公开(公告)号：US09608938B2

公开(公告)日：2017-03-28

申请号：US14457660

申请日：2014-08-12

Applicant: Sivakumar Venkatesan ,  Mayuresh Bakshi ,  Rohit Sharma

Inventor： Sivakumar Venkatesan ,  Mayuresh Bakshi ,  Rohit Sharma

IPC: H04L12/26 ,  G01R31/08 ,  H04L12/935 ,  H04L12/931 ,  H04L12/947

CPC classification number: H04L49/3009 ,  H04L43/0894 ,  H04L49/25 ,  H04L49/506

Abstract: A method and system for tracking and managing network flows including receiving a first flow counter value for a flow of first flows and determining that the flow is an elephant flow. The method further includes obtaining flow egress port information by determining an egress port on a switch for each of the first flows, obtaining port congestion information for the switch, where the port congestion information includes port congestion data for each egress port, and selecting, based on the port congestion information and the flow egress port information, a new egress port for the flow, and sending a request to update a switch chip on the switch, where the request specifies that subsequently received packets for the flow are to be forwarded out of the new egress port.

公开(公告)号：US08615009B1

公开(公告)日：2013-12-24

申请号：US12765636

申请日：2010-04-22

Applicant: Sankar Ramamoorthi ,  Satyadeva Konduru ,  Gregory Kotlyar ,  Satish Raghunath ,  Sivakumar Venkatesan ,  Ramakanth Gunuganti

Inventor： Sankar Ramamoorthi ,  Satyadeva Konduru ,  Gregory Kotlyar ,  Satish Raghunath ,  Sivakumar Venkatesan ,  Ramakanth Gunuganti

IPC: H04L12/28

CPC classification number: H04L45/74 ,  H04L45/306 ,  H04L45/60 ,  H04L45/745 ,  H04L69/22

Abstract: An example network device includes a network interface and a control unit that receives a packet having header information. The control unit includes a forwarding structure having a plurality of entries that each refers to one of a plurality of logical interfaces, a forwarding engine configured to access the forwarding structure to select a first logical interface to which to forward the packet based on the header information, wherein the first logical interface comprises a pseudo-device interface (PDI). The control unit also includes a PDI module that tunnels the packet to an external service complex (ESC) by at least applying a set of metadata to the packet, encapsulating the packet with a header, and forwarding the packet to the ESC via the network interface, and wherein the metadata allows the ESC to determine a set of services to be applied to the packet based on the metadata.

Abstract translation: 示例性网络设备包括网络接口和接收具有报头信息的分组的控制单元。 所述控制单元包括具有多个条目的转发结构，每个条目各自指代多个逻辑接口中的一个;转发引擎，被配置为接入所述转发结构，以基于所述报头信息选择转发所述分组的第一逻辑接口 ，其中所述第一逻辑接口包括伪设备接口（PDI）。 控制单元还包括PDI模块，其通过至少将一组元数据应用于分组来封装分组，并通过网络接口将分组封装到ESC，将分组隧道传送到外部服务综合体（ESC） ，并且其中所述元数据允许所述ESC基于所述元数据确定要应用于所述分组的一组服务。

公开(公告)号：US20160050150A1

公开(公告)日：2016-02-18

申请号：US14457660

申请日：2014-08-12

Applicant: Sivakumar Venkatesan ,  Mayuresh Bakshi ,  Rohit Sharma

Inventor： Sivakumar Venkatesan ,  Mayuresh Bakshi ,  Rohit Sharma

IPC: H04L12/803 ,  H04L12/859 ,  H04L12/933 ,  H04L12/709 ,  H04L12/26 ,  H04L12/721 ,  H04L12/931

CPC classification number: H04L49/3009 ,  H04L43/0894 ,  H04L49/25 ,  H04L49/506

Abstract: A method and system for tracking and managing network flows including receiving a first flow counter value for a flow of first flows and determining that the flow is an elephant flow. The method further includes obtaining flow egress port information by determining an egress port on a switch for each of the first flows, obtaining port congestion information for the switch, where the port congestion information includes port congestion data for each egress port, and selecting, based on the port congestion information and the flow egress port information, a new egress port for the flow, and sending a request to update a switch chip on the switch, where the request specifies that subsequently received packets for the flow are to be forwarded out of the new egress port.

