公开(公告)号：US09553734B2

公开(公告)日：2017-01-24

申请号：US14151610

申请日：2014-01-09

Applicant: DELL PRODUCTS L.P.

Inventor： Ramasubramani Mahadevan ,  Pathangi N. Janardhanan

IPC: H04L12/18

CPC classification number: H04L45/16 ,  H04L12/185 ,  H04L12/1886 ,  H04L45/74

Abstract: A switch (110R.3) uses a forwarding database (140F) to forward multicast packets. The switch participates in a multicast protocol (e.g. PIM) with other switches, and also uses a snooping database (e.g. for IGMP snooping) to learn about local group members, i.e. members for which the router serves as a local multicast router (e.g. the local members do not participate in the multicast protocol). When the switch learns of a local member of a multicast group, the switch updates the snooping database but may or may not install the snooping information in the forwarding database. Thus, the forwarding database is not necessarily provisioned for locally sourced data packets addressed to the group until receipt of a locally sourced data packet addressed to the group. If no such packets are received, the forwarding database is not provisioned for such packets, and therefore its size is reduced. Other features are also provided.

Abstract translation: 交换机（110R.3）使用转发数据库（140F）转发多播分组。 交换机与其他​​交换机参与组播协议（例如PIM），并且还使用侦听数据库（例如，用于IGMP Snooping）来了解本地组成员，即路由器作为本地组播路由器的成员（例如本地 成员不参与组播协议）。 交换机学习组播组的本地成员时，交换机会更新Snooping数据库，但可能会将Snooping信息安装在转发数据库中。 因此，转发数据库不一定被提供给寻址到组的本地源数据分组，直到接收到寻址到组的本地来源的数据分组。 如果没有接收到这样的数据包，转发数据库不会为这些数据包提供，因此它的大小减少了。 还提供其他功能。

公开(公告)号：US20160301608A1

公开(公告)日：2016-10-13

申请号：US14685369

申请日：2015-04-13

Applicant: DELL PRODUCTS L.P.

Inventor： Avinash Natarajan ,  Ramasubramani Mahadevan

IPC: H04L12/743 ,  H04L12/46 ,  H04L12/709 ,  H04L12/741

CPC classification number: H04L45/7453 ,  H04L12/2865 ,  H04L12/4641 ,  H04L45/24 ,  H04L45/245 ,  H04L45/54 ,  Y02D30/30 ,  Y02D50/30

Abstract: Various embodiments of the invention allow for rapid communication in virtual link trunking (VLT) networks in which network traffic flows over not all-homed VLT peer devices, while honoring Equal Cost Multi Path (ECMP) decisions and normal route decisions about next hops. Traffic flow is made deterministic and free of sub-optimal paths that otherwise cause unnecessary traffic over inter-node links in the VLT domain. In embodiments, this is accomplished by using receiving VLAN interface-IP addresses from VLT devices in order to create and use a sub-LAG egress table from which sets of ports that lead to intended VLT devices are derived. In embodiments, instead of a VLAN interface-IP addresses a routing MAC address is used when forming the sub-LAG.

Abstract translation: 本发明的各种实施例允许在虚拟链路中继（VLT）网络中的快速通信，其中网络业务流过并非全部归属的VLT对等设备，同时遵守等成本多路径（ECMP）决策和关于下一跳的正常路由决定。 业务流被确定并且没有子优化路径，否则会导致在VLT域中的节点间链路上的不必要的流量。 在实施例中，这是通过使用来自VLT设备的接收VLAN接口-IP地址来实现的，以便创建并使用从其导出到目的VLT设备的端口组的子LAG出口表。 在实施例中，代替VLAN接口-IP地址，当形成子LAG时使用路由MAC地址。

公开(公告)号：US20160162429A1

公开(公告)日：2016-06-09

申请号：US14565248

申请日：2014-12-09

Applicant: DELL PRODUCTS L.P.

Inventor： Karthik Krishnamurthy ,  Ramasubramani Mahadevan

IPC: G06F13/40

CPC classification number: G06F13/4045 ,  G06F13/4022 ,  G06F13/4068 ,  H04L45/16 ,  H04L45/245 ,  H04L45/44 ,  H04L49/201 ,  H04L49/70

Abstract: Aspects of the present invention include a port extender environment using the port extenders to dynamically select a data path. In embodiments of the present invention, each port extender can communicate data traffic to another port extender or to a host receiver. The communication path is selected in the port extender using a hashing system.

