公开(公告)号：US09455911B1

公开(公告)日：2016-09-27

申请号：US14143871

申请日：2013-12-30

Applicant: Google Inc.

Inventor： Leon Poutievski ,  Amin Vahdat

IPC: H04L12/24 ,  H04L12/741

CPC classification number: H04L45/745 ,  H04L45/02 ,  H04L45/28 ,  H04L45/64

Abstract: Aspects of the disclosure relate generally to switches responding to master controllers before establishing a secure connection in an in-band controlled network. In order to do so, a computing device may receive a request packet from a transmitting device, the request packet originating from a controller computer. When the request packet enters a designated port on the computing device, the computing device may identify routing data from the received request packet, the routing data indicating a path to respond to the controller computer. The computing device may next populate a table using the identified routing data of the request packet. The computing device may send a response packet to the controller computer using the identified routing data in the populated table. From here, a connection may be established between the computing device and controller computer.

Abstract translation: 本公开的方面通常涉及在在带内受控网络中建立安全连接之前响应主控制器的交换机。 为了这样做，计算设备可以从发送设备接收来自控制器计算机的请求分组。 当请求分组进入计算设备上的指定端口时，计算设备可以从接收到的请求分组识别路由数据，路由数据指示响应于控制器计算机的路径。 计算设备可以使用所识别的请求分组的路由数据来填充表。 计算设备可以使用填充表中的所识别的路由数据向控制器计算机发送响应分组。 从这里可以在计算设备和控制器计算机之间建立连接。

公开(公告)号：US09419887B1

公开(公告)日：2016-08-16

申请号：US13921585

申请日：2013-06-19

Applicant: Google Inc.

Inventor： Abdul Kabbani ,  Amin Vahdat

IPC: G06F13/00 ,  H04L12/729 ,  H04L12/721 ,  H04L12/701

CPC classification number: H04L45/125 ,  H04L45/00 ,  H04L45/12

Abstract: Systems and methods for managing computer network activity are disclosed. A data flow to be transmitted between a source node and a destination node can be identified. An initial graph representing the computer network can be determined. A shortest path between the source node and the destination node can be identified based on the graph. Bandwidth can be allocated to communication links along the shortest path between the source node and the destination node. It can be determined whether a bandwidth requirement associated with the data flow is satisfied by the allocated bandwidth. A residual graph can be determined. The steps of identifying a shortest path between the source node and the destination node based on the residual graph and allocating bandwidth to communication links along the shortest path between the source node and the destination node can be repeated.

Abstract translation: 公开了用于管理计算机网络活动的系统和方法。 可以识别要在源节点和目的节点之间传输的数据流。 可以确定表示计算机网络的初始图形。 可以基于图来识别源节点和目的地节点之间的最短路径。 带宽可以分配给沿着源节点和目的节点之间的最短路径的通信链路。 可以确定与分配的带宽是否满足与数据流相关联的带宽需求。 可以确定残差图。 可以重复基于剩余图来识别源节点和目的地节点之间的最短路径并且沿着源节点和目的地节点之间的最短路径向通信链路分配带宽的步骤。

公开(公告)号：US09253117B1

公开(公告)日：2016-02-02

申请号：US13718638

申请日：2012-12-18

Applicant: Google Inc.

Inventor： Leon Poutievski ,  Amin Vahdat

IPC: H04L12/933 ,  H04L12/24 ,  H04L12/751 ,  H04L12/26

CPC classification number: H04L49/15 ,  H04L41/12 ,  H04L41/28 ,  H04L43/04 ,  H04L43/10 ,  H04L45/02 ,  H04L45/026

Abstract: A network topology system for reducing the hardware links in a centrally-controlled network having a plurality of network switches and a plurality of centralized controllers controlling those switches. The global network of switches and controllers is broken up into domains, wherein a domain contains a subset of controllers and switches, and a switch in a domain is controlled by one or more controllers. In-band hardware links, which carry both data and control information, are used to connect the controllers and switches in a domain, and thereby reduce the number of hardware links from that used in networks which have separate hardware links for both data and control information.

Abstract translation: 一种网络拓扑系统，用于减少具有多个网络交换机的中央控制网络中的硬件链路以及控制这些交换机的多个集中式控制器。 交换机和控制器的全球网络被分解成域，其中域包含控制器和交换机的子集，域中的交换机由一个或多个控制器控制。 携带数据和控制信息的带内硬件链路用于连接域中的控制器和交换机，从而减少了在具有用于数据和控制信息的单独的硬件链路的网络中使用的硬件链路的数量 。

公开(公告)号：US09246770B1

公开(公告)日：2016-01-26

申请号：US14143384

申请日：2013-12-30

Applicant: Google Inc.

