Friday, 24 October 2008

Just pull this cable out and see what happens

Normally I'm not let loose in our data centres - I think it's because when faced with all those cables, I have a huge desire to pull some out! However, the other day I was invited into one of the machine rooms, and treated to a demonstration of some clever work our network team have been doing. We're gradually reconfiguring the network to increase resilience, doubling up network connections and replacing our user switches with new virtual routers. Eventually we'll have 2 Gb resilient connections to all buildings. Watched a nice demo of the way the two routers will work with one active and one inactive - if the active one fails, the inactive one immediately takes over. This was simulated by me pulling out the power cables to the active one (heavily supervised of course) and the other one kicking in straight away - barely a blip in the guitar solo being transmitted. Marvellous!

A lot of what our infrastructure team does is hidden to most of our users, with it only coming to their attention when something goes wrong. But it goes without saying that it's absolutely vital to the institution and much appreciated.

2 comments:

Anonymous said...

So that would explain the 15 e-mail alerts from nagios saying that services were down, causing momentry panic.............

Chris Sexton said...

oops, I thought it was only a test service