<ruby id="txhzx"><ins id="txhzx"><meter id="txhzx"></meter></ins></ruby>

    <noframes id="txhzx"><form id="txhzx"><nobr id="txhzx"><th id="txhzx"></th></nobr></form>

    <address id="txhzx"><nobr id="txhzx"><meter id="txhzx"></meter></nobr></address>
    <address id="txhzx"></address>

      <form id="txhzx"><listing id="txhzx"><meter id="txhzx"></meter></listing></form>

      <address id="txhzx"></address>
      <em id="txhzx"></em>

          <form id="txhzx"><nobr id="txhzx"><meter id="txhzx"></meter></nobr></form>

          <address id="txhzx"><nobr id="txhzx"><progress id="txhzx"></progress></nobr></address>

          <noframes id="txhzx"><address id="txhzx"></address>

            Application-Driven WAN Solution

            2018-08-21

            Overview

            ● The WAN network is primarily responsible for connectivity between different network sites, such as headquarters and branches, branch and branch, DC and DC, etc., and most of that is "passive" and bearing the application flow, but with the development of the application mode, such as cloud computing, mobile and applications, the flow method has changed. The network primarily needs to remain "active" to adapt to changes in application modes, to provide varied services for different applications. However, the current network management is primarily for devices rather than businesses and that makes the network architecture so complex, difficult to extend, unable to adapt to WAN network development.

            ● New H3C, with its long-term technology accumulation and rich experience in the field of WAN, in-depth research into the difficulties that exist in applications and user needs, developed a new generation of WAN (Application-Driven Wide Area Network) solution, AD-WAN.

            New H3C AD-WAN Architecture

            ● AD-WAN is a standard SDN network architecture, which is a hierarchical, open, and flexible architecture. As shown in the figure above, the AD-WAN solution is divided into three layers: Network Forwarding, Controller and Apps, and Management Orchestration Provisioning.

            o Network Forwarding, network equipment controlled and managed by SDN controllers, instead of a more traditional method, like SNMP, NETCONF and CLI. It includes SDN architecture methods, such as of BGP-LS, BGP Flowspec, PCEP, OpenFlow, etc. It supports Segment Routing and OpenFlow hardware forwarding and is optimized at the forwarding level to provide high-performance forwarding plane.

            o Controller and Apps: the entire architecture is based on ODL, which is an open source platform and supports various apps. According to multiple WAN network scenarios, such as a DCI network, backbone network, vertical industry, branch access, etc., customized and scenario apps are developed to meet the network requirements of users in different scenarios. The controller communicates with equipment through the standard south interface; provides customized API interface for users through the north interface, realizes and orchestrate system integration, and satisfies personalized user services.

            o Management Orchestration Provisioning: the orchestration layer, by invoking the app-provided API interface, realizes defining application strategies, management arrangements, real-time monitoring, visualization and troubleshooting, etc., simplifying operational management on the network.

            AD-WAN Solution Highlights:

            Automatic Equipment Deployment

            ● The AD-WAN Solution provides U-Disk Deployment solution. It integrates BIMS+IVM and generates U-Disk deployments. The devices automatically deploy by loading the configuration into the U-Disk, and then communicating with the AD-WAN simplifies the deployment process, especially for a large number nodes, reducing the cost of network construction, accelerating new services online.

            Multidimensional Visibility

            Application Visibility

            ● AD-WAN provides visibility to applications, viewing application traffic information, and bandwidth usage ratios, etc. It chooses an application and the application forwarding path visibility, so that users can keep track of the overall health of the application.

            Network Visibility

            ● AD-WAN provides a topology view to administrate networks. This means the transmission network is no longer a "black box". Users can troubleshoot quickly in a set period of time for WAN link problems and monitoring the underlying link quality.

            o Controller can collect physical information, including device info, device/link state, device resources, device/link historic information, etc.

            o Controller can collect network information, including neighbor ships, device/link states, device/link resources, etc.

            Fault Visibility

            ● The AD-WAN solution provides customized windows to show any content you need, including application traffic statistics and alarm statistics.

            Application & Policy Provisioning

            ● The network manager can define the application with a static or dynamic character, for example, 5-tuple definition, DPI. AD-WAN combines different applications with different policies to meet different service requirements. For example, it can initiate the delay policy + packet loss policy + bandwidth utilization for video meetings.

            Intelligent Traffic Steering

            ● The AD-WAN solution can pool link resources and centralize traffic controls. Traffic steering optimization automatically ensures that key services are of high enough quality, as defined by the application. The AD-WAN solution provides a variety of traffic steering policies:

            o Based on bandwidth utilization scheduling and double-link uplink, different services normally operate on the same link, but with the gradual increase of bandwidth utilization, the controller can optimize the network to change one service to another link when link utilization reaches its threshold.

            o Based on delay scheduling and double-link uplink, different services normally operate on the same link, and the controller can optimize the network to change one service to another link when a delay exceeds its threshold.

            o Based on packet loss scheduling and double-link uplink. different services normally operate on the same link, and the controller will optimize the network to change one service to another link when a link exceeds the threshold for the packet loss rate.

            o Based on period scheduling and double link uplink, periodic service scheduling guarantees business requirements and a certain period of time guarantees the reliability of high-priority business.

            ● In addition, all available link resources can be effectively utilized to improve link utilization, reduce WAN network OPEX costs, and protect user investment.

            Operation Environment

            Attribute

            Parameter

            Server configuration

            Processor: >=2.6GHz, >=8 per processor

            Memory: more than 64G

            Hard Disk: more than 512G

            Network port: more than 2*10GE ports

            Operating system: CentOS 6.5, Ubuntu 14.04

            Client

            Typical Application Scenarios

            Enterprise Branch / Pseudo wire

            Scenario Description:

            ● Hub-and-spoke two-level network, with mixed access of pseudo wires and Internet links, with a large number of branches and scattered locations.

            ● Mostly vertical traffic, services separated and isolated by ACL, DPI, etc.

            Key Requirements:

            ● Deployment automation, network visualization, simplified operations, and maintenance management.

            ● DPI identification, flexible routing, and improved bandwidth utilization.

            ● User access authentication, online behavior management and auditing.

            Typical Networking

            上海快三稳定计划_上海快三稳定计划|官网_首页