A Statistical Traffic & Bandwidth Capacity Analysis is a detailed investigation of your WAN
links that leverages an advanced analysis tool developed by PacketIQ
to provide an extremely accurate statistically-derived accounting of network usage levels
and patterns across the 24 hour business day based on multiple-day bandwidth data
exported from your existing SNMP or NetFlow/sFlow Network Management Systems.
Contact PacketIQ today to get started!
The major SNMP and NetFlow/sFlow Network Management Systems and other sample data sources currently supported include:
* All trademarks referenced above belong to their respective companies.
The PacketIQ analyzer can be adapted to import bandwidth usage data from almost
any NMS system that offers a Comma Seperated Value (.csv) export format.
One of the most useful features of a Statistical Traffic & Capacity Analysis is the Time of Day Analysis reports,
which provide a graph and statistical values of usage over a 24 hour period - statistically derived from multiple days of sample data.
This offers an easy to comprehend visualization of typical daily usage patterns and levels without having to try to
discern these by eye-balling the numerous daily spikes found in typical graphs.
A Merge feature combines data from multiple links to measure total loading (useful for company merger modeling),
redundant link failure loading, or combining data for multiple time periods into one report.
The Multi-Link feature combines usage data from sequential WAN links
to determine true bandwidth availability across multiple hops - essential when loading varies
signficantly across links and/or time zones.
The reported link speed and calculated utilization levels can be easily adjusted within the PacketIQ analyzer.
This is very convenient if the bandwidth data samples are being collected from an attached router interface (100 Mbps or 1 Gbps) instead of directly
from the WAN link port (which may be 6 Mbps, for example).
View bandwidth usage levels in local time for a given user location - not the timezone for the NMS collector.
This is essential for accurate Time of Day and usage pattern analysis.
Sample data can be filtered by date, day of week, or time of day to highlight or exclude traffic periods
and patterns of interest. For example, it may be desirable to filter 30 days of sample data to only report
on the periods of 6AM to 6PM Monday-thru-Friday for more accurately measuring business day levels, or to specify
settings to exclude a periodic backup period from the analysis.
The average and peak values provided by most NMS solutions are practically useless for representing bandwidth usage levels.
PacketIQ reports offer Min-Max-Average-Standard Deviation-90th-95th-99th Percentile values for each data flow direction.
Percentile values, preferably 95th or 99th, combined with appropriate filtering, are the best representation of
nominal usage levels across a large data set.
Percent Change, Slope, Y-Intercept, and First/Last trend values
across the imported data set support long-term trending analysis & forecasts.
All of the parameters, selected filters, bandwidth and Time of Day statistical values,
trend data, and 'standardized' sample data for each WAN link is saved into .csv formatted files for archive, historical trending,
and other analysis functions.
Standard and Time of Day Analysis reports can be provided in .pdf documents or .bmp, .jpg, and/or .png formats.
Usage levels for all of the individual LAN/WAN links included in the analysis are summarized in a percent utilization graph and accompanying table that includes
specific values and Red-Yellow-Green link status indicators for quick interpretation.
This top-level summary is followed by the individual standard and Time of Day statistical analysis reports for each link.
Excerpt from a
Bandwidth Statistical Analysis Report Click for a larger image
© Copyright 2005-2013 PacketIQ Inc. All rights reserved.