Analysing Incidents

It is import to know when an Incident occurs, but further knowledge as to the cause of a fault will help resolve it. Within the web interface you have access to many graphs to help gain information regarding a device.

After you have navigated to the Metric entity that the Incident refers to, a series of graphs will be displayed.

The top half of the Metric page shows Current Data (previous 48 hours).

The lower half of the window shows Historic Data. There are graphs for: 48 Hours up to, Week of, Month of, and Year of, all relative to a reference date. The reference date can be set at the bottom of the window.

For Example:

The LithiumCorp North Sydney office has been reporting inttermittant ADSL dropouts. Upon investigating the device, the availability looks normal.

By selecting the ADSL Circuit 'ATM1/0' we can view the current and historic state of the ADSL circuit. Straight away you can see there is a problem with the 'Upstream Noise Margin'.

Scrolling down the page you can see this has been an issue throughout April, and the very last graph shows the interference began around the end of March.

By setting the reference date to a date at the end of March (30th), you can see that at around 8pm the Upstream capacity drops from 24dB down to 7dB.

Changing the Metric (at the bottom of the page) to 'Upstream Capacity Used' (A measure of the current link speed /  Maximum theoritacal link speed for the line) you can see that it jumps from 34% almost instantaenously to 94%. The Downstream Capacity graph also shows a similar trend.

These results indicate that an unfiltered device may have been added on the ADSL line at around 8pm on the 30th March. After further investigation with the site, it was found that there was a fax machine added on the line around this time.