We're monitoring a local Jetty (Java webserver) process using an application status page. When everything's going well, it includes the string "OK", which we check for. This should be a clearly successful status test. When everything's not going well, we get some sort of 4xx or 5xx error message. This should trigger alerts immediately. When some things are going well and others aren't fully up to speed (slow database), we'll get a "DATABASE_TEST_RAN_LONG", which isn't ideal, but at least for a few occurances (n <= 5) we can live with. In particular, we DON'T want a single result sounding off pagers in the middle of the night. The current test looks like: define command{ command_name check_jetty command_line /usr/lib/nagios/plugins/check_http -H '$HOSTADDRESS$' -u /serviceStatus -e 200 -s OK } What would be a sane process of getting Nagios to: - Report all clear when we get a 200 status and "OK" text on page? - Wait for 6 consecutive instances of "DATABASE_TEST_RAN_LONG" before alerting for that result. - Alert immediately on any cases not matching one of the above? I don't believe we can capture this in a single test unless I'm missing something. Thanks in advance. -- Dr. Ed Morbius Chief Scientist Krell Power Systems Unlimited