Abstract translation: 本发明的方面包括使用端口扩展器来动态地选择数据路径的端口扩展器环境。 在本发明的实施例中，每个端口扩展器可以将数据业务传送到另一个端口扩展器或主机接收器。 使用散列系统在端口扩展器中选择通信路径。

公开(公告)号：US20160043935A1

公开(公告)日：2016-02-11

申请号：US14920688

申请日：2015-10-22

Applicant: DELL PRODUCTS L.P.

Inventor： Ramasubramani Mahadevan ,  Pathangi N. Janardhanan

IPC: H04L12/761 ,  H04L12/18 ,  H04L12/709

CPC classification number: H04L45/16 ,  H04L12/1886 ,  H04L45/245 ,  H04L45/54 ,  Y02D50/30

Abstract: Each router in a group of routers (110R.2, 110R.3) includes an interface (P4) assigned a common virtual address for multicast (VAM). The common VAM is advertised in router advertisings for multicast traffic; other addresses are advertised for unicast. The member routers in the group share multicast forwarding databases. Increased throughput is therefore provided for multicast traffic without traffic replication. Any router in the group can service requests from end-point sources and receivers of multicast traffic, e.g. requests to join or leave a multicast group, or requests presented as multicast packets for forwarding to a multicast group.

Abstract translation: 一组路由器（110R.2,110R.3）中的每个路由器包括分配了多播（VAM）的公共虚拟地址的接口（P4）。 通用的VAM在路由器广播中被广播用于组播流量; 其他地址被广播为单播。 组中的成员路由器共享组播转发数据库。 因此，对于没有流量复制的组播流量，提供了增加的吞吐量。 组中的任何路由器可以服务于来自端点源和多播业务的接收机的请求，例如， 请求加入或离开组播组，或者作为组播数据包呈现的请求转发给组播组。

公开(公告)号：US09154408B2

公开(公告)日：2015-10-06

申请号：US13777301

申请日：2013-02-26

Applicant: Dell Products L.P.

Inventor： Ramasubramani Mahadevan ,  Pathangi Narasimhan Janardhanan ,  Avinash Natarajan

IPC: H04L12/26 ,  H04L12/703 ,  H04L29/14 ,  H04L12/707 ,  H04L12/709

CPC classification number: H04L45/28 ,  H04L45/22 ,  H04L45/245 ,  H04L49/00 ,  H04L69/40

Abstract: A system and method for traffic polarization during failures includes a communication network cluster including a first network switching unit configured to communicate with a first network node via a first network link and a second network node via a second network link, a second network switching unit configured to communicate with the first network node via a third network link and the second network node via a fourth network link, and a first intra-cluster link coupling the first network switching unit to the second network switching unit. The second network switching unit is further configured to respond to a failure in the fourth network link by broadcasting a route withdrawal message to the first network switching unit and the first network node. The first network switching unit is configured to accept rerouted network traffic from the first network node via the first network link resulting from the route withdrawal message.

Abstract translation: 用于故障期间的交通偏振的系统和方法包括：通信网络集群，包括：第一网络交换单元，被配置为经由第一网络链路与第一网络节点通信，并且经由第二网络链路与第二网络节点进行通信;第二网络交换单元， 经由第三网络链路和第二网络节点经由第四网络链路与第一网络节点通信，以及将第一网络交换单元耦合到第二网络交换单元的第一集群内链路。 第二网络交换单元还被配置为通过向第一网络交换单元和第一网络节点广播路由提取消息来响应第四网络链路中的故障。 第一网络交换单元被配置为经由由路由提取消息产生的第一网络链接接受来自第一网络节点的重新路由的网络业务。

公开(公告)号：US20150195093A1

公开(公告)日：2015-07-09

申请号：US14151610

申请日：2014-01-09

Applicant: DELL PRODUCTS L.P.

