It is important that the length of time and the baseline information being
gathered are sufficient to establish a typical picture of the network. This
period should be at least seven days to capture any daily or weekly trends.
Weekly trends are just as important as daily or hourly trends.
As an
example, the engineers in Building 7 run a massive backup and software refresh,
on Sundays at 2 A.M., on all 200 of their workstations. Because the backup
server happens to be in Building 9, the push and backup saturate the corporate
backbone. If the baseline was only performed from Monday through Friday, the
saturation would be missed.
A baseline needs to last no more than six
weeks unless specific long-term trends need to be measured. Generally, a
two-to-four-week baseline is adequate.
Examples of baselines for CPU utilization over various intervals of time can
be provided by the Multi Router Traffic Grapher (MRTG).
This tool can be
found at http://mrtg.hdl.com/mrtg.html.
It is recommended that baseline measurement is not performed during times of
unique traffic patterns. For example, it would not be beneficial to perform the
baseline over a holiday or during December if most of the company is on
vacation.
Subsequent baselines should provide varied durations and
timelines. This will allow discovery of the optimal analysis collection
interval for the network.