Walking through Wan history, from the early days…
Leased lines
• Customers subscribe to ‘dedicated’ point-to-point links • Cost prohibitive for customers
Frame Relay and ATM services
What is a Virtual Private Network? Branch office Corporate headquarters PSTN Infrastructure Point-to-point link (leased line)
Started in the 1980’s
FR/ATM Network • Customers subscribe to point-to-point links to construct Hub and Spoke or Mesh• Offered over a shared infrastructure that offers multiplexing advantages • Cheaper alternative to Leased line
Mobile Users and Telecommuters Branch Office Corporate Headquarters Suppliers, Partners and Customers Intranet Extranet Shared Infrastructure
Started in the 1990’s
Walking through Wan history, up to present day…
IP-VPN
• Customers subscribe to multipoint Layer 3 MPLS service • IP only
• CPE = router, routing interaction with Service Provider
Branch Office Corporate HQ Branch Office VRF VRF VRF IP IP VPN VPN L3 Router VPLS
• Customers subscribe to multipoint Layer 2 MPLS service:
VPLS VPLS VB VB VB L2 Switch Branch Office Corporate HQ Branch Office
• All sites appear to be connected to a single bridged domain • Multi-protocol
• CPE = switch or router, flexible, dynamic • Simple, familiar interface
• Clear demarcation
Advantages: Flexibility, control, quality of service, optimum performance, ultra-low latency, scalable & flexible bandwidth, network resilience, multiple service options
UK WAN services evolution
1500 2000 2500 E U R m il li o n Ethernet FR•
The Ethernet market is growing rapidly
•
Ethernet is replacing FR/ATM and IP/VPN
0 500 1000 1500 2008 2009 2010 2011 2012 2013 E U R m il li o n FR ATM Broadband SDH, WDM
Drivers of the UK WAN evolution
Forrester forecasts the Enterprise data traffic in Europe to grow by 30-70%
per annual to 2012 – typically doubling every 18 months.
Without innovation, costs to the enterprise will rise inexorably.
This is why Carrier Ethernet is on every agenda.
Carrier Ethernet provides optimum availability, flexibility, speed of response
and resilience without compromises, at an affordable price.
Forrester expects VPLS-based services delivered over MPLS networks to
become the dominant means of delivery for Ethernet services over the
next 24 months.
Drivers of the UK WAN evolution
Carrier Ethernet > mass adoption due to low cost, performance
Carrier Ethernet > mass adoption due to low cost, performance
and flexibility. Ethernet over MPLS = VPLS.
VPLS is becoming mainstream.
VPLS is a high security, low latency means to connect sites or services either point-to-point or as a mesh.
We use Virtual Private LAN Service technology to create a WAN with many of the features and functionalities of a LAN.
Our Layer 2 VPLS-enabled Next Generation Network (NGN) provides Ethernet based multipoint-to-multipoint communication over
Virtual Private LAN Service (VPLS)
We are all familiar with the concept of simply plugging our PCs, servers and printers into our office LANs and getting instant plug and play connectivity between all of our devices. Well VPLS now extends that functionality across the wide area, pulling all of your remote sites back together and allowing you to think of them as a single large campus serviced by one or more local area networks.
multipoint-to-multipoint communication over IP/MPLS networks. It makes it possible for
geographically dispersed sites to share an Ethernet broadcast domain by connecting multiple sites through pseudo wires to a central bridge on an any-to-any basis – think of it as a private virtual VPLS cloud.
Last Mile OA & M
The Metro Ethernet Forum describes Service Management as one of the key features that characterises a Carrier Class Ethernet network and up to now, the last mile has been something of a management blind spot. In order to redress that balance, we now deploy a fully managed Ethernet Demarcation Device or EDD on all optical access circuits.
• 802.3 AH Ethernet OAM tools for the first mile.
• Multiple services presented on separate ports.
We can now present different services on physically separate Ethernet ports whilst maintaining 100% separation of each layer 2 service. This offers the perfect insulation of secure LAN to LAN traffic from public Internet or for example a totally separate Voice VPN,
without as in other solutions having to use complex and slow Access Control Lists at layer 3.
• 100% separation of layer 2 services.
• Bandwidth management per service CIR, EIR and Burst size.
• Traffic classification based on TOS/DSCP/802.IP/802.1Q.
• VLAN forwarding/swapping/stacking (Q in Q).
• Performance management.
• Status monitoring.
