noticeboard.ru.ac.za

The recent spate of frequent power outages is playing havoc with the University's network, and by implication, the telephone system.

Each time power is restored we find that there are a number of areas of campus in which the network has not recovered. The causes for this range from corrupted configurations to physical damage to our infrastructure.

You do not need to report such building-wide network outages to us -- our network monitoring system automatically detects these and notifies us. You can see the status of most buildings at http://bb.ru.ac.za/switch/switch.html

However, you should be aware that it sometimes takes us several hours to visit all the affected portions of campus and attend to the faults. The more faults (and the more complex the faults) the longer it takes. We've also got a limited number of spares in stock to handle damaged equipment, but the frequency of the power outages is depleting these rapidly.

When networking fails in a building, it means that the phones (like computers) cannot reconnect to the PBX. If this happens, the easiest solution is to wait until networking is fully restored and then unplug your phone and plug it back in. You may also need to log out of the phone and log back in. (The same applies to your computer.)

For information on power redundancy on the telephone system, please see http://www.ru.ac.za/phoneproject/howitllaf...ou/poweroutages

Our empirical experience is that the shorter power outages cause the most damage. Thus an outage of a few seconds or minutes is likely to have more impact on our network than an outage of several hours. This is unfortunate, because these tend to have the least impact on other areas of campus.

Note that when attending to faults we tend to prioritise academic departments over student residences (such an approach is documented in and supported by the Student Networking terms and conditions). We also tend to prioritise areas with big lecture venues or computer laboratories over the rest of campus, and larger buildings over smaller buildings. However, to maximise staff efficiency, we usually attend to all the faults we can in a single area before moving on.

Our intention in doing this is to try and get as many people reconnected as quickly as possible. It unfortunately means that people on the periphery of campus might have to wait some time before a technician is available to attend to their fault.