公开(公告)号：US08842577B2

公开(公告)日：2014-09-23

申请号：US13471724

申请日：2012-05-15

Applicant: Ali Sajassi ,  Samer Salam ,  Keyur Patel

Inventor： Ali Sajassi ,  Samer Salam ,  Keyur Patel

IPC: H04L12/28

CPC classification number: H04L45/66 ,  H04L45/04 ,  H04L45/50 ,  H04L61/6022

Abstract: In one embodiment, a method includes obtaining an indication that a device is attached to a provider edge. The device has a media access control (MAC) address, wherein the device was previously attached to a first provider edge (PE) of a Multiprotocol Label Switching (MPLS) network. The method also includes issuing a first advertisement that identifies the MAC address. The first advertisement includes a first MAC address mobility attribute arranged to indicate a number of times the MAC address has moved with respect to the MPLS network.

Abstract translation: 在一个实施例中，一种方法包括获得设备附接到提供商边缘的指示。 该设备具有媒体访问控制（MAC）地址，其中该设备先前连接到多协议标签交换（MPLS）网络的第一提供商边缘（PE）。 该方法还包括发出标识MAC地址的第一广告。 第一广告包括被配置为指示MAC地址相对于MPLS网络移动的次数的第一MAC地址移动性属性。

公开(公告)号：US08804496B2

公开(公告)日：2014-08-12

申请号：US13315948

申请日：2011-12-09

Applicant: Jim Guichard ,  Luca Martini ,  Keyur Patel ,  Thomas David Nadeau

Inventor： Jim Guichard ,  Luca Martini ,  Keyur Patel ,  Thomas David Nadeau

IPC: G01R31/08

CPC classification number: H04L45/04 ,  H04L1/22 ,  H04L45/22 ,  H04L45/28 ,  H04L45/42 ,  H04L45/44 ,  H04L45/50 ,  H04L45/68

Abstract: The protection of multi-segment pseudowires by utilizing backup paths is disclosed herein. Disclosed embodiments include methods that establish at least one backup path for multi-segment pseudowires, the establishing being performed prior to detection of failure in the primary path. Upon detecting a path failure, the detected failure is signaled to the head-end, a backup path is chosen, and reachability information associated with the chosen backup path is signaled across the backup path before reverse traffic is switched to the backup path. In other disclosed embodiments, apparatus are configured to establish, prior to detection of failure in the primary path, at least one backup path for the multi-segment pseudowire.

Abstract translation: 本文公开了通过利用备用路径保护多段伪线。 公开的实施例包括为多段伪线建立至少一个备用路径的方法，该建立是在检测到主路径中的故障之前执行的。 在检测到路径故障时，将检测到的故障信号发送到前端，选择备份路径，并且在将反向流量切换到备份路径之前通过备份路径发信号通知所选备份路径的可达性信息。 在其他公开的实施例中，设备被配置为在检测到主路径中的故障之前建立用于多段伪线的至少一个备用路径。

公开(公告)号：US08639822B2

公开(公告)日：2014-01-28

申请号：US12986929

申请日：2011-01-07

Applicant: Anantha Ramaiah ,  Keyur Patel ,  Shrirang Bage

Inventor： Anantha Ramaiah ,  Keyur Patel ,  Shrirang Bage

IPC: H04L29/06

CPC classification number: H04L49/355

Abstract: A method and non-transitory computer-readable medium are disclosed for extending a hold timer that binds an application-layer session when a transport-layer out-of-order message queue includes an out-of-order message for the application-layer session. An application receives an application-layer message from transport protocol logic that is configured to deliver in-order application-layer messages to the application. The received application-layer message is a next in-order application-layer message for an application-layer session that is bound by a hold timer. After an amount time has passed, the application detects an expiration of the hold timer. In response, rather than immediately tearing down the application-layer session, the application inspects an out-of-order queue of the transport protocol logic. The hold timer is extended when the out-of-order queue includes an out-of-order application-layer message for the application-layer session. In one aspect, the transport protocol logic includes an application programming interface that provides the application with access to the out-of-order queue.

