Additional thought... by watching the client-side response time, we could declare a server outage when a watcher determines that the response times are greater than some statistical mean. For example, if the upper limit is 3 sigma above the 20 minute rolling average (and the variance is suitable for this -- IMHO likely) then a client-based watcher thread could raise a flag.
-- richard This message posted from opensolaris.org _______________________________________________ perf-discuss mailing list perf-discuss@opensolaris.org
