For my last two blogs, I've expanded upon Nick Lippis' commentary on single network vendor strategies. Nick's stance, which I agree with, is that having dual network hardware vendors actually decreases network availability and limits services the network can provide. In the second blog, I explained how this concept should apply inside technology tracks, but different vendors for each track is appropriate.
So, how does the single network vendor concept extend to WAN service providers? The typical "dual-carrier" strategy required two carriers' circuits at every site. This strategy had its roots in the 1990s when carriers were building the first global WANs for companies, but, at the same time, experiencing global problems. Smaller, local problems also occurred. Building completely diverse networks with a single carrier was cost prohibitive, if not impossible. Carriers just didn't have the infrastructure to provide diverse services, so you used two. Then, if one carrier had a bug, outage, or problem that took down most of its network, your sites were still available. The service might have been degraded on one carrier, but WAN services were more about connectivity than performance. Most sites had a 64kbps or 128kbps circuit with a smaller frame-relay PVC. QoS was not possible either. Internet access was nice, but not mission critical. As long as the site had connectivity it could do work.
Plus, costs were high for these puny circuits, so having two carriers to compete against each other added another nice benefit.
Some of these factors still apply today, but other factors may be more important now. Carriers today have large global networks and can generally provide complete diversity, particularly in North America. Remote areas of the world still have issues, but that is part of the business risk of putting your main manufacturing site in Outer Mongolia. Furthermore, carriers today provide enhanced services - QoS, routing, dynamic capacity, etc. Using two carriers may relegate you to the lowest common feature denominator between both carriers; something I discussed about QoS before. Only by using a single carrier could you reap the benefits of these enhanced services.
The biggest risk in having a single WAN carrier is assumed to be a lack of diversity. However, most outages occur on the local access circuits (particularly when CO technicians start to work, but I digress). No matter the WAN carrier chosen, circuits are at the mercy of the local LEC (US) or PTT (International). Furthermore, the WAN carrier, to keep costs low, is going to purchase access circuits from the LEC at wholesale rates. This low cost alternative almost guarantees both circuits will ride the same cable into the site. So, while your complaint will be with the WAN carrier, the problem is really the LEC. Having two WAN carriers would not fix this problem. When the second carrier goes to buy their access circuits from the LEC at wholesale, guess which cable that circuit will be on?
Thus, having two carriers for "diversity" doesn't buy you anything except fewer services. Instead, you can force the single carrier to provide your sites complete diversity as part of the contract. Then the onus is on them to make sure every site is built diverse including the LEC circuits. With two carriers the responsibility is your own to validate circuit designs between two carriers and a LEC.
I used to be a believe in dual carrier approaches, but I've changed over the last couple years. Global outages are once-a-decade events, if ever. Carriers have expansive global networks to provide enhanced services. It's simpler to build a global enterprise network with a single carrier. The biggest risk is not technical, but service delivery. By having a single carrier you are reliant on that carrier delivering your new circuits to a site on time. No circuits, no site (been through that one before). However, that problem can be mitigated by aggressive SLAs.
Take advantage of the features and simplicity a single WAN provider can offer. Re-bid your network every two to three years via RFP to keep costs down, but then pick one carrier. If you have a large network, think about a single carrier per region or per business unit with inter-connectivity at hub sites or via a VNO. You'll have a better network.
Michael Morris is a communications engineering manager at a $3-billion high-tech company. His background is in enterprise WANs working with telcos and developing large-scale routing designs. He has worked on networks at government and corporate organizations, including networks at two Fortune 10 companies. In his current role, he leads a team of 10 engineers responsible for large-scale IT networking projects and architectural standards for data networks, storage area networks, IP telephony, contact centers, and security. Michael is CCIE #11733 and recently became one of the first three Cisco Certified Design Experts (CCDE) ever (#20080002). He has 11 years experience in networking and communications, including four years as a paratrooper in the U.S. Army. He has a bachelor's degree in MIS from the University at Buffalo and is working on his MBA from NC State University. In 2008, he was awarded the Network Professional Association (NPA) Professional Excellence and Innovation Award for his work on network architecture, templates and enterprise MPLS design.
The Leader in Network Knowledge