Inventor： Amin Vahdat ,  Leon Poutievski

IPC: H04L12/28 ,  H04L12/24

CPC classification number: H04L41/04

Abstract: Aspects of the disclosure relate generally to selecting and identifying a primary controller server among a group of servers on a network. In order to do so, a first computer may set a first number to zero when a storage system associated with the first computer is initialized. The first computer may conduct an election protocol to select a primary controller computer from a plurality of computers. After selecting the primary computer, increase the first number by a given value and send it to the selected primary computer. The selected primary computer may set a second number corresponding to an absolute time when the storage system associated with the first computer is initialized. The selected primary computer may receive the first number from the first computer. The selected primary computer may send the numbers to a client device. After sending, the primary computer may establish a connection with the client.

Abstract translation: 本公开的方面通常涉及在网络上的服务器组中选择和识别主控制器服务器。 为了这样做，当与第一计算机相关联的存储系统被初始化时，第一计算机可以将第一个数字设置为零。 第一计算机可以执行选举协议以从多个计算机中选择主控制器计算机。 选择主计算机后，将第一个数字增加给定值并将其发送到所选的主计算机。 所选择的主计算机可以设置与初始化与第一计算机相关联的存储系统的绝对时间对应的第二数字。 所选择的主计算机可以从第一计算机接收第一号码。 所选择的主计算机可以将数字发送到客户端设备。 发送后，主计算机可能会建立与客户端的连接。

公开(公告)号：US09246760B1

公开(公告)日：2016-01-26

申请号：US13904804

申请日：2013-05-29

Applicant: Google Inc.

Inventor： Abdul Kabbani ,  Amin Vahdat

IPC: H04L12/28 ,  H04L12/24

CPC classification number: H04L41/082 ,  H04L41/0896 ,  H04L41/12

Abstract: Systems and methods for reassigning and implementing communication links in a network are described herein. The systems and methods described herein, reduce the amount of throughput loss that results when a network is expanded. The systems and methods generate distributions for the communication links that, when implemented, reduce the number of broken paths that occur in the expanding network. More practically, the systems and methods provide a heuristic for remapping communication links from a first link distribution to a second link distribution.

Abstract translation: 本文描述了用于在网络中重新分配和实现通信链路的系统和方法。 这里描述的系统和方法减少了网络扩展时产生的吞吐量损失量。 系统和方法生成通信链路的分发，当实现时，减少在扩展网络中发生的断开路径的数量。 更实际地，系统和方法提供了用于将通信链路从第一链路分发重新映射到第二链路分发的启发式。

公开(公告)号：US09210487B1

公开(公告)日：2015-12-08

申请号：US14644797

申请日：2015-03-11

Applicant: Google Inc.

Inventor： Xiaoxue Zhao ,  Amin Vahdat ,  Hong Liu

IPC: H04J14/00 ,  H04B10/08 ,  H04J3/14 ,  H04Q11/00

CPC classification number: H04Q11/0005 ,  H04L49/1515 ,  H04Q2011/0007 ,  H04Q2011/0024

Abstract: Embodiments provide a methodology for designing a large-scale non-blocking OCS using a multi-stage folded CLOS switch architecture for use in datacenter networks and fiber-rich backbone network POPs. One aspect employs a folded CLOS architecture because of its ease of implementation, enabling the topology to scale arbitrarily with increasing number of stages. The fraction of ports allocated for internal switch wiring (overhead) also increases with the number of stages. Design decisions are made to carefully optimize the insertion loss per module, number of ports per module, number of stages and the total scale required. Other embodiments include folded CLOS switch architectures having at least two stages. In one example, power monitoring may be included only on the leaf switches.

Abstract translation: 实施例提供了一种使用多阶段折叠CLOS交换机架构设计大规模非阻塞OCS的方法，用于数据中心网络和富含纤维的骨干网POP。 一个方面采用折叠式CLOS架构，因为其易于实现，使得拓扑结构随着阶段数量的增加而任意扩展。 分配给内部开关布线（开销）的端口的分数也随着级数的增加而增加。 进行设计决策，仔细优化每个模块的插入损耗，每个模块的端口数量，级数和所需的总尺寸。 其他实施例包括具有至少两个阶段的折叠CLOS开关架构。 在一个示例中，功率监视可以仅包括在叶片开关上。

公开(公告)号：US20150281066A1

公开(公告)日：2015-10-01

申请号：US14478217

申请日：2014-09-05

Applicant: Google Inc.

