I am fortunate enough to have the opportunity to work for a large and successful real estate firm in the country’s most active real estate market. The Bay Area of California is home to Silicon Valley: where technological innovation is birthed, and globally recognized tech brands are headquartered. After having spent the better part of 1.5 years in this market, one thing in particular has surprised me: as is the case with several established medium-sized organizations across the country, the time tested reliability of MPLS still maintains its grip on professionals tasked with managing their organizations’ infrastructure.
Who among us hasn’t shifted most—if not all—of our applications offsite? When the inflection point of combined CAPEX on hardware and its associated cost of maintenance arrived at the cost-savings and reliability that could be achieved through an OPEX spend in the cloud, all of us made the gradually obvious decision to migrate our resources. Yes, there is a certain level of security that comes with managing corporate owned devices on a single network; yet the workload and maintenance involved with buildingIPSec tunnels for remote VPN access between sites, running consistent patches on local file servers used for DFS and shared folders (in a Windows environment), and ensuring that backup servers are consistently functioning in the event of local failures has been made so much easier by an idea that has been that has been plastered on billboards and digital skyscraper ads everywhere: move to the cloud.
"Security is a breeze with SD-WAN’s built in end-to-end encryption and unified security protocols"
So we did. We mitigated organizational security needs with Multi-Factor Authentication for cloud access and, in some BYOD environments, mobile device management. Yet as the demand for bandwidth, which is inherent to cloud access, grew, many organizations in the Valley have remained hesitant to move off of MPLS. Longstanding relationships with ISPs, legacy personnel, and the fear of unreliable connectivity all play a role in this stasis.
All of this is to say that, in the vast majority of use-cases I’ve seen here in Silicon Valley, fear of the unknown (and relatively untested, some would argue) is what has prevented many organizations from transitioning to an arguably superior alternative: SD-WAN.
With many businesses having shifted their assets and resources into the cloud, the need for Real-time Transfer Protocols (RTP) within those organizations has essentially been relegated to on premise VoIP instances and local file share. The “necessity”of MPLS in these cases can either be mitigated with redundancy via multiple transport routes in a single SD-WAN appliance or a lower bandwidth (cheaper) MPLS line prioritized exclusively for RTP layered on top of a (much cheaper) coax, business fiber, DSL, and/or LTE line.
Having taken into account the only argument for the need of an MPLS line, the benefits of SD-WAN begin to unfurl at an enthusiastic pace and every other reason in favor of MPLS becomes moot.
Security is a breeze with SD-WAN’s built in end-to-end encryption and unified security protocols. Line provisioning times go from weeks to, in many cases, seconds. Packet loss and jitter is corrected in real time via Policy-based Routing (PbR). Site outages, caused by weather or any variety of backhoe, become a thing of the past with redundant lines. Perhaps most importantly, DIA lines can sometimes achieve 4X the Mbps speeds of MPLS at a fraction of the cost, making the cost of transport less expensive at best, and cost comparable at worst.
It’s time we let go of the affectionate nostalgia we have for MPLS. It slows us down, it costs us more money, and it doesn’t meet our needs. There’s a new product in town, it goes by the name of SD-WAN, and it’s time we all gave it a call.