JOHN WILLIAMSON DISAGGREGATED NETWORKS
TO DO LIST… Achieving widespread deployment of advanced open disaggregated networking solutions doesn’t appear to be exactly straightforward. Parts of the deal are still missing or in ongoing development. “One of the challenges is that there is no one set of specifications that address all service provider application requirements. There is a very broad spectrum of use cases,” offers Xenos. “A second challenge is ensuring that all system vendors adopt a common set of interfaces, and that there is consistent translation of the features into the products so that they all behave in the same way. Integration testing is required to make sure that the network behaves as the service provider expects it to.” As an illustration, Xenos describes how different vendors’ products can define the routing of a signal across a specific path in different ways; with one product you might need to enter the nodes to be included but in another you might need to enter nodes to be excluded. “For larger more complex networks, network management and control along with service orchestration and automation have probably been the most challenging,” admits Hughes. “One of the significant challenges is the development, testing, and standardisation of open APIs, both north- and south- bound,” suggests Brown. “Another very important element is development of common equipment models for configuration and monitoring.” Alarm correlation is also an issue that needs to be addressed according to Shore. “Sometimes a transponder can see an error due to an issue on the line system,” he explains. “You need alarm correlation to prevent technicians from chasing down transponder alarms that are the result of optical line system failures, and vice versa.” Mixed blessings? It’s also possible that the open disaggregation story may not end equally
“While many vendors - including Fujitsu - are embracing concepts of disaggregation and view OLS/OpenROADM as opportunities, it can pose a threat to some incumbent vendors’ market share, creating friction towards widespread adoption of ODTN,” adds Hughes. Even so, no one in the industry is expecting the open disaggregated optical network train to be de-railed. “It will be exciting to see ODTN and open optical networks move from PoCs to commercial network deployment announcements,” says Brown. “The innovators and early adopters are clearing the operational issues and proving the revenue-generating benefits that the pragmatists need to pull the trigger and unlock the potential value of their network.” “The number one attraction of disaggregation is accelerating the speed of innovation,” concludes Shore. “That provides benefits for everyone.” OPEN PLAN The Open and Disaggregated Transport Network (ODTN) kicked off in May 2018 as an initiative led by the operator members of the Open Networking Foundation (ONF). The ODTN scheme leverages the ONF’s Open Network Operating System (ONOS®) SDN Controller, and draws on efforts such as the Transport API (TAPI), OpenConfig, Telecom Infra Project (TIP) and the work of the OpenROADM MSA. The scope of the ODTN, as described in the relevant ONF Project Charter, encompasses: • Disaggregated DWDM systems, including but not limited to transponders and open line systems, amplifiers, multiplexers, all-optical switches and ROADMs • Open source network operating system for control and configuration of the DWDM system • Open and common data models, APIs and protocols. “In terms of ODTN there’s been a lot of work completed to define a partial disaggregation model which separates the open optical terminal, which would be like the transponders which are controlled with an OpenConfig interface, and the open line system which has the SDN controller and is controlled through a TAPI interface,” summarises Xenos. “With initiatives and efforts such as ODTN, by controlling the transponders and optical line systems and using common data models, it’s going to simplify and accelerate the process of adding new equipment to the network from any vendor, and accelerate the process of prototyping and turning up new services for end users.” Xenos also points out that a specific objective of one of ODTN’s operator sponsors was “full operational automation”.
any optical line system. The evolution to transponder that could operate independently from line systems really began around the second-generation coherent transceivers capable of 200G transmission. Those transceivers were extremely automated and adaptable – a trend that has only accelerated as we’ve progressed to the latest – 5th – generation of coherent optics.” DEGREES OF SEPARATION Opinion varies as to how far disaggregation can and should go. Initially, recalls Shore, network operators started deploying technically disaggregated networks even if both the line system and the disaggregated transponder were from the same vendor. This practice continues today, although there’s now more of a move towards putting third party transponders over their deployed line systems to take better advantage of the latest technologies. Asked if he thought the end was in sight for service providers to solely deploy complete, single vendor disaggregated solutions, Shore replied: “Without question. And likely within the next few years.” Hughes reckons most medium-to- large customers today require optical networks to be at least partially open and disaggregated to ensure that they have the flexibility to choose the best and most economical platforms as they become available, whether it be line systems or transponders. Can disaggregation move to another level? “A more fully-disaggregated solution not only physically separates the line system from transponders, it also physically separates functional elements within a system,” states Hughes. “For example, the line system can be broken down into the ROADM, add/drop complex, Raman amplifiers and OTDR as separate physical solutions. This extra flexibility allows service providers to better right-size the sites in their network.” By contrast, Shore believes there are limits to amount of useful disaggregation that can be achieved in the optical layer as network operators are unlikely to mix- and-match different vendors’ equipment at this layer. For example – they would be unlikely to pair one vendor’s ROADM with another vendor’s amplifier.
happily for all solutions suppliers. “Open disaggregation helps the
innovative specialist who does one thing really, really, well, but doesn’t necessarily have a broader network solution,” contends Shore. “It hurts vendors who go a mile wide and an inch deep, and who try and win by selling networks and not the best technology.”
Rob Shore Senior Marketing VP, Infinera
Dave Brown Senior Manager, Optical Product Marketing, Nokia
Rob Hughes Optical Solutions Lead, Fujitsu Network Communications
Helen Xenos Portfolio Marketing, Ciena
20
| ISSUE 22 | Q4 2020
www.opticalconnectionsnews.com
Made with FlippingBook Publishing Software