How to decide which providers/vendors to receive your RFP?
Perhaps an obvious component but remains challenging. There are numerous resources we know of; including data on major service providers, resellers and VNO’s (Virtual Network Operators).
The task of provider research is further compounded due to the numerous available service options, i.e. SD-WAN services, MPLS, VPLS and Internet IP VPN.
Network Union have researched the market, details of which are available via our provider PDF which is free to all blog readers. Whether you use our data or move forward with providers you already have in mind, what follows is some of the core components to consider.
Calculating the true network reach of providers via their PE deployments?
With service provider marketing suggesting capability such as 99% coverage of the UK, understanding network reach is important.
These kind of statements, professing to offer such a high coverage, are typically talking about wholesale agreements with the likes of BT Open Reach, COLT, Virgin Wholesale and so on.
As the graphic demonstrates, a small network operator is able to market their capability offering complete coverage of the UK with only two Provider Edge nodes.
Note: PE nodes are a providers entry point into their core network.
The greater percentage of PE coverage, the greater their capability to terminate your office connectivity on a more local basis. Why does this matter?
In the main, latency, cost and diversity are impacted by a greater distance between your site and the prospective providers PE node. If your business is global, the potential issue is compounded due to the need for greater tail distances as offices could be in far flung locations.
If we specifically cover latency, the Global Enterprise is typically more at risk because of the potential for longer tail circuits vs. a UK organisation. In short, costs are also higher when PE coverage is lower due to the longer distances between office and PE node.
In order to gain clarity, we suggest asking for specific detail on true PE coverage within your RFP.
Selecting the right products, MPLS vs. VPLS vs. SD-WAN vs. Hybrid.
In the early 2000’s, the typical WAN RFP / ITT revolved around Internet IP VPN - the technology began to replace legacy networks such as Frame Relay and ATM services.
As the decade progressed, we witnessed the take-up of MPLS and VPLS services due to their ability to support QoS (Quality of Service) in conjunction with privacy. However, today, we are witnessing an Internet IP VPN resurgence due to the interest in SD-WAN technology.
The capability of devices such as Cisco iWAN with security, deep packet firewall inspection, and application prioritisation available within a single box is producing an enticing product offering. In addition, the Internet today is a well scaled, bandwidth-rich platform when compared to a decade ago which translates into a much more viable platform.
The average user is connected on a 24/7 basis to cloud resources via both corporate and BYOD (Bring your own device) hardware. The connection could originate from the office, home, hotel or anywhere with 3G or 4G wireless connections. In the majority of cases, the connection is provided by the Internet to both personal and work cloud resources.
With this in mind, it makes sense that the popularity of SD-WAN services via iWAN is increasing. In the majority of cases, the Internet is complimented by main office MPLS or VPLS connectivity where predictability of traffic performance is required.
- SD-WAN is typically associated with Internet-based connectivity - open sourced software is supported in conjunction with vendor features. We’re not quite there yet.
- MPLS VPN is a layer 3 any to any routed private WAN with end to end Quality of Service.
- VPLS is a layer 2 any to any LAN extension service with Quality of Service. Note: VPLS does ultimately have scaling issues with any to any topology.
- IPSec VPN is a security and encryption protocol used to secure packets as they traverse the Internet or other untrusted networks.
Project management and migration.
One of the main reasons for maintaining the status quo and not changing provider is the risk associated with migration. As readers will already know, migrating and managing the process of taking infrastructure to a new WAN provider is problematic for a variety of reasons. There is often an unwillingness to assist from the losing provider, then there are the resources required from your business to attend calls and work week in and week out on project delivery.
Presales work and technical design authority (TDA).
Network Resiliency and Diversity.
- Diverse circuits delivered with no single point of failure where possible.
- Carrier diversity - not a guaranteed method of achieving diversity but may work in some use cases.
- EFM (Ethernet First Mile)
- 3G / 4G
Further solutions include ADSL and EFM backup. We urge caution with regards to bandwidth vs the primary circuit as congestion could potentially occur.