This week, I saw a Heavy Reading white paper, in partnership with the New IP Agency (NIA) entitled “CSP Attitudes Toward Digital Transformation: A Reality Check”. In a short five pages, it’s a blistering indictment of the media and vendor marketing-led hype around next-generation CSP operations. And I agree with it wholeheartedly.
The paper summarizes the challenges and opportunities in helping the industry move forward to embrace the industry buzzwords – automation, network virtualization, big data, devops, etc. – through a complementary set of technologies that are created in a community-development model (read: Open Source) or vendor-development model.
I wanted to take a few minutes to share with you my views on the message, and how I internalize this for the development of the Dell EMC strategy for CSPs.
A Dell EMC Perspective
First, my views – the paper is exactly on-point from both my own conversations with CSP CTIOs and my own observations into what is perceived as a slow industry transformation. Noting that such massive transformations do not happen overnight, I think that CSPs are largely where they should be – recognizing the inevitability of the transformation (45.6%), but skeptical of the conflicting messaging (40.9%), concerned on locking into a specific direction so early in the transformation (43.6%) and rationalizing their own implementation plans.
There has been an absolute deluge of vendor perspectives, standards, open source projects, press coverage, and analyses on transforming network operations within CSPs since 2012 when the first NFV white paper was written. While the ambitions and goals are likely pure – to help the industry move forward – in the end, the outcomes have been distracting, keeping the industry in a constant mode of evaluation and validation, and leading to growing confusion on where to start, and start again, and again.
CSPs see vendors moving “up-stack” (enterprise infrastructure vendors delivering telco cloud solutions), “down-stack” (OSS vendors delivering network infrastructure solutions), and “cross-stack” (Systems Integrators moving from complex delivery to complete solutions) – all the while an ever-increasing landscape of startups are being funded in an attempt to disrupt the Network Equipment Providers (NEPs). Never mind the fact that these “stacks” that these vendors are moving within are inconsistent to begin with!
More importantly – I don’t see that the barriers to entry in the CSP space have changed drastically due to the incorporation of increasingly enterprise-originated cloud technology, based on the survey results (61% think that enterprise vendors are not yet equipped to build telco-class cloud networks). At least, not yet.
So – the questions that arise are really threefold:
(An aside: I do think that the paper inaccurately discounts the importance of Open Source in the survey to garner a specific perspective that over half the respondents view open source as “ok, or has limited application.” More importantly, my read on this is that an overwhelming percentage (86.6%) see a role for Open Source – some are crisper on what that role is (44.3%) and some are still developing their own opinions on what that role is (42.3%). Less than 15% believe that Open Source has limited or no role in telecom. Maybe I am just more optimistic than Heavy Reading – or maybe there is more raw data that I don’t have exposure to.)
With a logical concern that I am just professing vendor-marketing hyperbole (Note: I’m not), I’d like to share my view into how I internalize the questions above to help in the evolution of the Dell EMC CSP strategy.
The Industry Call-to-Action: Build Once, Reuse Many
The industry goal should be to build a common platform – regardless of current and/or future procurement processes, operational models, software stack selection, or technology direction – with the end goal to make infrastructure easily consumable and services easily deployable and manageable.
This common underlying infrastructure platform (dare I say, “IaaS for network virtualization”?) that spans currently-accepted principles (hypervisors, network overlays, service chaining, orchestration, etc.) but incorporates a path towards currently-incubated principles (containers, micro-services, automation and DevOps) without enforcing either a specific direction or espousing a particular end-state seems to be just what the industry needs. This allows for what I expect to be near-term divergence, and eventual convergence, in industry direction to happen.
The obvious need is for an infrastructure vendor who has assembled and built a set of expertise and knowledge around the vendor software stacks – both NFV and SDN – as a means of rationalizing an infrastructure deployment framework that is:
This is our goal.
The Dell EMC Response
Building a single, standard cloud infrastructure platform that has been certified by the predominant virtualization vendors, validated by the major Network Equipment Providers (NEPs), integrated with leading startups, and delivered through current trusted telco partners – namely, Systems Integrators, OSS vendors, and the NEPs themselves – for the deployment of telco cloud is a significant challenge – but it is one that Dell EMC has taken on. We’ve done it, and we have data points confirming that such an approach has merit. Now, we are focused on replicability, scalability (scaled out for higher capacity, scaled in for the network edge), and iteration.
Our ability to abstract the industry confusion and create a cloud infrastructure platform that allows telecom service providers to adopt new technologies at a pace that matches their organizational readiness without forcing a specific path at initial purchase has become a key differentiator for Dell EMC. To be forthcoming, it has been a challenge to break through all the other vendor hype out there by embracing such a practical, achievable objective.
|Update your feed preferences|