Layer 3 v Layer 2 Comparison
Layer 3 2547bis Layer 2 VPLS
Protocol
Limited to IP protocol only so special handling needed to make some applications available on the network
Multi-protocol Ethernet – if the application can run on your LAN then it can run on the WAN
Architecture
Routed solution means customer routers have to navigate service provider routers to connect with each other
Bridged solution so customers routers only see routers of their own network – transparent WAN to customer
Control Customers lose control of IP routing Customer retains control of IP routing
Change: Adding Sites
Complex as all service provider routers connecting sites need routing changes
Simple, only the service provider router the site connects to needs changing
Change: IP changes
Service provider needs to agree all IP address changes
IP addressing is simpler and no need to involve service provider
Faults: Management
Very difficult to fix faults – costing time and money for service provider
Service provider does not need to deal with customer routing issues – fewer fault calls, quicker and cheaper fault fixing
Faults: Calls
No clear demarcation between customer and service provider network – resulting in customer fault calls when the issue is on the customer network
Clear demarcation of service provider and customer results in less stress for customer and less cost for service provider
The service creation platform
A single Ethernet tail circuit connects
customers from their premises into
our Next Generation Network, known
as our Service Creation Platform.
Connection into the service creation
platform gives our customers access
to a highly flexible suite of
to a highly flexible suite of
plug-and-play applications and services
available on a wires-only or fully
managed basis.
Case study: Clients current WAN
Internet
Hub Site
Centralised hub
Site services Partner Site
PSTN
• 1 hub site & 3 spoke sites.
• One spoke site is also used for Disaster Recovery and should DR ever be invoked, the staff and services from the affected site would be relocated there.
Spoke Site Spoke Site (DR)
• The hub serves the spoke sites with centrally held file, print and application services.
• Internet access was provided at the hub site using a 2Mb Internet leased line.
Each spoke site used a 2Mb point to point leased line to privately connect back to the Hub site accessing the central services.
A dedicated tie-line voice circuit was used to connect the PBXs at the hub and DR sites, with ISDN 30’s providing access to the Public Service Telephone Network to the PBXs at each site.
Case study: Clients current WAN
The problems:
The client’s bandwidth requirements had long outgrown the 2Mbps capacity of the existing leased lines and the network performance was suffering – not to mention the poor IT support team who had to field the complaints from users.
An upgrade using existing legacy technology meant either adding a second 2Mb leased line to each site (doubling costs) or upgrading to 34Mb circuits – either option being far in excess of the available
•Insufficient bandwidth capacity •Multiple circuits
•Network performance
•Inflexible capacity and functionality •Single points of failure throughout •Restricted growth of the company •Costs exceeding available budget
to 34Mb circuits – either option being far in excess of the available budget.
There was no way of replicating the hub site comms at the DR site to offer true Business Continuity without having to install a second infrastructure of leased lines.
The client’s services had outgrown their comms room and become more and more critical to the business, so a fully equipped hosting environment was also required. Building an on-site datacentre simply wasn’t practical with the available space not to mention the spend that the project would have meant.
Case study: The solution
Business Class Internet Voice Network iSCSI Voice VPLS Voice VPLS DSL Network Data VPLS Data VPLS Supplier/Exchange/ Partner/Client Site Hub site MPLS/VPLS Network Spoke Site (DR) Co-Location iSCSI Hub siteCase study: The solution
The higher capacity circuits used allow the clients business to operate more efficiently and support growth.
The secondary circuits offer maximum availability and a 5 x 9s SLA. Re-locating the hub site services to a central on-net facility negates both ‘bottlenecks’ at the hub and also wasted bandwidth through the ‘trombone’ effect that is inherent with hub and spoke designs. It also maximises business continuity as no site is dependant on any other for day to day operations.
Such is the design, that should the main site ever fail, every other Such is the design, that should the main site ever fail, every other site will continue to operate as usual, connecting to the centralised services with no interruption.
Extra bandwidth capacity was provisioned at the D.R. site to support a DR invocation and accommodate the extra staff that would be located there should the need ever arise.
The client can also rest happy in the knowledge that their services are housed in a custom built facility with full n+1 power resilience, physical security and environmental controls.
•Full resilience
•Higher capacity with instant upgrade path •Removes bottlenecks and ‘Trombone Effect’ •No inter-site operational dependence •Full D.R. and business continuity supported •High availability hosting environment
•Secure – client is in control of routing and data paths
•Supports client’s growth
•Reduced spend. Delivered at 90% of current spend, and providing a 500% increase in capacity •Full scalability of both comms and data storage capacity
Benefits of our solutions
You are in control – With control of your own IP routing, your IT department can be much more agile in responding to varying levels of customer demand.
Optimum IT efficiency – Visibility of your network information, control of your IP routing, and the centralisation of many functions, enables your IT department to maximise its resource and create greater efficiency.
Unrivalled security – Our data traffic travels across a VLAN which is the same as a private network so the risk of internet security breaches is minimised.
Low prices – Our network design can dramatically reduce circuit prices by up to 70% compared with Low prices – Our network design can dramatically reduce circuit prices by up to 70% compared with
traditional leased lines.
Optimum performance - The VPLS technology we deploy means we can offer huge increases in network speeds - up to 10Gbps.
Ultra-low latency – We provide industrial strength resilience with a carrier-class switched
infrastructure so we can offer very low latency (67 millisecond round trip from London to New York). Low network down time – We can boast 99.999% network uptime, which means as our customer, you