Abstract translation: 公开了一种方法和非暂时性计算机可读介质，用于在传输层无序消息队列包括用于应用层会话的无序消息时扩展绑定应用层会话的保持定时器 。 应用程序从传输协议逻辑接收应用层消息，该消息被配置为向应用程序发送按顺序的应用层消息。 接收到的应用层消息是由保持定时器绑定的应用层会话的下一个按顺序应用层消息。 经过一段时间后，应用程序会检测到保持定时器的到期。 作为响应，应用程序不是立即拆除应用程序层会话，而是检查传输协议逻辑的乱序队列。 当无序队列包含用于应用层会话的无序应用层消息时，保持定时器被扩展。 在一个方面，传输协议逻辑包括向应用程序提供对无序队列的访问的应用编程接口。

公开(公告)号：US20120281539A1

公开(公告)日：2012-11-08

申请号：US13098628

申请日：2011-05-02

Applicant: Keyur Patel ,  Robert Raszuk ,  Ryan Bickhart ,  Alton Lo

Inventor： Keyur Patel ,  Robert Raszuk ,  Ryan Bickhart ,  Alton Lo

IPC: H04L12/26

CPC classification number: H04L45/026 ,  H04L45/04

Abstract: In an embodiment, a method comprises: receiving, at a data packet router, a path advertisement comprising information about an available path; determining whether the path advertisement comprises an originator identifier of an originator of the available path; in response to determining that the path advertisement comprises the originator identifier of the originator of the available path, determining whether the originator identifier of the available path is a router identifier of the data packet router, and in response to determining that the originator identifier of the available path is the router identifier of the data packet router, not accepting the path advertisement; wherein the method is performed by one or more processors of the data packet router.

Abstract translation: 在一个实施例中，一种方法包括：在数据分组路由器处接收包括关于可用路径的信息的路径通告; 确定路径广告是否包括可用路径的始发者的始发者标识符; 响应于确定路径广告包括可用路径的发起者的始发者标识符，确定可用路径的始发者标识符是否是数据分组路由器的路由器标识符，并且响应于确定所述可用路径的始发者标识符 可用路径是数据包路由器的路由器标识符，不接受路径广告; 其中所述方法由所述数据分组路由器的一个或多个处理器执行。

公开(公告)号：US20120209568A1

公开(公告)日：2012-08-16

申请号：US13026351

申请日：2011-02-14

Applicant: KARLA K. ARNDT ,  JAMES M. CAFFREY ,  KEYUR PATEL ,  ASPEN L. PAYTON

Inventor： KARLA K. ARNDT ,  JAMES M. CAFFREY ,  KEYUR PATEL ,  ASPEN L. PAYTON

IPC: G06F19/00

Abstract: Techniques are described for monitoring a performance metric. A multiple modeling approach is used to improve predictive analysis by avoiding the issuance of warnings during spikes which occur as a part of normal system processing. This approach increases the accuracy of predictive analytics on a monitored computing system, does not require creating rules defining periodic processing cycles, reduces the amount of data required to perform predictive modeling, and reduces the amount of CPU required to perform predictive modeling.

Abstract translation: 描述了用于监视性能指标的技术。 多重建模方法用于通过避免在作为正常系统处理的一部分发生的尖峰期间发出警告来改进预测分析。 这种方法提高了监视计算系统上预测分析的准确性，不需要创建定义周期性处理周期的规则，减少执行预测建模所需的数据量，并减少执行预测建模所需的CPU数量。

公开(公告)号：US07697439B2

公开(公告)日：2010-04-13

申请号：US11126486

申请日：2005-05-10

Applicant: Luca Martini ,  Keyur Patel

Inventor： Luca Martini ,  Keyur Patel

IPC: H04L12/26

CPC classification number: H04L45/123 ,  H04L45/04 ,  H04L45/12

Abstract: A method for determining a cumulative network layer reachability cost of traversing one or more autonomous network routing systems comprises receiving first network route information from an originating customer edge router, wherein the first network route information identifies a route in a customer network; determining a transit cost from the provider edge router to an endpoint associated with the originating customer edge router based upon a metric value received in a cost community attribute; repeating the determining step for each of one or more autonomous systems that lie in a path from the originating customer edge router to a destination customer edge router, to result in determining an accumulated transit costs across one or more autonomous systems; and determining, based at least in part upon the accumulated transit cost, a best path from the provider edge router to the endpoint associated with the originating customer edge router.

