Hi, While testing and deploying Nagios I used the following service definition: max_check_attempts 1 normal_check_interval 1 retry_check_interval 1 Everything was configured and tested including nagiosgrapher (rrdtool). It passed proof and concept and we have now moved it into production. We tested this against a small set of 200 servers. Then we went to our "production" values of: max_check_attempts 3 normal_check_interval 5 retry_check_interval 1 Data collection still works, but when we try to view a graph it's basically empty data (nothing drawn) for the period in which the new normal_check_internal was changed. If we change it back to 1 from 5, it starts to display data correctly again. How do I *keep* the new normal_check_internal value, but make it work again? Googling led me to this: <a href="http://forum.centreon.com/archive/index.php/t-4506.html">http://forum.centreon.com/archive/index.php/t-4506.html</a> which essentially says nagios.cfg:normal_check_interval and nrpe.cfg:step are tied together and if you change normal_check_internal you need need to lose all your data. The existing data isn't super critical. We're happy to lose it. I tried removing ./var/rrd directory (well rename) but then all the graphs broke with 'no such device' error. What other directories do I need to delete to clear all nagiosgrapher/rrdtool data? Cheers