At NANOG 79 a speaker from Google pointed out that if a component had a 50 year MTBF, Google would still experience a failure of this component every day due to the scale of their network. But very high availability of the network is critical to the customer experience, so how do you ensure availability while knowing parts of the system will fail? The key is building for resiliency.
This need for resiliency was highlighted during the response to the COVID pandemic over the past year as the world moved from the traditional office-centered network infrastructure to a distributed work from home environment. That created a jump of over 25% bandwidth demand in a weekend generated by newly critical cloud applications. As we emerge into the “Next Normal,” a continued focus on resiliency will mean that the focus will shift from “just-in-time” to “just-in-case.” This will place a focus on creating agility where it wasn’t possible before.
In a recent report from McKinsey on how resilience will be critical to success in the “Next Normal”, Katy George states that companies should “move with speed to digitize their operations end-to-end to create step changes in transparency and flexibility.” She goes on the suggest that “companies should use this resilience to create new value for customers and increase the company’s competitive advantage.” (The need for operational resilience | McKinsey)
Figure 1: The above steps show that the physical cross connect remains the broken link in the rapid provisioning of cloud services
While many elements of the network infrastructure responded incredibly well to the pressure of the pandemic, there are a few weak spots. Cloud-based applications such as Zoom were able to respond rapidly, but every cloud service relies on a physical fiber interconnect at some point to connect customer to servers and storage. As shown by the chart in figure 1 above, while servers, storage and switches can be provisioned instantly with software-defined-networking, the physical cross connect is still installed manually. While one challenge of manual installation is the speed, as anyone working with the network has learned from experience when someone touches a fiber more often than not, they cause a problem. These problems can range from a dirty connector to an error in the database to disrupting traffic on a neighboring fiber connection, all of which can be a challenge to troubleshoot and solve. Automating the fiber cross connect can fix this broken link and bring transparency and flexibility to the fiber layer.
The Telescent Network Topology Manager (NTM) is a NEBS Level 3 certified automated cross connect system and offers remote cross connect management. The system monitors, maps and controls the network during installation, reconfiguration and churn of cross connects to enable a Lights-Out Data Center with machine accurate inventory while preserving the value of the installed fiber plant. The NTM consists of 1,008 input and 1,008 output ports, each connected by a short, uninterrupted internal fiber connection. When a reconfiguration is requested, a robot-driven gripper removes an internal fiber from the original port and moves it to the desired new port. The patented algorithm identifies the unique path to route the fiber around the 1,007 other internal fibers in the system to establish a non-blocking, any-to-any connection. The Telescent NTM is strictly non-blocking and completely automated configuration and reconfiguration is possible with the Telescent system without any on-site, manual labor using the original external fiber connections. All connections in the Telescent NTM are latched, maintaining connection in the case of any power outage and the system is fully field maintainable. The NTM can be configured in a leaf-spine network to allow scaling to >10,000 cross connects while still preserving the ability for any-to-any connectivity. Since the reconfigurations are controlled by software, all connections are recorded in a database and the state of any connection can be verified at any time providing machine accurate inventory for accurate billing to customers while also avoiding errors with reconfiguration. Power monitoring is included with the NTM, and an optional OTDR system allows additional diagnostic capability of the attached fiber plant.
By automating the fiber interconnection, the Telescent NTM brings flexibility and visibility to the cross connect layer. Combined with existing SDN controller, this allows full automation and control across the entire network. This also allows cloud service providers to offer new benefits to customers, such as the option for rapid proof of concepts or shorter-term contracts that can be installed or decommissioned quickly. Together, this new flexibility should improve the company’s value to their customers and increase its competitive advantage.
To conclude with another quote from the McKinsey article, “resiliency is being ready for tomorrow,” says George, “that’s how long-term strategic advantage is won.” The Telescent NTM bring agility where it wasn’t possible before and creates the flexibility to respond to the needs of tomorrow.