From: Iain Hallam <iain.hallam@bristol.ac.uk>
I've now read the code for Monitor.pm and Handling Thresholds (?) and, from
my current understanding, it seems that the current value ($value) can be
compared with one previous value ($cmp_value). If the old value is always
set to be 24 hours ago, but is not typical of that time while the current
one is, the difference between the two values will go beyond the threshold
and trigger an alarm. What would be more useful is a way of aggregating the
data from the last day with previously collected data, into a rolling set of
average data points for an average 24-hour period. This data-source would
then become better, over time, at showing the average traffic (for example)
over a 24-hour period and alarms would be triggered less often. It might be
neccessary to have a longer data span in order to account for week-ends,
when the devices would be used less, and to find some way to expire old data
from the averages so that trends can be monitored.
If I've got this completely wrong, I'd appreciate some feedback, and I'd
also like to know about the current state of monitoring alarms and things.
If anyone has been working on it since 0.70, could you email me (or the
list), please?
Many thanks from a newbie cricketer :)
Iain.
-- Network Supervisor University of Bristol Computing Service iain.hallam@bris.ac.uk--------------------------- ONElist Sponsor ----------------------------
Hey Freelancers: Find your next project through JobSwarm! You can even make $$$ in your sleep by referring friends. <a href=" http://clickme.onelist.com/ad/jobswarm1 ">Click Here</a>
------------------------------------------------------------------------
This archive was generated by hypermail 2b29 : Mon Mar 06 2000 - 19:01:06 PST