mysql backup script

[Steve Buehler]

I am having a problem with a backup script that is written in a shell 
(/bin/sh) script to backup my mysql databases.  For some reason on any day 
with an even number I get the following error:
MySQL could not be stopped, exiting...
It is really weird because it will run on odd number days without a 
problem.  Should I put some kind of a wait in the script after it stops the 
mysqld and before it checks to make sure the pid file is still there?  If 
so, does anybody know how?  I am not sure that that will solve the problem 
though because it exits afterward and the mysql daemon is still running 
without having to restart it.  Below is the relevant part of the script.
The script is run from cron with this line
0 1 * * * /root/backup/backup.sh >/dev/null 2>&1
But since the logs do show it running, That shouldn't be the problem.
Thanks
Steve

#########  Perform myisamchk
#mysqladmin -p$MYSQLPWD shutdown
/etc/rc.d/init.d/mysqld stop
if [ -f $mysqlpid ]; then
         echo -e "MySQL could not be stopped, exiting... " >> 
/tmp/backuplog$date
         exit 1
else
         echo -e "Checking MYI tables for all databases " >> 
/tmp/backuplog$date
         myisamchk --silent --force --fast --update-state -O key_buffer=64M \
         -O sort_buffer=64M -O read_buffer=1M -O write_buffer=1M \
         /var/lib/mysql/*/*.MYI
fi
echo -e "Checks complete, starting MySQL " >> /tmp/backuplog$date
/etc/rc.d/init.d/mysqld restart
sleep 20
if [ ! -f $mysqlpid ]; then
         echo -e "MySQL failed to start, exiting... " >> /tmp/backuplog$date
         exit 1
fi
##########