While I was down at one of the other offices today I decided to resolve an irritating problem, on one of the domains on a production server I had tried on more than one occasion to change the network speed to gigabit without success. At each attempt both consoles would lock, setting the port on the switch back to 100 Mhz resolved the problem. So after I had been on site for a couple of hours of I went in search of an answer, a not so short search revealed the problem.

The cable from the port on one of the core switches went to a hub, plugged into this hub were several cables – these included the two System Consoles, one of the domain network interfaces and at least one other server. I was somewhat bemused by this configuration, it wasn’t one that I’d seen before. Once the problem had been identified it was quicly resolved, the assignment of a new port for the domain NIC and off we went.

The other bits I just left behind, in truth the pain of explaining that the hub was a single point of failure was more than I could contemplate. The original request for the networks team to investigate this was placed last Thursday, it was requested as a matter of urgency – if I hadn’t done it myself today I’d still be waiting. My thoughts as I reconfigured the interface were just how hard can it be to move a cable!