How will CSPs manage the avalanche of real-time data that will be created at the network edge?
To embed our video on your website copy and paste the code below:
<iframe src="https://www.youtube.com/embed/G3yvFe3omZA?modestbranding=1&rel=0" width="970" height="546" frameborder="0" scrolling="auto" allowfullscreen></iframe>
Constantine Polychronopoulos, VP & CTO, Telco Business Unit, VMware
Predictions are that at least 25 per cent of all data will soon be created in real-time at the network edge. So the big question is what role will CSPs have in supporting and managing that huge mass of edge-generated data? Well, as Constantine Polychronopoulos points out, service provider are already generating data from edge-based devices and the inescapable reality is that as time goes on more and more and more data will be generated at the edge, whether that data is real-time or not, and it will have to be managed or the entire edifice will crumble.
As an example Constantine Polychronopoulos says that in the near future, in excess of 1TB of data per day will be generated by autonomous vehicles alone - and that is only a tiny part of the information avalanche that will have to be managed. What's more, back-hauling data to the cloud is so expensive as not to be a viable option and decisions about real-time applications will have to be actuated at the edge, or they will simply overload and fail. And that's why wireless CSPs will have a vital role because use cases are being delivered by and on 4G, and, soon, 5G devices and systems.
Even now though there is still debate about where "the edge" actually is and what it actually means because there are several incarnations of it - that the "near edge", the "far edge" and "the core". Constantine Polychronopoulos says VMware believes in enabling a hybrid multi-cloud environment with seamless provisioning capabilities to permit apps to run in highly-distributed clouds so that it will not matter where the edge may be at any one time. What does matter is the technology that makes the edge a reality - and that means automation and, most importantly, the ability to harness data locally and also to provide local storage and computing power and facilities and have an efficient model to deal with distributed data management.
Constantine Polychronopoulos adds that the continuing evolution of virtualisation is a fundamental requirement for the software-defined future, be that hypervisor-, containerised- or cloud native-based, as long as the performance is first-class and the tools are available to be able, seamlessly, to abstract physical infrastructure from virtual infrastructure. This will the focus on the virtual infrastructure and concomitant value-added services. Cloud-native containers are a natural evolution of virtualisation and VMware is integrating cloud native technology to provide a view of cloud-native abstraction to the user even though it runs on the VMware stack which comes with benefits including full segmentation, security and microservices that virtualisation provides together with all the advantages of cloud native and kubernetes-based workloads.
Filmed at DSP Leaders Forum 2019
Sign up to receive TelecomTV's top news and videos, plus exclusive subscriber-only content direct to your inbox.