On Wed, May 15, 2013 at 1:38 PM, Thomas Johansson <thomasj at isy.liu.se> wrote: > On 2013-05-15 17:44, Matt wrote: >> When I run them all at once they bog the system and cause some of >> latency graphs on equipment being monitored to look poor. > > Wait randomized interval of 60 seconds and start > > perl -e "sleep(int(rand(60)))" && (perl /scripts/my_script.pl) Also you can add a wait to the script after starting some number of background commands to wait until they have all completed before you start another batch. But, rather than using so many home-grown scripts, maybe you should look at OpenNMS, cacti, etc. for monitoring frameworks to manage it all for you. -- Les Mikesell lesmikesell at gmail.com