Inventor： Bikash Koley ,  Steven Padgett ,  Ankur Jain ,  Arjun Singh ,  Amin Vahdat ,  Mahesh Kallahalla ,  Mukarram Tariq

IPC: H04L12/741 ,  H04L12/947

CPC classification number: H04L45/74 ,  H04L45/02 ,  H04L45/742 ,  H04L49/25 ,  H04L49/70

Abstract: An autonomous network and a corresponding routing method include determining routing paths by a controller, and providing the determined routing paths to a data packet processor located remotely from the controller. The data packet processor routes outgoing data packets, based on information from the controller, through a plurality of switches remotely from the data packet processor. Each switch includes a plurality of network interfaces. For an outgoing data packet, the data packet processor determines a network interface over which to transmit the data packet, and adds an indication of the determined network interface in a header of the data packet. The data packet processor forwards the modified data packet to the switch including the determined network interface. The switch identifies the network interface based on the indication, and transmits the outgoing data packet over the identified network interface.

Abstract translation: 自主网络和对应的路由方法包括由控制器确定路由路径，并将确定的路由路径提供给远离控制器的数据分组处理器。 数据分组处理器基于来自控制器的信息，通过远程从数据分组处理器的多个交换机来路由输出数据分组。 每个交换机包括多个网络接口。 对于输出数据分组，数据分组处理器确定发送数据分组的网络接口，并将确定的网络接口的指示添加到数据分组的报头中。 数据分组处理器将修改的数据分组转发到包括确定的网络接口的交换机。 交换机根据指示识别网络接口，并通过识别的网络接口发送出站数据包。

公开(公告)号：US10097474B1

公开(公告)日：2018-10-09

申请号：US13886069

申请日：2013-05-02

Applicant: Google Inc.

Inventor： Abdul Kabbani ,  Amin Vahdat

IPC: G06F15/16 ,  H04L12/891

Abstract: A system includes a plurality of rate limiters and a rate limiter management module. The rate limiter management module causes a processor to analyze and compare a plurality of flows each having at least one aggregation attribute, to aggregate a first group of the plurality of flows into a first aggregate flow based on the at least one aggregation attribute, and to assign the first aggregate flow to a rate limiter having at least one membership criterion. The first aggregate flow includes flows having the at least one aggregation attribute that satisfies the at least one membership criterion.

公开(公告)号：US09942156B1

公开(公告)日：2018-04-10

申请号：US15183228

申请日：2016-06-15

Applicant: Google Inc.

Inventor： Junlan Zhou ,  Min Zhu ,  Amin Vahdat

IPC: H04L12/26 ,  H04L12/803 ,  H04L12/721 ,  H04L12/741 ,  H04L12/733

CPC classification number: H04L47/125 ,  H04L45/20 ,  H04L45/38 ,  H04L45/74 ,  H04L47/11

Abstract: Exemplary embodiments allocate network traffic among multiple paths in a network, which may include one or more preferred paths (e.g. shortest paths) and one or more alternative paths (e.g., non-shortest paths). In one embodiment, network traffic in form of flows may be allocated to the preferred paths until the allocation of additional network traffic would exceed a predetermined data rate. Additional flows may then be sent over the alternative paths, which may be longer than the preferred path. The paths to which each flow is assigned may be dynamically updated, and in some embodiments the path assignment for a particular flow may time out after a predetermined time. Accordingly, the flow traffic of each path may be balanced based on real-time traffic information.

公开(公告)号：US09503395B1

公开(公告)日：2016-11-22

申请号：US14457392

申请日：2014-08-12

Applicant: Google Inc.

Inventor： Junlan Zhou ,  Amin Vahdat ,  Arjun Singh

IPC: H04L12/58 ,  H04L12/933 ,  H04L12/54 ,  H04L12/823

CPC classification number: H04L49/1515 ,  H04L12/56 ,  H04L47/32 ,  H04Q11/08 ,  H04Q2213/1338

Abstract: The present technology considers multi-stage network topologies where it is not possible to evenly stripe uplinks from a lower stage of the network topology to switching units in an upper stage of the topology. This technology proposes techniques to both improve overall throughput and to deliver uniform performance to all end hosts with uneven connectivity among the different stages while delivering uniform performance to all hosts. To achieve improved network performance in case of asymmetric connectivity, more flows may be sent to some egress ports than to others, thus weighting some ports more than others, resulting in Weighted Cost Multi Path (WCMP) flow distribution.

Abstract translation: 本技术考虑到多级网络拓扑，其中不可能均匀地将上行链路从网络拓扑的较低级别划分到拓扑的上级中的交换单元。 该技术提出了提高总体吞吐量并为所有终端主机提供统一性能的技术，不同阶段之间的连接不均衡，同时为所有主机提供均匀的性能。 为了在不对称连接的情况下实现改进的网络性能，可以向某些出口端口发送比其他端口更多的流，因此比其他端口更多地加权一些端口，导致加权成本多径（WCMP）流分布。