|
|
|
|||||||||||||||||||||||||||||||||||||||||
| Network |
The network on which our servers reside is fully switched, with multiple redundant routers, and multiple connections to different ISPs, allowing us to provide absolute availability of the network with no down time. The network is designed around a high-speed switched architecture. Each server is plugged into at least a 100mbit, fully-switched port on an edge switch. Each edge switch is then in turn plugged into two separate aggregation switches at 1gbit, with failover redundancy allowing either 1gbit link to fail. The core network is run exclusively on Cisco equipment. The network design is fully redundant and high performance, with low latency and high throughput. Once traffic has reached the core switches it is routed by multiple, redundant routers on to the internet over a number of different connections. The network transit connections are currently to Level 3, Tiscali, and Deutsche Telekom and is also peered with approximately 200 ISPs at the peering points LINX, XChangePoint, and LoNAP. We are connected to this range of providers via redundant routers. With this configuration if one of the routers fails, the secondary takes over the full routing within a matter of seconds – meaning that a router failure doesn’t compromise network availability. In total we have 16,000mbits of network connectivity available. |