pfelt at feltonline.com
Wed Apr 26 04:40:03 MEST 2006
i'm looking for more information on the holt-winters forecasting
algorithm. specifically, how does it constitute a failure? i've got an
rrd in which i've created holt-winters rras. it is functioning great in
drawing the confidence bounds between which the data should fall (as
described in: http://cricket.sourceforge.net/aberrant/rrd_hw.htm
the only problem that i'm having is the failures. i've plotted the
graph for today and at about 12:00pm the data spiked to about 4 Mbit.
the upper confidence bound is at about 3Mbit. I checked the rrd via a
dump and at that time the failure rra does not show a failure at that time.
did i do something wrong here?
Unsubscribe mailto:rrd-users-request at list.ee.ethz.ch?subject=unsubscribe
Help mailto:rrd-users-request at list.ee.ethz.ch?subject=help
More information about the rrd-users