I have just started to shell script a few months ago and have written the script below, which works well for what its for. However I would like it if i did not have to ssh twice to the server for each server in the list ( way more then listed in script ).
Basically the script searches through a sendmail log for the
( COMMAND="$1" ) string once found i have it awk for the message ID, and remove any messages that are not complete. Than it dumps the ID's into a file which i ssh and grep the maillog file a second time for the messge ID. Now with the messge ID i have all the logs associated with the message put together seperated by the "--------" lines.

#!/bin/ksh
######################################################################
#
# Description:
# A simple script to search through all mailservers in the list below
# for a specified string input as a command line argument.
#
######################################################################
# written by:Me
######################################################################

if [[ $# -eq 0 ]];then
print "Script syntax:
./mail-script <search_string>"
exit
fi
COMMAND="$1"
echo Searching all mailhubs for an instance of: "$COMMAND"
echo ""

MAILSERVER="mailserver-1 mailserver-2 mailserver-3 mailserver-4"

for I in $MAILSERVER
do
# incase file exists from fase start, remove to start fresh
rm $HOME/hub-list 2>/dev/null
echo " "
echo "*********************************************************"
echo " $I --- $COMMAND "
echo "*********************************************************"
echo " "
ssh root@$I "grep -i $COMMAND" /log/maillog | \
awk '{print $9}'|grep -v connect |grep -v disconnect >>
$HOME/hub-list
while read -u3 hub
do
ssh root@$I grep $hub /log/maillog
echo " "
echo "--------------------------------------------------"
echo " "
done 3<$HOME/hub-list
done
echo " @@@@@ The Script has Finished @@@@@ "


I have been toying with trying to keep it all in the same command line, can this be done. My example below does not work but you can see what i was heading
grep -i Something /log/maillog | awk '{print $9}'|grep $9 /log/maillog

Any help would be appreciated, as i am still learning.