The third annual Big Communications Event in Austin this past month was rife with talk of orchestration, the digital telco and SDN. Terminology, especially at events, is always dizzying, mainly because terms ‘stick' but their meanings are often very particular to the speaker and/or listener. I had one lengthy debate involving the meaning of SDN and whether it natively referred to the aspect of orchestrating connectivity between two or more points (which I prefer to call service orchestration) or if it was the traditional separation of control and data planes.
Either way, performance stood out in almost every keynote and dominated a significant portion of the Service Lifecycle Orchestration panel on which I sat next to many of my MEF associates, Ralph Santitoro from Fujitsu and Axel Clauberg from Deutsche-Telecom. Why? Well for one, it's the main differentiator (in terms of guaranteed performance) between the Internet and business services, be it the New IP or MEF's Third Network vision. Therefore, I made a decision to discuss the proverbial elephant in the room on the panel; the much-feared Operations Support System (OSS).
OSS and all it entails is such a strong aspect of the telco's DNA that its ability to address the digital transformation of communications service providers, a la SDN, NFV and even LSO, is of key concern. This is because of how challenging (both in terms of cost and time) OSS can be to change and how critical it is to the current revenue generation of most CSPs. Hence, I introduced a variation on a famous thought experiment which I termed “Schrodinger's OSS”. (For the original reference to Schrodinger's cat I encourage you to check out this quick video.)
Take this much-feared OSS and place it in a box. Just like Schrodinger's cat, as long as one does not open the box, there is an equal chance that the OSS is dead or alive. In fact, the OSS is both dead and alive at the same time until such time as one opens the box to discover the result.
In the context of SDN and NFV, the OSS is often perceived as “dead,” forcing the CSP to re-invest and build a whole new next-‘next-gen' OSS under the moniker of “service orchestration.” This is when the savvy CFO notices the latest purchase order line items have a striking similarity to the last set of ‘next-gen-OSS' purchase orders performed about 5-10 years ago. At the same time, the OSS is also alive in the sense that all of the current revenue (often in the billions of dollars/euros) flowing through a CSP's environment is dependant on that OSS, meaning that any changes to this delicate relic, even if for the greater good of the digital telco, may in fact harm the existing business.
That's why one can find an OSS box in every reference architecture diagram and why its usually off to the side because the CSP knows they can't get rid of it but doesn't know how it fits within all of what's new. It's for this reason InfoVista began collaborating with industry players to develop solutions for transitioning the industry from the old workflow to the new. Few CSPs have the luxury of Greenfield networks, so its important to find a way to adapt, whilst managing a hybrid network of both physical and virtual components, and semi-programmatic provisioning to full SDN-like orchestration.
Our standards-aligned multi-vendor proof-of-concept with Oracle Communications and Juniper Networks, recently featured at TM Forum Live in Nice earlier in May, provides some of these transitional strategies, including the on-going assurance of network performance, both physical and virtual, the increasingly dynamic nature of service instantiation and the mapping of customers and services to the underlying resources and relevant service and resource KPIs. Either way, its time to open that box, and make sure everything performs as it should!