Link: Home Link: NMS MRTG Link: Portal Library

Network / Server Availability Report

Four reports are presented below. The reports show overall availability and downtime for the ATM switches, VLANs (network segments), and servers on the network. Please read the note below before viewing the reports.

2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000

The detail section of each report has two collums per type of device or network segment. Devices or segments are listed by name with their cumulative down time next to the name. Down times are shown in seconds. This is done to make math easier.

The summary section of each report (at the top) shows the number of days in the report period, number of devices monitored, the number of "device seconds", and the total down time for all devices. "Device seconds" is the number of seconds in that period times the number of devices managed. It is a measure of the maximum amount of time that a device could be available during that time period. Last is the percent availability. Percent availability is computed using the following formula:

device seconds - total down
100
= availability %

The overall average is computed similarly across all time periods. The averages present availability in a 7x24 working environment. Attempting to separate events based on working hours vs. off hours is simply too difficult with the tools available. To get a "worst case" working hours availability figure, run the formula given above with double the number of down seconds. This will give an availability percentage with all down time incidents effectively placed within the 7 AM - 7 PM window we consider "working hours".

A note about these statistics: These statistics are not derived from Divine Truth. They are approximations calculated as carefully and precisely as possible. These statistics are good enough for duscussions of trending and general availability. They are not an absolute accounting. There are several ways in which the statistics could be influenced. One station on the network captures the availability data; if some thing between that station and a managed device is down, then that managed device will be reported as down when it may not be. Also, the report generation algorithm will throw out corrupt data, so downtime events might not be counted. Examination of the report generator error log shows that this happens periodically. In summary, these statistics are a good representation of what is going on in the network but they are not a perfect representation.




Metro Hartford Information Services | Updated 3-1-10