Don’t do SDN. Quite simply there is nothing to be done as such. If the current industry hype is compelling you to “do SDN” or “get SDN” you may find you already have it (or a version of it). If you are a user of server virtualisation solutions with hypervisors and virtual switches you are already leveraging networking elements defined and delivered by software (but elements the MUST still drive hardware). To extend the discussion further if your organisation uses carrier based services (delivered by one of the major telecoms companies) you are already using network services like MPLS and VPLS that massively leverage elements defined in software to deliver the networking outcome you need (many call this network function virtualisation but this is somewhat semantic).
Therefore are you missing anything now or are you already a customer of the next big thing but were blissfully unaware? Enter that horrible response “Yes and Yes”, modern enterprise customers have embraced software defined networking ideals for quite a while however equally the software defined storyboard has been somewhat invisible to all but those learned technologists employed to design, build and support the platforms in question.
But now those more recent networking elements defined in software and grouped together under the SDN banner, paint a totally different picture even if many of the legacy network infrastructure elements are retained. The brave new defined world of SDN is all about open standards (preventing vendor lock in), accelerated innovation (by using open source ideals), potential for cost reduction (due to the hardware abstraction or any network hardware vendor ideology), true network agility (massively reducing time to market of applications and new business services) and most compelling of all, application awareness (to ensure applications control the network not vice versa).
It means that striving to “do SDN” makes little sense unless you are clear on the business outcome aligned aspects that are essential to realise. With that in mind the “big tip” is to understand the SDN or network virtualisation elements that can deliver tangible value against a realistic operational plan. This must be the primary action for now, not an unchecked move to a new platform based on a features biased evaluation.
To that end now is the time to evaluate how ready your current networking platform (and security footprint) that underpins your business is to deliver the speed, agility and dynamism your business requires. And maybe is not a valid response. By understanding and leveraging the most viable elements of traditional networking approaches, interfaced with validated software defined and network virtualisation outcomes, the best of both worlds has the potential to deliver the best in the world outcome for your organisation. The new dawn of the software defined IT enterprise will potentially be your best dawn ever……
Big claims maybe – try me !
Until next time
All change” – the new story in town is a “software defined something” story with software defined networking (SDN) top of the hype wave for many. Is now really the time for SDN?
Is SDN an essential networking concept or technological advancement? In reality, “No” due to more than enough tacit and continually evolved operational and technical insight already available to deliver enterprise networks of today. However do the concepts presented by SDN have the potential to deliver real world networking operational and financial benefits by enabling previously unthinkable or challenging networking outcomes, “Yes”. At present the hype is clouding reality and sadly causing a fair disconnect in the minds of networking stakeholders. We observe a common cycle (and not unusual with new technology introduction) with numerous prospects for the SDN story deciding in advance based on the hype and momentum that its essential to them, followed by unsuccessful attempts to translate the new technical highlights into real world justifications for change.
The likelihood of success by adopting this “hype and hope” approach is rare. There are some organisations (few) who may be fortunate to find a short term SDN silver bullet that can be used to validate “why SDN” – but for the masses the path to success must start with a true understanding (really isolated) of the future business needs of the organisation with an appraisal of whether the existing networking footprint can enable it. The current functional networking environment for many is more than adequate and whether delivered by teams of highly qualified engineers or as an outsourced footprint – “works” (and in the main works well). But for many “works” in tomorrow’s business template requires a networking environment with way more facets including flexibility, optimum security, agility, “application intimacy”, ease of change and true integration with other infrastructure environments top of the success criteria. It points to the importance of justifying SDN adoption or evolution based on less “technical me too” advancements but more on the key activities (outcomes) the networking layer must underpin for today and tomorrow with an unemotive appraisal of the new challenges that will be faced.
Now we arrive at an area of controversy & confusion – cost reduction. Many of the earlier value statements surrounding SDN laboured or majored in cost reduction. Yet early indication suggests that is less likely in the short term (due to the need over time to depreciate and decommission key existing networking infrastructure investments). That’s not to say costs cannot be saved – by delivering newer greenfield type environments with SDN fundamentals at the core, a cost conscious footprint can be designed and deployed. However that is currently more challenging with more common networking infrastructure templates based on well funded, stable platforms with well defined operational processes and well formed financial controls. Sending good networking platforms out to pasture for “SDN promises” will prove a somewhat false economy.
SDN is here to stay and the benefits to networking deployment and service delivery will be in time unquestioned. But now is the time to understand how the SDN high notes mentioned by many can deliver real world benefits to real customers based on real use cases. Most compelling is the flexibility a software based networking environment will deliver to cloud or virtualised environments. The potential for “application intimacy” and real world reductions in “human latency” or configuration delays enabled by SDN makes selection of future network infrastructure elements with SDN compatibility a must. And lastly with multiple vendors purveying SDN strategies that maintain loose compatibility with each other but are often best delivered as a single vendor outcome, careful consideration is required before any definitive strategy/product/vendor selections are made.
SDN has made the networking arena interesting and exciting to a degree not seen for many years (possibly not since the change from hubs to switches). It will ensure existing market players innovate and accelerate to remain viable in the future and enable a window of opportunity for new market entrants to lead a charge based on a new world order. Networking will never be the same again.
Until next time