Our Reports


The Network’s Cloud-like Future: SD-WAN to Networks as a Service

Author: John Burke, CIO & Principal Research Analyst

IT and the enterprise need the network to be delivered like a service and managed accordingly—to create a Network as a Service (NaaS) model. This means having some kind of clear Service Level Agreement (SLA) to define what services will be delivered and how delivery will be measured. Yet, many organizations have no useful SLA on the network, especially the wide-area network, nor the ability to define and manage to one. The biggest obstacle to meaningful SLAs and to a NaaS model broadly is the way most legacy WAN management focuses on the wrong things. Where they exist, network SLAs are usually about device uptime and link performance. It’s no surprise they focus on the low-level details in this way: network engineers tend to work manually, site by site and component by component, rather than holistically.

SD-WAN addresses these shortcomings. In addition to sharply reducing the number of outages in WAN sites, and reducing the cost of expanded capacity, they provide highly automated, centralized management for both links and devices, and simplify creation and tracking of real, business-meaningful SLAs. That’s a huge step towards delivering NaaS, but it is only part of the picture: “XaaS” implies getting capacity on demand, and abstraction of supplier relationship details. SD-WAN doesn’t solve these problems. In fact, SD-WAN can increase the burden of relationship management! With SD-WAN, IT has means and incentive to use more and more last- mile providers, forcing the organization to monitor and manage that many more contracts and bills, and to develop and maintain a working support relationship with each. And, to manage elasticity on its own, the enterprise must negotiate short- term/open-ended contracts, get the links provisioned quickly, and integrate links into the WAN transparently. SD-WAN technology only helps with that last part.

For NaaS to be a reality, then, someone else has to handle the details of those provider-facing relationships, both ongoing and transient, with SD-WAN as one part of the delivery infrastructure. IT leaders, WAN architects and engineers should:

  • Evaluate SD-WAN technologies immediately, looking for:
    • cost savings vs continuing on as they are
    • how much downtime they might avoid using it
    • how much IT staff time could be freed up by using it
    • how much more quickly they could add new services to the network
  • Calculate the cost of managing a WAN vendor relationship
  • Consider the desired state for the WAN in terms of path- and provider-diversity and site resilience, how many vendors will be required to achieve it, and the associated costs
  • Evaluate using managed SD-WAN as a stepping stone to NaaS in the future
  • Begin planning for a truly dynamic NaaS-based future as service provider offerings mature, and 5G rolls out

Table of Contents
  • COMPASS DIRECTION POINTS
  • TABLE OF FIGURES
  • EXECUTIVE SUMMARY
  • THE ISSUE: NETWORKS NOT ACTING LIKE CLOUD SERVICES…YET
  • OBSTACLES TO NAAS: LEGACY WAN MANAGEMENT FOCUSES ON COMPONENTS AND CONNECTIONS
    • MEANINGFUL SLAS
    • EVERY ROUTER A ROSE—NURTURED AS A UNIQUE INDIVIDUAL
    • EVERY CONNECTION UNIQUE
  • SD-WAN SAYS: MANAGE WAN SERVICES TO MEET APPLICATION NEEDS
    • POOLING CONNECTIVITY, BALANCING CONSTANTLY
    • PLOUGHING UNDER ROSES, PLANTING CORN: MANAGING THE WHOLE WAN
    • CENTRALIZED POLICY FOCUSES ON APPLICATION NEEDS, BUSINESS VALUE
  • FROM SD-WAN TO NETWORK AS A SERVICE
    • THE EXPLODING PROVIDER POOL EXPLODES RELATIONSHIP MANAGEMENT
    • ELASTIC WAN
    • NAAS PROVIDERS COMPLETE THE TRANSFORMATION
  • CONCLUSION AND RECOMMENDATIONS


You are currently viewing a preview of this content. Nemertes Clients, please log in for full access to all research content. If you are not a client, please click below to purchase access to this research report. We also invite you to become a client.

 



Purchase Access

$1,995.00