I am excited about curved-screen 8K televisions, sophisticated home robots, inexpensive tablet computers, innovative smart watches and sensor-laden fitness bands, self-driving electric cars, and high-flying camera-toting drones. Perhaps you are too: Those are some of the must-have gadgets for 2015, as demonstrated at January’s International CES conference in Las Vegas. If you want a good overview of the best of CES, see Mashable’s coverage, and be prepared to start buying new gadgets for your home. However, if you want to know what’s best for telecommunications in 2015, forget wearables and other Internet-connected gizmos. Let’s talk about networks!
Lifecyle Service Orchestration (LSO) will make SDN & NFV accessible at all levels.
The complexity of a network can be incredible. Think about the challenges of defining new services. Creating and implementing them, whether through traditional network devices, SDN-enabled devices, or a hybrid enterprise or carrier network. You also have to manage those services, ensure adherence to service level guarantees, bill for them – and then tear those services down when they’re no longer required.
LSO is the key to creating and managing services in a complex environment, especially when SDN and NFV are part of the mix. Defining LSO capabilities and supporting APIs is critical to overcoming today’s operational support system (OSS) challenges and bridging the gap to future SDN/NFV-enhanced networks. LSO can instruct the network at every spot not only to implement services, but also to gather data for real-time visualization and management. In fact, you can think of LSO as bringing together the best of traditional network consoles and Big Data analytics. An example is CENX, whose Cortx Service Orchestration can take over end-to-end lifecycle management of a modern network, as well as help with migration and SLA enforcement issues.
A visionary initiative in this area is the MEF’s “Third Network,” which combines the on-demand agility and ubiquity of the Internet with the performance and security assurances of carrier Ethernet 2.0 (CE 2.0). Third Network services promise to offer a dynamic, cloud-centric experience with unprecedented levels of user control over network resources, with connectivity orchestrated between both physical and virtual service endpoints and across multiple providers. The MEF Third Network initiative includes creation of interoperability specifications that vendors can use to orchestrate sharing and combining resources across, well, everything. The MEF’s scope encompasses network equipment makers, software companies, telcos, data center suppliers, cloud service providers, even computer manufacturers. The MEF is working in collaboration with CEF and other industry stakeholders that form the MEF Unite initiative to define essential LSO and management capabilities necessary to achieve the key aims of the Third Network.
High-speed metro will drive enterprise connectivity.
Let’s think big. An enterprise can run all the fiber it wants through its building or around its campus. What about when connecting directly to, say, cloud-based data centers over Ethernet? Too often, connectivity is fast but not fast enough to tear down the performance walls between an enterprise data center and a cloud center – or between different cloud providers.
High-speed interconnections, based on optical technology, are at the heart of the modern cloud data center, and also the connection between those data centers and their performance-driven enterprise customers. One of the company’s in this aspect of high-speed metro networking is Infinera, which has staked out a position with high-performance, low power consumption, and port density. For example, there’s a new product family called Cloud Xpress, which is designed for the Metro Cloud, the virtual transport network that interconnects multiple data centers within a metro area.
Service level guarantees will drive everyone crazy.
Nearly every enterprise is relying upon cloud services of some sort or another. It’s all something-as-a-service, whether it’s infrastructure, security, analytics, virtualization, storage… it goes on and on. In the next year, we will continue to see a proliferation of everything as-a-service, with boundaries that blur. While organizations like the CloudEthernet Forum are dedicated to standardizing definitions of cloud services, the reality is that enterprise CIOS and service provider marketers will have trouble speaking the same language.
There is a huge downward push on pricing, but at the same time there’s a huge upward spike in service utilization and complexity. What does that mean? Crazy service level guarantees that will be hard to understand, hard to measure and hard to enforce, especially when many of those SLAs span multiple service providers. Separate clouds. Interdependent APIs. Hard-to-predict performance and availability. The European Commission is studying Cloud Service Level Agreement standardization guidelines, as part of its Europe 2020 initiative. This can’t come too soon, and with luck, we’ll see real global standards take form in 2015.
IoT will begin to consume significant bandwidth and other back-end resources.
The bulk of IoT devices today – fitness bands and smart watches, home thermostats and smart televisions – are only the beginning. Many wearable devices communicate with a smartphone via Bluetooth; apps on the phone talk to the cloud, but there aren’t many packets. The same is true with other IoT devices located in the home or office, which talk to the router via WiFi or (rarely) a physical cable. Remember those?
Today, the IoT is barely measurable. That will change and grow. Devices equipped with SatNav/GPS capabilities send location information frequently. Those are small messages, but there are a lot of them. Devices equipped with cameras will stream data into the cloud. That will be huge volume – think Netflix in reverse, and without the benefit of edge caching. More mobile devices will have their own cellular data connections. A study published in late 2013, from the Polytechnic Institute of New York University, highlighted an exponential jump in IoT bandwidth consumption between now and 2020. A story in Campus Technology cites Gartner predicting that IoT devices will grow to 26 billion units by 2020, from 0.9 million in 2009. To quote from the story, “How Will Campus Networks Handle the Internet of Things’ 26 Billion Devices?”:
“IoT deployments will generate large quantities of data that need to be processed and analyzed in real time,” noted Fabrizio Biscotti, research director at Gartner. “Processing large quantities of IoT data in real time will increase as a proportion of workloads at data centers, leaving providers facing new security, capacity and analytics challenges.”
It’s a thought that many IT managers, and cloud providers, may not be thinking about. Gear up the SDN, light up the metro fiber, and pay attention to what’s happening at CES… because all those massively distributed packets are about to hit our networks.