I am capturing latency on my router interfaces and keeping/graphing Min, Avg, Max, and MDev for RTA in RRD.
Most of my routers are averaging 30-50ms on RTA. Some of my very remote routers (e.g China) run at an average RTA of 175-200ms. I have an enterprise application that has a "perceived user experience" that goes down the chute when the latency hits a sustained spike above the steady state level. Rather than create a single value threshold on the top-most template and clone and copy down to the exception routers, I would like to have an expression that is sensitive to the steady state of the individual routers. I have in mind something like Max Threshold = Average Avg RTA + X * Average MDev RTA where X is some multiple of the mean deviation. However these values do not appear to be available to a Python expression in the max threshold field since these are values being generated by RRD. Has anybody done anything like this? -------------------- m2f -------------------- Read this topic online here: http://forums.zenoss.com/viewtopic.php?p=27003#27003 -------------------- m2f -------------------- _______________________________________________ zenoss-users mailing list [email protected] http://lists.zenoss.org/mailman/listinfo/zenoss-users
