Over the last number of year’s enterprise IT departments have become more and more comfortable moving some of their more important workloads into a cloud environment. The benefits of this are widely known and range from reduced infrastructure maintenance costs to increased resource utilisation allowing the enterprise to only pay for the resources used when they are needed and scale down when not.
I have been considering the impact of this business model on cloud based connectivity, There are numerous companies providing private Wide Area Connectivity to the enterprise, based either on RFC2547 IP-VPN, Traditional point to point circuits or NGN based VPLS and layer2 service. Most deployments I have come across the customer pays a fixed amount for a connection, of a specific size into each of the branch offices it decides to connect into the Wide Area Network.
I think the market is getting closer to a utility based pricing model, where enterprise IT departments will have the ability to control the bandwidth and levels of QOS available to each of their branch offices as their business needs dictates.
The best way to explain anything is through a use case; take for example a chain of cinemas. This industry has recently migrated away from shipping large physical film reels around the country often days or weeks before film launch to a digital distribution model where movie content is distributed either encrypted Blu-ray disks or via a MPLS WANS where it is cached on local digital players for streaming to the high definition digital projectors. In the MPLS distribution model the amount of time it takes to stream the content to the premises is not that important as it is cached locally and can be done overnight, so this traffic would traverse the enterprise as a best effort traffic class with error checking done to account for any lost packets in transit.
Lately I have seen cinema’s start showing live content, Ballets, concerts etc etc. for this distribution best effort traffic class would be unacceptable as any lost packets would be blatantly obvious to the viewing audience as onscreen artefacts or audio defects.
In a utility based modem the enterprise It department would be able to provision an Expidited Forwarding (EF) queue to that branch for the period of time the live shows are running to ensire all packets in the video and audio stream get delivered in real time to the theatre without having to pay for that high level of service all year round.
So what would need to happen for this “Vision” to come true. The technology side of things is easy enough, telcos and service providers need to collaborate on creating an open API to their orchestration engines to allow enterprises order bandwidth on demand. The more challenging side of the house is contracts, billing mechanisms, network capacity management. All of which are items that can be overcome.
The main question I would leave you with, “Is the market ready ?”
Feel free to drop me a line with your thoughts to firstname.lastname@example.org