Abstract translation: 用于确定遍历一个或多个自主网络路由系统的累积网络层可达性成本的方法包括从始发客户边缘路由器接收第一网络路由信息，其中所述第一网络路由信息识别客户网络中的路由; 基于在成本共同体属性中接收到的度量值，确定从所述提供者边缘路由器到与所述始发客户边缘路由器相关联的端点的传输成本; 对位于从起始客户边缘路由器到目标客户边缘路由器的路径中的一个或多个自治系统中的每一个重复确定步骤，以导致确定跨越一个或多个自治系统的累积过境成本; 以及至少部分地基于累积的转移成本确定从提供商边缘路由器到与始发客户边缘路由器相关联的端点的最佳路径。

公开(公告)号：US07697416B2

公开(公告)日：2010-04-13

申请号：US11518382

申请日：2006-09-08

Applicant: Mike Shand ,  Stewart F. Bryant ,  David D. Ward ,  Gargi Nalawade ,  Robert Raszuk ,  Keyur Patel ,  Clarence Filsfils ,  Stefano B. Previdi

Inventor： Mike Shand ,  Stewart F. Bryant ,  David D. Ward ,  Gargi Nalawade ,  Robert Raszuk ,  Keyur Patel ,  Clarence Filsfils ,  Stefano B. Previdi

IPC: H04L1/00 ,  H04L12/26 ,  H04L12/28 ,  H04L12/56

CPC classification number: H04L45/04 ,  H04L45/02 ,  H04L45/28

Abstract: In one embodiment, an apparatus and method are described for constructing a repair path in the event of non-availability of a routing domain component of a routing domain comprising, as components, links and nodes. The apparatus is arranged to receive respective network repair addresses from each of the far-side and near-side advertising node for use in the event of non-availability of a routing domain component between the advertising node. The apparatus is further arranged to advertise the near-side advertising node network repair address to one or more far-side nodes via a path external to the routing domain.

Abstract translation: 在一个实施例中，描述了在路由域的路由域组件的不可用性的情况下构建修复路径的装置和方法，该路由域组件包括作为组件的链路和节点。 该装置被安排为从广播节点和广播节点之间的路由域组件不可用的情况下接收来自远侧和近侧广告节点的相应网络修复地址。 该装置还被布置成经由路由域外部的路径将近侧广告节点网络修复地址通告给一个或多个远侧节点。

公开(公告)号：US07656819B2

公开(公告)日：2010-02-02

申请号：US11267865

申请日：2005-11-04

Applicant: Keyur Patel ,  Chandrashekhar Appanna ,  John Scudder

Inventor： Keyur Patel ,  Chandrashekhar Appanna ,  John Scudder

IPC: H04L12/28 ,  H04L12/56 ,  H04J3/16 ,  H04J3/22

CPC classification number: H04L45/04 ,  H04L45/02 ,  H04L45/54

Abstract: A method of improving convergence in networks that use path vector protocols. In one embodiment using Border Gateway Protocol (BGP) for inter-domain route and reachability communication, nodes that need to remove routes using EBGP NLRI announcements send BGP WITHDRAW messages containing a new route originator attribute that identifies an originator of a specified withdrawn route. A receiving node removes the specified route from its routing information base (RIB), and matches the originator to other paths in the RIB. Matching paths are marked as ineligible for consideration in a best path computation, thereby preventing the use of routes that are unreachable but not yet withdrawn by an originating node. The approach causes faster BGP convergence, reduces unnecessary route flapping, and alleviates unnecessary route dampening.

