Solar Powered Network

The key features of the solar powered network are redundancy and efficiency. The entire configuration has been designed to offer multiple layers of redundancy at each stage. From the three Internet backbones, to two dual port Ethernet cards in each server; you could almost say redundancy is our mantra! By using the latest virtualization technologies to reduce the number of physical servers we need and so reduce the requirements for energy and cooling. All this adds up to mean that we offer a 99.9+% Uptime guarantee.

Redundancy

There are 3 separate Internet backbones with different carriers, one wired and two wireless, running Border Gateway Protocol, which routes all traffic coming in and out of the network on the shortest possible path. This also means that if one or two Internet backbone links goes down, the other will still handle the traffic. The Internet connections then connect to two redundant Cisco 7200 VXR series routers, allowing one to take over from the other during a failure.

The Cisco routers connect to seperate trunked switches, allowing either to take over should one go down. Next, two separate Cisco ASA 5500 series firewalls block all but needed ports and monitor each other. If one firewall falls over, the other takes over automatically. From the firewalls, traffic goes to another pair of trunked switches.

Each of our servers connects to these switches through two dual port NIC cards, one dual port NIC card in each PCI-133 slot. The ports in each NIC card being teamed for failover with corresponding ports in the other NIC card. Then the opposite ethernet port on each card is connected to dual Cisco Catalyst 4500 series switches. Each of these two port separate NIC card teams are then teamed together for complete redundancy.

24 hr Network Monitoring

To ensure that this all runs smoothly, we have redundant monitoring servers that monitor each piece of equipment on the network. The monitoring is so advanced it will notify the network operating centre if there are any issues with the switches, routers, servers, bandwidth, etc. Potential problems are then pro actively dealt with ensuring our clients the greatest possible uptime available.

Redundant, Efficient Servers

Lightbeing Creations' commitment to the environment doesn't end at our servers. To reduce our cooling and electrical needs, we utilise a solution from VMWare, NetApp and IBM, which allows us to run large numbers of virtual servers on a relatively small number of physical servers. Most data centres experience "server sprawl" because every time there's a new application to run, it's given its own server. Which means in the average data centre, the servers are running at 4 percent to 8 percent capacity. With our solution, instead of assigning each server an application or a task, virtualisation lets us put individual physical servers into their own "virtual machines," with a 30:1 ratio of virtual servers to physical servers. This increases the physical servers efficiency and dramatically shrinking our server footprint. The goal is 70 to 75 percent utilization rather than 4 to 8 percent, and through this virtualisation process we achieve it.

An addition to reducing the need to buy so many servers, there's a very big reduction in energy and cooling. Each old server consumed 400 watts of power, whereas our IBM servers consume 625 watts, but by replacing 30 of the old servers with one of these new ones, there's a huge overall saving in energy requirements. Plus fewer servers generate less heat and so require less cooling, further reducing our energy requirements.

We use high specification, multi-processor AMD Opteron powered IBM servers which use sixty percent less energy and generate fifty percent less heat. These are clustered together, each with redundant connectivity to our SAN (Storage Area Network). This high-speed sub-network of shared storage devices that contain nothing but RAID hard disks for storing data, which makes all storage devices available to all of the clustered servers. As the data does not reside directly on any of the clustered servers, any server can go down and the other servers in the cluster will take over and balance the load.

• Redundant Bandwidth
• Redundant Cisco 7200 VXR series routers
• Redundant Cisco ASA 5500 series firewalls
• Redundant Cisco Catalyst 4500 series Gigabit switches
• Redundant Packeteer packet shapers
• Redundant Network Cards in all servers
• Redundant Power Supplies in all servers
• Redundant Clustered NetApp SAN (storage area network)
• Redundant Backup Servers
• Redundant APC Network Line Conditioners