By now it’s apparent that carriers are in the process of rebuilding their networks with a software-defined approach. Software-defined networks (SDNs) coupled with network function virtualization (NFV) software running on advanced processors will not only reduce the cost of delivering network services, it can pave the way for advanced, gigabit applications such as 5G. But how do we assure service in a more demanding, virtualized world?
Service assurance is the key to the success of SDN. It will drive automated control into the network – including across domains such as the data center and optical networks – and enable telecom carriers high-quality network performance. This is why leading service providers are starting to ramp up their deployment of SDN tools.
The effort to assure services can deliver the kind of quality that users expect will depend on using virtualization capabilities to microsegment network applications, analyze them, and automate changes in real time.
SDN Service Assurance ROI
Naturally, carriers face challenges in implementing virtualization and its related functions including microsegmentation, analytics, and network control at scale. While virtualization has matured in the data center, it’s more complicated when it extends into the carrier network across multiple domains including multi-vendor equipment, including optical.
To achieve this goal, SDN vendors will have to move to assurance tools that integrate IP and optical networks. It also has to be integrated with service fulfillment and be automated. This is a topic SDxCentral has written about in its report on LSO. Building and deploying the underlying modern network infrastructure may prove to be relatively simple compared to building the management framework required to manage SDN service assurance for networks at scale.
The return on investment is there. By building SDN-enabled services such as bandwidth calendaring and bandwidth on-demand services, carriers can optimize their networks and make the most out of the infrastructure.
Distributed Assurance Needed
Carriers of all types of and sizes are now wrestling with how best to approach this next major challenge of dynamic service assurance – and a wide range of tools and software are involved. The question is – how does one attain wide-ranging service assurance, that works across multiple domains with equipment from many different vendors?
The answer is that service assurance needs to be distributed in the telco cloud, offering a centralized software platform developed by third parties. These platforms create rules that can be executed in specific domains to “close the loop” on service assurance. This is a goal that Nokia hit on this week with the addition of service assurance to its Network Services Platform (NSP).
Leading carriers are clearly thinking how to build a more open ecosystem that can integrate SDN-based service assurance tools. As an example, AT&T is leaning toward a preference for using open source software. AT&T has outlined its plans for a larger open source control, orchestration, management and policy project, dubbed ECOMP, through which the carrier plans to automate support for service delivery, service assurance, performance management, fault management associated with managing its SDN.
In addition to developing their own platforms, AT&T and other carriers are also working with specific SDN vendors to achieve service assurance goals. All of the major SDN vendors are working on this area. A recent report from IHS Markit identified Ciena, Cisco and Nokia as the leading providers of SDN orchestration software. Other key players in service assurance, including Netscout and Accedian, are integrating monitoring and assurance tools with SDN.
Service Assurance in the Third Network
The Metro Ethernet Forum refers to the management layer needed to orchestrate virtualized networks as The Third Network. Based on the idea that the network should be managed a service, the idea is that the monitoring and managing network virtualization will require dedicated circuits to monitor traffic patterns in a way that advanced analytics tools can then leverage to optimize traffic flows. Once those patterns are optimized, carriers will then be able to guarantee service level agreements (SLAs) to the network microsegment level.
The core issue carriers now face is to what degree they will want to engineer service assurance tools from the ground up themselves. It may be feasible today for a major carrier to stitch various open source components to create an SDN.
As such it’s likely that tier one carriers may start to rely more on commercial vendor expertise. As for tier two carriers and cloud service providers (CSPs) that are trying to deliver “over-the-top” network services that do an end run around carriers, reliance on commercial vendor expertise to enable service virtualization assurance is all but a guaranteed requirement.
In the meantime, as service virtualization assurance issues start to raise their ugly head it’ll be interesting to see to what degree carriers will actually be able to deliver on promises to deliver 5G networking services by the end of the decade, which at this juncture is now just over 1,000 days away.