Abstract translation: 一种用于跟踪和管理网络流的方法和系统，包括接收第一流的流的第一流计数值，并确定流是大象流。 该方法还包括通过为每个第一流确定交换机上的出口端口，获取交换机的端口拥塞信息，其中端口拥塞信息包括每个出口端口的端口拥塞数据，以及基于 关于端口拥塞信息和流出口端口信息，用于流的新出口端口，并且发送更新交换机上的交换芯片的请求，其中请求指定随后接收到的流的分组将被转发出 新出口端口。

公开(公告)号：US20090185560A1

公开(公告)日：2009-07-23

申请号：US12016852

申请日：2008-01-18

Applicant: Sivakumar Venkatesan ,  Ramakanth Gunuganti

Inventor： Sivakumar Venkatesan ,  Ramakanth Gunuganti

IPC: H04L12/56

CPC classification number: H04L45/60 ,  H04L45/24 ,  H04L49/90 ,  H04L49/9094

Abstract: Systems and methods consistent with the present invention provide a better fragment drop heuristic that determines a per-fragment determined “remainder time” value to trigger potential drops on the whole bundle. A per-bundle drop timeout value is assumed. This value is to be configured based on differential delay considerations of the various links that constitute the bundle. The arrival time of a fragment to a reassembly algorithm triggers a remainder timer. When the reassembly algorithm instance actually processes the fragment, the “remainder time,” which is difference of a bundle drop timeout and time elapsed on the remainder timer, is used to determine whether the fragment and the other fragments of the packet should be dropped.

Abstract translation: 与本发明一致的系统和方法提供了更好的片段丢弃启发式，其确定每片段确定的“剩余时间”值以触发整个束上的潜在丢失。 假定每个捆绑包丢弃超时值。 该值将根据组成捆绑的各种链路的差异延迟考虑进行配置。 片段到重组算法的到达时间触发余数计时器。 当重组算法实例实际处理片段时，使用余数定时器上的捆绑丢弃超时与经过的时间差的“余数时间”用于确定分组的片段和其他片段是否应该被丢弃。

公开(公告)号：US08537673B1

公开(公告)日：2013-09-17

申请号：US12105746

申请日：2008-04-18

Applicant: Sivakumar Venkatesan

Inventor： Sivakumar Venkatesan

IPC: G01R31/08

CPC classification number: H04L45/24 ,  H04L45/245

Abstract: Resource utilization required for processing packets associated with multilink bundles of a router are tracked. Corrective actions can be taken to provide fair usage of the available resources by comparing actual usage to a predetermined threshold value, and taking corrective actions if that threshold is exceeded.

Abstract translation: 跟踪处理与路由器的多链路捆绑相关联的数据包所需的资源利用率。 可以采取纠正措施，通过将实际使用情况与预定阈值进行比较来提供可用资源的合理使用，并且如果超过该阈值，则采取纠正措施。

公开(公告)号：US08233502B2

公开(公告)日：2012-07-31

申请号：US12016852

申请日：2008-01-18

Applicant: Sivakumar Venkatesan ,  Ramakanth Gunuganti

Inventor： Sivakumar Venkatesan ,  Ramakanth Gunuganti

IPC: H04L12/28 ,  H04J3/24 ,  H04J3/00 ,  H04J3/04

CPC classification number: H04L45/60 ,  H04L45/24 ,  H04L49/90 ,  H04L49/9094

Abstract: Systems and methods consistent with the present invention provide a better fragment drop heuristic that determines a per-fragment determined “remainder time” value to trigger potential drops on the whole bundle. A per-bundle drop timeout value is assumed. This value is to be configured based on differential delay considerations of the various links that constitute the bundle. The arrival time of a fragment to a reassembly algorithm triggers a remainder timer. When the reassembly algorithm instance actually processes the fragment, the “remainder time,” which is difference of a bundle drop timeout and time elapsed on the remainder timer, is used to determine whether the fragment and the other fragments of the packet should be dropped.

Abstract translation: 与本发明一致的系统和方法提供了更好的片段丢弃启发式，其确定每片段确定的“剩余时间”值以触发整个束上的潜在丢失。 假定每个捆绑包丢弃超时值。 该值将根据组成捆绑的各种链路的差异延迟考虑进行配置。 片段到重组算法的到达时间触发余数计时器。 当重组算法实例实际处理片段时，使用余数定时器上的捆绑丢弃超时与经过的时间差的“余数时间”用于确定分组的片段和其他片段是否应该被丢弃。