This paper proposes a traffic-driven optical IP networking architecture for service provider networks. Its design is derived from the optical GMPLS architecture, which provides high performance but is not scalable since both optical paths and IP routes need to be arranged in a mesh topology. To improve scalability, we first modified the configuration so that paths and routes can be arranged in a tree topology. However, this approach may degrade performance due to traffic concentration at each tree's root. To prevent such performance degradation, we further modified the architecture so that both cut-through optical paths and cut-through IP routes can be assigned reactively, according to traffic demand, and these can work together in cooperation. As a result, our architecture achieves both high performance and scalability, in that the whole network performance can be maintained without a massive increase in the number of optical paths and IP routes, even if the number of customer networks grows.

Packet processing on commodity hardware is a cost-efficient and flexible alternative to specialized networking hardware. However, virtualizing dedicated networking hardware as a virtual machine (VM) or a container on a commodity server results in performance problems, such as longer latency and lower throughput. This paper focuses on obtaining a low-latency networking system in a VM and a container. We reveal mechanisms that cause millisecond-scale networking delays in a VM through a series of experiments. To eliminate such delays, we design and implement a low-latency networking system, kernel busy poll (KBP), which achieves three goals: (1) microsecond-scale tail delays and higher throughput than conventional solutions are achieved in a VM and a container; (2) application customization is not required, so applications can use the POSIX sockets application program interface; and (3) KBP software does not need to be developed for every Linux kernel security update. KBP can be applied to both a VM configuration and a container configuration. Evaluation results indicate that KBP achieves microsecond-scale tail delays in both a VM and a container. In the VM configuration, KBP reduces maximum round-trip latency by more than 98% and increases the throughput by up to three times compared with existing NAPI and Open vSwitch with the Data Plane Development Kit (OvS-DPDK). In the container configuration, KBP reduces maximum round-trip latency by 21% to 96% and increases the throughput by up to 1.28 times compared with NAPI.

This paper proposes a managed IP multicast platform that enables IP multicast services to be used for business. Nowadays, many business applications have switched from traditional network platforms to the IP platform. Among these applications, one-to-many or many-to -many types of applications are especially essential to business users. These applications may use IP Multicasting for transmitting data to many users. However, for business applications, it is difficult to use the present IP Multicast services, because they lack many requirements for business usage. The requirements are address management, authentication, time management, and guaranteed throughput. To satisfy the business users, we made the design of a managed IP multicast platform that will meet these requirements. Our platform, which separates the routing control layer and the packet forwarding layer, is called GMN-CL (Connection Technologies for Global Mega-media Network). The routing control layer manages routing information and controls network routing centrally, so it can understand the whole network situation and perform efficient routing. The packet forwarding layer can concentrate completely on forwarding, so the forwarding speed and copying speed is higher than when using routers. We have implemented our design of a managed IP multicast platform over GMN-CL. This paper reports the system design, implementation, and evaluation.