Yukihiro Nakagawa

A management method of IP multicast in overlay networks using openflow

Overlay networks stretch a Layer 2 network and increase mobility of virtual machines. VXLAN (Virtual eXtensible LAN) is one of Layer 2 overlay schemes over a Layer 3 net- work proposed in IETF and its definition covers 16M overlay networks or segments which solves 4K limitation of VLANs. However VXLAN uses IP multicast for the isolation of net- work traffic by tenant in the shared network infrastructure. IP multicast requires great amount of resources such as IP multicast table and CPU therefore the scalability is to be limited by handling of IP multicast. We propose to manage IP multicast in overlay networks using OpenFlow instead of using dynamic registration protocol such as IGMP. We describe our implementations of VXLAN controller, edge switch with VXLAN gateway and OpenFlow switch. Our method using OpenFlow eliminates periodical Join/Leave messages and achieves more than 4k tenants in our Layer 2 network at server edges, which was not possible before.

DomainFlow: practical flow management method using multiple flow tables in commodity switches

A scalable network with high bisection bandwidth and high availability requires efficient use of the multiple paths between pairs of end hosts. OpenFlow is an innovative technology and enables fine-grained, flow level control of Ethernet switching. However, the flow table structure defined by OpenFlow is not hardware friendly and the scalability is limited by the switch device. OpenFlow is also not sufficient for fast multipath failover. To overcome these limitations, we propose DomainFlow in which the network is split into sections and exact matches are used where possible to enable practical flow management using OpenFlow for commodity switches. We applied a prototype of DomainFlow to multipath flow management in the Virtual eXtensible LAN (VXLAN) overlay network environment. The total number of flow entries was reduced to 1/128 using currently available commodity switches, which was not possible before.

A Single-Chip, 10-Gigabit Ethernet Switch LSI for Energy-Efficient Blade Servers

The use of virtualization technology has been increasing in the IT industry to consolidate servers and reduce power consumption significantly. As a virtualization platform, a large-scale blade server is suitable because it can hold a dozen blades in a chassis with well managed configuration, enabling easy provisioning. To realize an energy-efficient blade server, the network component must deliver both high performance and reduced power consumption. We developed the fifth generation single-chip 10GbE switch LSI that supports 26 10GbE ports with built-in 10 Gb/s serial back plane interfaces. Using this highly integrated switch LSI, we also developed a single-wide 10GbE switch blade for the blade server. The switch blade delivers 100 percent more performance per watt than other 10GbE switch blades in the industry. This paper describes the features of the switch LSI, the high-speed IO circuit of its built-in interfaces and 10GbE switch blade.

Dynamic virtual network configuration between containers using physical switch functions for NFV infrastructure

Container-based virtualization simplifies the deployment of applications and will have a significant impact on the future of data center networks. In an NFV infrastructure, it is required to dynamically extend/reduce resources granted to the virtual network function (VNF) as needed. In this paper, we explored SDN technology to construct lightweight virtual networks between Linux containers in dynamic container placement to isolate tenant traffic. We used an SDN controller to distribute logical endpoint information to physical switches before the host physical location was determined for the container. And we used physical switch functions to promptly configure virtual networks between containers when a container was dynamically placed on a host and started in a split second. We also actively probed the container and automatically removed the virtual network when the container stopped. By using our prototype, virtual networks were automatically constructed with no interaction with the SDN controller as soon as 500 of the containers were started.

Managing storage flows with SDN approach in I/O converged networks

The demand for efficient use of information and communication technology (ICT) resources has recently been growing. Equipment costs at datacenters have been reduced to satisfy this demand and all ICT resources have been required to be controlled on demand. Network virtualization and software defined networking (SDN) technologies have been applied to datacenter networks for the same reason. On the other hand, efficient networks and operations also need to be provided, including traditional storage systems. Ethernet fabric and I/O convergence have been applied to satisfy these requirements. However, the SDN approach has mainly been applied to local area networks and flexible control has so far not been accomplished for converged networks. We introduced an SDN approach similar to OpenFlow and applied this to Fibre Channel over Ethernet (FCoE) networks, which is a method of achieving I/O converged networks. This method of controlling flows can coexist with OpenFlow based SDN flow control. We explain the differences between our method of flow control, OpenFlow, and path optimization with FCoE and their effects as examples in this paper.

