Network Upgrade

 

We have some very exciting news to share with you: We have recently signed contracts and placed orders for a substantial overhaul of the Simwood core network.

Whilst this does not directly affect our voice services, it does affect the network they run on but more importantly provides a solid platform for the future growth of both our business and hopefully yours.

The transition to IPv6 was an opportunity to re-examine our network from top to bottom and evaluate its role. Whilst our existing network was built from the ground up for voice and has served us well, these changes transform it into a strategic asset to cope with the demands we see coming and the opportunity they present.

Background

IPv6 aside, the Internet is changing at an alarming rate. According to Cisco, total traffic in 2014 will be 4 times that in 2009 but of more significance video in 2014, in its various forms, will represent 91% of traffic. That 91% is of course an average and with video streaming of major events, flash traffic will cause substantial peaks. It is clear therefore that in terms of Internet traffic alone, whilst 5 years ago video and other performance sensitive applications did not exist materially, they will represent the majority of traffic going forwards. Networks will need to cope with increased throughput of more quality sensitive traffic.

At the same time as the demands on the network per unit of traffic rise, the selling price continues to fall. There has to be an inflection point where users become used to buying different capacities of a given quality of service but in the mean-time on economic grounds alone networks need to be able to effectively prioritise traffic and to do so right the way to the customer premises.

Finally, Enterprises are increasingly converging services on to single connections, typically guaranteed capacity ethernet circuits. That exaggerates both of the above requirements for QoS and capacity across the network. We already see our Enterprise customers wishing to replicate SAN transactions between sites, or migrate running virtual machines between sites, both of which impose tremendous capacity requirements and QoS demands that supersede even voice and video. These demands are not just on the network itself but within the customer’s own circuit where they may also be running voice and other applications.

It is clear therefore that a network built even a few years ago is likely not suitable for the challenges ahead. It needs substantially more capacity than was imagined, needs to prioritise traffic efficiently across that capacity, and needs to be able to offer customer specific guaranteed performance across the network, with additional customer specific QoS within that. Moreover, it needs to provide today’s ethernet fabrics, using current standards which simply didn’t exist a few years ago.

What are we doing

In the first instance we are expanding on our current footprint of Telehouse East, London and Synergy House, Manchester. We have secured additional space within Telehouse East and begin installing new fibre etc. tomorrow. We have also secured some of the last space in Scolocate in Edinburgh. This gives us ample power and space across our three sites for core services. Whilst much has been pre-sold to existing customers, we do have some co-location space available if required.

We have also procured new fibre connectivity between London and Manchester to augment capacity there, as well as new fibre between London and Edinburgh, and Edinburgh and Manchester. The result is a high performance, scalable optical ring between the three sites.

Most significantly, we have placed orders with Brocade to equip each site with Foundry NetIron routers. These will enable us to offer a traffic engineered MPLS core, assuring performance across the network, with capacity in each site, and potential capacity between sites of tens of times our existing peak usage.

We have also ordered the Brocade network management system to enable us to effectively extend related services into and across all sites. More on those later.

Both LINX and AMS-IX use Brocade, as do 90% of IXPs apparently, and it seems 90% of the Global 1000 rely on Brocade. We think we’re in good company and very excited about the position this puts us in to support our customers in capitalising on the market changes we see coming.

How?

We’re building out the new network in parallel to the existing to minimise disruption. Carrier Ethernet customers will migrate immediately along with some virtualisation customers. Beyond that our voice services will migrate progressively as appropriate and we will notify changes in due course.

***

We’re tremendously excited about this and think you’ll love the results. As always, thank you for reading and please do not hesitate to contact us with any questions.