Abstract translation: 一种改善使用路径向量协议的网络收敛的方法。 在使用边界网关协议（BGP）进行域间路由和可达性通信的一个实施例中，需要使用EBGP NLRI通知去除路由的节点发送包含标识指定撤销路由的发起者的新路由发起方属性的BGP WITHDRAW消息。 接收节点从其路由信息库（RIB）中删除指定的路由，并将发起方与RIB中的其他路径相匹配。 匹配路径被标记为不符合最佳路径计算的要求，从而防止使用始发节点不可达但尚未撤销的路由。 该方法引起BGP收敛速度更快，减少不必要的路由振荡，减轻不必要的路由衰减。

公开(公告)号：US07318108B2

公开(公告)日：2008-01-08

申请号：US11023056

申请日：2004-12-22

Applicant: Arjun Sreekantiah ,  Ruchi Kapoor ,  Keyur Patel ,  Robert Raszuk

Inventor： Arjun Sreekantiah ,  Ruchi Kapoor ,  Keyur Patel ,  Robert Raszuk

IPC: G06F15/173

CPC classification number: H04L45/02 ,  H04L45/04

Abstract: A method is disclosed for providing prioritized routing table convergence in Border Gateway Protocol (BGP) hosts, as part of a BGP process that is executed by a route processor in a network, wherein the processor hosts a plurality of BGP routing tables each associated with a specified sub-address family (SAFI). A convergence priority order value is created and stored in association with each routing table. As part of performing a convergence phase of BGP processing, the convergence priority order value of each routing table is obtained. The convergence phase is then performed for the routing tables only according to an order that is represented by the convergence priority order values. Therefore, convergence proceeds more quickly and values needed by other routing processes become available earlier.

Abstract translation: 公开了一种用于在边界网关协议（BGP）主机中提供优先级路由表收敛的方法，作为由网络中的路由处理器执行的BGP进程的一部分，其中处理器承载多个BGP路由表，每个BGP路由表与 指定的子地址族（SAFI）。 与每个路由表相关联地创建和存储收敛优先级顺序值。 作为执行BGP处理收敛阶段的一部分，获得每个路由表的收敛优先级顺序值。 然后，仅根据由收敛优先级顺序值表示的顺序为路由表执行收敛阶段。 因此，收敛进展得更快，其他路由进程所需的值越来越早。

公开(公告)号：US20070214280A1

公开(公告)日：2007-09-13

申请号：US11371522

申请日：2006-03-09

Applicant: Keyur Patel ,  Robert Raszuk ,  Clarence Filsfils

Inventor： Keyur Patel ,  Robert Raszuk ,  Clarence Filsfils

IPC: G06F15/173

CPC classification number: H04L45/22 ,  H04L45/04 ,  H04L45/124 ,  H04L45/28

Abstract: A technique institutes backup Border Gateway Protocol (BGP) paths for non-multipath BGP Fast Convergence in a computer network. According to the novel technique, a protecting node, e.g., a border router, calculates a best BGP path to a destination prefix, and also calculates one or more backup BGP paths (i.e., non-multipath paths) to the destination prefix. The best path is selected as a utilized path for the destination prefix, e.g., the path over which to send traffic destined for the prefix from the border router. Upon determining that the best BGP path has failed, one of the backup BGP paths is selected as the utilized path for the destination prefix.

Abstract translation: 技术为计算机网络中的非多路径BGP快速收敛提供了备份边界网关协议（BGP）路径。 根据新技术，保护节点（例如边界路由器）计算到目的地前缀的最佳BGP路径，并且还计算到目的地前缀的一个或多个备用BGP路径（即，非多路径路径）。 选择最佳路径作为目的地前缀的利用路径，例如，从边界路由器发送去往前缀的流量的路径。 在确定最佳BGP路径失败后，其中一条备份BGP路径将被选择作为目标前缀的使用路径。