Hi,
I am new to DB2 adminstration..As far as i know these are the steps to take to tune a bad SQL::
1)Use Visual explain and analyse path
2)Use db2advisor for index suggestion
3)Update statistics

Could you please elaborately tell me as to how to tweak any bad SQL..Thanks in advance

If you want a generic answer, you may be better off reading a book ...

Sathyaram

Reading a book? I guess that would be a good way to pass the time while waiting for the SQL to run.

There are a few rules-of-thumb, of course, like avoiding stage-2 predicates in "sensitive" places, or avoiding multiple scans through large tables. Sometimes, it's just a matter of rewriting the SQL query (e.g., join instead of subquery or vice versa) and trying out which one takes longest to terminate ;-)

Thanks for the help.

Other tips -
1) Understand your database schema and data (as much as possible)
2) Understand the DB2 optimizer and various switches (configuration parameters)
3) Understand what the query (or SQL statement) is trying to do
4) Make use of the tools available
5) Use DB2's advisor (its free !)

Here's what I do for my DB2 v8.2.4 environments (actual script) -
==============================
#!/bin/ksh

PROGNAME="`basename $0`"
if [[ ! $# -eq 2 ]]
then
echo ; echo ; echo ;
echo "$PROGNAME:2 Parameters required"
echo "Usage: $PROGNAME <dbname> <query-file>"
echo ; echo ; echo ;
exit 1
fi

DBNAME=$1
QUERY_FILE=$2

db2 connect to $DBNAME
db2 set current explain mode explain
db2 -tvf $QUERY_FILE
db2 terminate
db2exfmt -1 -d $DBNAME -o ${QUERY_FILE}.explain
db2 terminate
db2advis -d $DBNAME -i ${QUERY_FILE} -m MICP -b SYSTOOLSPACE -c SYSTOOLSPACE -f -o $QUERY_FILE.db2advis.sql | tee $QUERY_FILE.db2advis
db2 terminate
==============================