory
du -ks * sort -rn head -10

This command will list out all files greater than mentioned size.
find . -size +10240

Keep Troubleshooting !!!

Old Oracle Apps 11i Looks

Hey you might be wondering what I am going to talk now. Yup its about oracle Apps front end look and feel. Some of the users still using same old grey and blue looks and DBA wonder if he missed something. No not at all, just check these two profiles and you are done.

Change the following profiles at user level first-
"Java Look and Feel " to oracle and
"Java Color Scheme" to blaf

If it works for you then change it Site level to have others enjoy the same.!!!

Keep Troubleshooting !!!

Script for Checking Java Mailer Status

---------------------------------Script Start-----------------------------------
#!/bin/ksh
#Script to monitor WF Mailer Status
. .env
cd /tmp
sqlplus /nolog << EOF
connect apps/password
set heading off
set pages 0
set feedback off
spool monitor_mailer.lst
select COMPONENT_STATUS from FND_SVC_COMPONENTS
where COMPONENT_ID='10006';
spool off
EOF
MONITOR=`grep -v "SQL>" monitor_mailer.lstsed s/\ //g`

if [ $MONITOR != "RUNNING" ] ; then
echo $MONITOR mailx -s "WFMAILER: Check Workflow Mailer STATUS" sudubey@gmail.com
fi
/bin/rm -f /tmp/monitor_mailer.lst
-----------------------Script End -------------------------------------------------

Put this in crontab for every 10 min or so. It will send a status whenever its not running. No more OAM checking !!!!

Note: This has been tested in Sun Solaris

Keep Troubleshooting !!!

Apps Password changed from sqlplus

Many of you might have done this mistake, or no option left like in my client usually happen in every six month, it force you to change apps password via sqlplus. Now what application login will stop working, isn't it. Yes !!

Do this :
1. Change to old apps password via sqlplus;

SQL> Alter user apps identified by oldappspasswd;

2. Run autoconfig with this password in all tiers.

3. Restore customizations if any.

4. Start all services and you all set to login to your apps frontend login.

Hope this helps you all.

Keep Troubleshooting !!!