A Hybrid Image Compression Method For Medical Images

This paper describes an efficient compression method for medical images. This method can compress images with small bit rates and generate reconstructed images of high quality. Discrete Cosine Transform (DCT) and Block Truncation Coding (BTC) are widely used in various fields. These methods, however, have the following problems. With the DCT method, the reconstructed image quality is quite good, except that the quality of sharp edges in the image is clearly inferior. The BTC method can reconstruct sharp edges, but the reconstruced image is not suitable for medical images. To solve these problems, we have proposed a new hybrid compression method. In this method an original image is divided into sub-images. The high frequency component of each sub-image is evaluated. If the component is small, DCT is performed on the sub-image. When the high frequency component is large, BTC is applied to the sub-image itself, then DCT is performed on the difference between the original and the reconstructed sub-images. In our experiments, this method proved to be effective.

A compression algorithm for medical images and a display with the decoding function

This paper describes an efficient compression method for medical images and a high-speed display with a decoding function. An image is divided into blocks first and then whether each block has sharp edges is examined. The discrete-cosine-transform coding (DCT), which is efficient for a part having small gray-level variation, is applied to a block having no sharp edge. The block-truncation coding (BTC) which preserves sharp edges is applied to a block having edges, and then the DCT is applied to their error images. This combined procedure makes a reconstruction of the whole image in a high quality. The display developed by the authors receives the compressed data from a host computer and reconstructs images of good quality at high speed using four pipeline-based microprocessors based on their method. Experimental results show that the image quality of parts without sharp edges is as high as the DCT method; and that of parts with edges is better than the DCT and BTC methods. The whole process, which includes transferring data from the host computer to the display, reconstruction, and display of the data, took less than 5 s for a 1/10 compressed image. This demonstrates an excellent performance of the proposed reconstruction display.

Prototype of image compression system for medical images

We previously developed an image reconstruction display for reconstructing images compressed by our hybrid compression algorithm. The hybrid algorithm, which improves image quality, applies Discrete Cosine Transform coding (DCT) and Block Truncation Coding (BTC) adaptively to an image, according to its local properties. This reconstruction display receives the compressed data from the host computer through a BMC channel and quickly reconstructs good quality images using a pipeline-based microprocessor. This paper describes a prototype of a system for compression and reconstruction of medical images. It also describes the architecture of the image compression processor, one of the components of the system. This system consists of the image compression processor, a host main-frame computer and reconstruction displays. Under this system, distributed processing in the image compression processor and the image reconstruction displays reduces the load on the host computer, and supplies an environment where the control routines for PACS and the hospital information system (HIS) can co-operate. The compression processor consists of a maximum of four parallel compression units with communication ports. In this architecture, the hybrid algorithm, which includes serial operations, can be processed at high speed by communicating the internal data. In experiments, the compression system proved effective: the compression processor compressed a 1k x 1k image in about 2 seconds using four compression units. The three reconstruction displays showed the image at almost the same time. Display took less than 7 seconds for the compressed image, compared with 28 seconds for the original image.

Flow-Aware Congestion Control to Improve Throughput under TCP Incast in Datacenter Networks

Data enter networks (DCNs) host cloud computing and various applications. In this environment, there are mice and elephant flows and the partition/aggregate communication pattern is common, so TCP in cast may often occur. Massive mice flows directly affect the elephant flows under in cast, and the standard TCP mechanism or SDN approach with centralized controller is insufficient to solve this problem. In this paper, we focused on how a flow-aware congestion control method that takes into account flow characteristics can improve throughput under in cast. It is required to establish our long-term goal that provides different network level services based on flow characteristics. We propose a switch-based method that identifies target flows and provides different congestion control using explicit congestion notification (ECN). Although our method uses ECN, it does not require any change to TCP stacks on end nodes or a dedicated queue for the targets. In our experiments, we focused on the throughput of flows, and selected elephant flows as targets because their throughput is more sensitive than mice flows. We compared the experimental results from our method to those of standard TCP, ECN, and a dedicated queue for elephants. Our method improved the aggregate throughput of elephant flows by 1.7× to 5.1× and the good put of elephant flows by 1.9× to 5.4× compared to those of the above methods under our in cast scenario.

Surface correspondence based on three-dimensional structure inference in animation images

The authors propose a method of correspondence between surfaces of two successive frames in cell animation images using estimated three-dimensional structures. In this method, local three-dimensional relationships between adjacent surfaces are first inferred from each two-dimensional input image. Then, each surface in an image is made to correspond to one in another image by referring to the relationships. By this method, the candidates for correspondence can be limited, and surfaces can easily be matched with each other. Experiments show that the proposed method is effective: the accuracy of correspondence improves by more than 20%.<<ETX>>