[snakes1100]

I would agree with Marco, you should look at what is actually causing the server to suck up all your resources during the backup time, log into the server via ssh before the cron starts and try to find out what pid (s) are doing the damage.

You could temporarily use this as well to do what you asked. You will need to place this in a file and set a cron to run it 15 seconds before the db dump starts, you will also have to make a dupe conf file so as Paul points out, your visitors get a message as to whats going on. Set the new conf file to point at default index.html for all calls to apache on the server.

Simple & crude, should work for your needs.

#!/bin/sh
/usr/local/apache/bin/apachectl stop
sleep 30
/usr/local/apache/bin/apachectl -f /usr/local/new.conf start
sleep 128000
/usr/local/apache/bin/apachect stop
sleep 30
/usr/local/apache/bin/apachectl start