High-bandwidth resilient Internet access in data centres
IP transit provides direct access to the Internet, allowing data exchange globally via ultra high-speed connections with low latencies and close to the Internet backbone for most direct paths. We take pride in optimum and efficient global traffic routing.
Our IP Transit service is available in configurations tailored to the differing needs of enterprises, content providers, hosts and other ISPs:
For simple setups, providing a single or dual-redundant IP gateway with included IP addressing, usually connected directly to a firewall or switch
Offers resilient uplinks at the IP (layer 3) level for customers with BGP capable routers or firewalls, whilst including IP addressing
For networks with their own ASN and IP allocations, allows for advertising those IP prefixes globally and receiving full-tables or default-only, enabling dynamic routing and path selection across multiple ISPs
| Static | Private BGP |
Public BGP |
|
|---|---|---|---|
| IP addresses provided |  |
|
 |
| Your own IPs (/24 or larger) | |
|
|
| BGP equipment required | |
|
|
| Redundant uplink options | |
|
|
| Default route only | |
|
|
| Full IP tables option | |
|
|
| Multi-homing with other ISPs | |
|
|
| Your own AS number required | |
|
|
Latency in data centers is crucial for the speed and reliability of data exchanges, representing the delay between a request and its response. It is influenced by a number of factors, such as congestion and transmission medium (ie. fibre, copper or wireless) but a primary factor is distance (and therefore choice of route).
For example, connecting London to New York (~5,500 km apart) involves a fixed route via an undersea cable across the Atlantic. Light travels through fiber at two-thirds the speed of light, adding ~1 ms of latency per 200 km. This gives a theoretical minimum latency of 55 ms for a round trip, though real-world delays range from 68-75 ms due to processing and other factors.
In contrast, a connection from London to Singapore may involve several potential routes. Here are three real-world routes from some of the worlds top carriers, highlighting the differences in achievable speeds.
An ISP's choice of international carriers affects latency, as limited carriers may only provide slower routes. More carriers mean better options, as some perform better to specific locations than others.
Here are real examples of average ping times taken over different carriers' paths between London and major foreign cities:
| Carrier A | Carrier B | Carrier C | |
|---|---|---|---|
| Barcelona | 25ms | 29ms | 34ms |
| Miami | 113ms | 104ms | 99ms |
| Stockholm | 30ms | 25ms | 28ms |
| Tokyo | 239ms | 250ms | 219ms |
We use four of the top five carriers in the world today (ranked by AS cone size) as well as London peering for efficient routing within the UK:
This gives us very flexible routing options. Whilst many ISPs choose the cheapest paths, we prioritise performance-based routing for faster and more reliable paths, minimising latency and avoiding congestion.
We offer three models for accounting IP transit utilisation:
We match the paid bandwidth allowance to the port speed. For example, you order a 10GE unmetered port and it comes with 10Gbps of bandwidth allowance. This has the advantage of a totally fixed cost which is also slightly cheaper than a burstable port.
We throttle the bandwidth of the port to a bandwidth of your choosing. For example, you order a 1GE port and a 200Mbps bandwidth, and we artifically restrict the throughput to 200Mbps upload and download. You can therefore commit to any bandwidth level, which is less expensive than a flat rate port, whilst never exceeding your paid allowance. However, it does not allow for any burst headroom and this method can suffer from microburst packet losses.
This is the most common model. You pay for a chosen contractual bandwidth (the committed data rate or CDR) and we provide the next port speed up from your commitment. For example, you commit to 200Mbps and we provide a 1GE port. This provides you with burst headroom up to the full port speed for traffic spikes, but usage above your chosen CDR may become chargeable as overages. We calculate your utilisation on the port by applying the 95th percentile model.
The chart above illustrates a typical daily distribution of traffic, with regular peaks during daytime hours and an unexpected traffic spike on the Wednesday, reaching 2Gbps. The Committed Data Rate (CDR) is the contracted bandwidth being paid for.
To ensure quality of network for everyone, it is vital to plan (and cost) for capacity requirements suggested by the CDR. However, having some headroom for occasional bursts of traffic avoids congestion.
To facilitate a fairer usage of this burst capacity, we provide the top 5% of your usage above the CDR for free. To achieve this, we sort your bandwidth utilisation throughout the month from highest to lowest rates, then ignore the top 5%. The next highest value (the 95th percentile) becomes the recorded monthly data rate.
An overage charge is only applied if your 95th percentile rate exceeds the CDR (0.15Gbps or 150Mbps in the example above). This fairer approach permits you to have occasional traffic spikes without paying for extra capacity that you might only need for a total of up to 36 hours per month.
Our strategically located data centres and BT fibre exchanges allow us to efficiently connect content providers, ISPs and businesses across London:
London Docklands
London Central
Inner London
Outer London
Slough
* Includes buildings with connection price parity to direct points of presence
† Locations accessible at (possibly higher) campus connection pricing
This service can be Managed By Enlink which brings the following benefits:
Planning & design
We carefully assess your technical and business requirements ensuring a service tailored to your needs, balancing performance, scalability, and cost-effectiveness
Procurement & installation
We take care of everything, from sourcing equipment to managing installations, delivering a stress-free deployment
Ongoing management & support
Enjoy peace of mind with our continuous management, proactive diagnostics, upgrades and patches, and 24x7 fault support
Time and resource savings
We take pride in being conscientious, saving you valuable time and avoiding the headaches and costs of downtime
Use our simple wizard to ensure a quick turnaround of your quote, typically within one business day. If you need additional help or guidance, request a callback.
We may be able to offer you enhanced discounts if multiple services interest you:
