Lumen’s aspiration for enterprise is a perfect combination of cloud, edge and network.
Lumen is pushing hard to effect a vision of a seamlessly interconnected world of industrial sites, clouds, edges and the facilities that make them work together. In many ways, cloud – in all its versions – is the killer app for fast, everywhere networks. Networks, on the flip side, are either the boon or bane of cloud. Good networks make clouds sing. Weak networks reveal cloud’s Achilles’ heel.
Lumen wants to be the global carrier to knit them together, into a seamless, latency-appropriate, whole.
This week, Lumen announced yet another brick in its slow-but-consistent rollout of its platform for “Industry 4.0”: Lumen Edge Bare Metal. This capability provides a higher level of performance, latency, security and most importantly predictability and joins their existing offerings including internal cloud capabilities, AWS, Azure and GPC – among with the Lumen optical network. Its plan differs in detail and positioning, but is not unlike Verizon’s VNS – blogged on here.
Lumen is also both clear and realistic: its unique differentiator is its network, combined with the ability to orchestrate all the pieces, both Lumen and third party, together into a logical, end-2-end whole. In the network domain, Lumen makes strong points, arguing that the global penetration of its optics – fully owned and controlled, not leased – is second-to-none. In support of this bit of chest thumping, they provided some eye-catching numbers, including:
- 450,000 route miles of global fiber…
- . . . that Lumen owns, engineers, monitors and maintains
- More connections to public and private datacenters globally than any other carrier
- 60 edge nodes underway, featuring bare metal and (to come) multi-tenant with QoS managed containers
- Critically, since it is so often a bottleneck, extensive peering relationships and POPs, with peering directly to > 6,400 ASs (autonomous routing systems)
That last datapoint is, to me, critical. Peering can often become a choke point for traffic, and a lack of peering to ASs also means that indirect routes (more hops, more latency) can be expected. So, peering depth and extent is a techie detail that is worth paying attention to if you want applications to perform well. Furthering this point is Lumen’s stated goal of being within 5ms of most industrial endpoints, providing excellent latency performance for all but the most time-critical transactions. OK, arbitrage trading need not apply, but for most everything else, looking good.
“Secret Sauce: Orchestration.” Come again?
As noted above, Lumen refers to orchestration as its “secret sauce”. Say again? OSS is important? Actually, yes. Lumen’s aspiration, if not always 100% reality yet, is to have API integrations such that it (or the end user) can orchestrate end users’ applications consistently across compute zones (private, public, edge), transport paths, transport overlays (e.g.: SDWAN) and vertical workloads (e.g.: SaaS). This has significant advantages in simplicity, consistent and E2E QoS and monitoring, and as we argue here, security. The simple fact is that the majority of security vulnerabilities stem from errors and omissions in configuration/settings, and automation, end-to-end, can vastly reduce these errors. So end-to-end orchestration, if it includes security handling, will become more and more crucial to improving security and eradicating the evils we deal with on a daily basis.
Lumen has a large array of partners and suppliers in this endeavor, including IBM, VMware, Itential, ServiceNow, Nokia Deepfield, AWS, Azure and more. We hope to delve more into the “how?” and what we can learn from the messy details of implementation soon. Reality is always harder than theory 🙂
One clear take-away is that the telecom industry is beginning to recognize that those derided dumb pipes need not be dumb, and either way are valuable and becoming more obviously valuable by the year. Lumen may be on a journey, but it’s a journey worth taking.