Inventor： Ramasubramani Mahadevan ,  Pathangi N. Janardhanan

IPC: H04L12/18

CPC classification number: H04L45/16 ,  H04L12/185 ,  H04L12/1886 ,  H04L45/74

Abstract: A switch (110R.3) uses a forwarding database (140F) to forward multicast packets. The switch participates in a multicast protocol (e.g. PIM) with other switches, and also uses a snooping database (e.g. for IGMP snooping) to learn about local group members, i.e. members for which the router serves as a local multicast router (e.g. the local members do not participate in the multicast protocol). When the switch learns of a local member of a multicast group, the switch updates the snooping database but may or may not install the snooping information in the forwarding database. Thus, the forwarding database is not necessarily provisioned for locally sourced data packets addressed to the group until receipt of a locally sourced data packet addressed to the group. If no such packets are received, the forwarding database is not provisioned for such packets, and therefore its size is reduced. Other features are also provided.

Abstract translation: 交换机（110R.3）使用转发数据库（140F）转发多播分组。 交换机与其他​​交换机参与组播协议（例如PIM），并且还使用侦听数据库（例如，用于IGMP Snooping）来了解本地组成员，即路由器作为本地组播路由器的成员（例如本地 成员不参与组播协议）。 交换机学习组播组的本地成员时，交换机会更新Snooping数据库，但可能会将Snooping信息安装在转发数据库中。 因此，转发数据库不一定被提供给寻址到组的本地源数据分组，直到接收到寻址到组的本地来源的数据分组。 如果没有接收到这样的数据包，转发数据库不会为这些数据包提供，因此它的大小减少了。 还提供其他功能。

公开(公告)号：US20140219081A1

公开(公告)日：2014-08-07

申请号：US13758884

申请日：2013-02-04

Applicant: DELL PRODUCTS L.P.

Inventor： Avinash Natarajan ,  Ramasubramani Mahadevan

IPC: H04L12/24

CPC classification number: H04L41/0659 ,  H04L12/44 ,  H04L12/4641 ,  H04L12/56 ,  H04L45/245 ,  H04L45/28 ,  H04L45/66 ,  Y02D30/30

Abstract: An information handling system is provided. The information handling system includes a first network switch configurable, the first network switch having a computer processor in communication with a plurality of ports for receiving and sending frames and a memory coupled to the computer processor and including a media access control (MAC) address table. The computer processor is configured to read frames, each frame having a header, and is also configured to associate a MAC address with a parent link aggregation group (LAG) except while a downstream link failure is detected. The computer processor is configured to associate the MAC address with a sub-LAG while a downstream link failure is detected. Methods for directing traffic through a virtual link trunking (VLT) domain during failure of a downstream link are also provided.

Abstract translation: 提供信息处理系统。 所述信息处理系统包括可配置的第一网络交换机，所述第一网络交换机具有与用于接收和发送帧的多个端口通信的计算机处理器和耦合到所述计算机处理器的存储器，并且包括媒体访问控制（MAC）地址表 。 计算机处理器被配置为读取帧，每帧具有报头，并且还被配置为除了检测到下游链路故障之外，将MAC地址与父链路聚合组（LAG）相关联。 计算机处理器被配置为在检测到下游链路故障时将MAC地址与子LAG相关联。 还提供了在下游链路故障期间通过虚拟链路中继（VLT）域引导业务的方法。

公开(公告)号：US11469923B2

公开(公告)日：2022-10-11

申请号：US16882423

申请日：2020-05-22

Applicant: DELL PRODUCTS L.P.

Inventor： Damodharan Sreenivasagaperumal ,  Ramasubramani Mahadevan

IPC: H04L12/46 ,  H04L45/74 ,  H04L45/00 ,  H04L45/16

Abstract: A set of remote Virtual Extensible LAN (VxLAN) tunnel endpoints (VTEPs) and an ingress VTEP associated different Ethernet Segments (ESs) elect amongst themselves designated forwarder (DF) for forwarding broadcast, unknown-unicast, and multicast traffic (BUM) traffic by triggering an RFC 7432 election mechanism on each of the VTEPs. In embodiments, DF election involves exchanging configuration information, such as Type-4 routes for ESs via Border Gateway Protocol (BGP), without being confined to a particular ES that is local to all VTEPs, i.e., irrespective of local ES and internet identifiers. This allows performing targeted forwarding of BUM traffic to intended VTEPs which avoiding unnecessary ingress replication of BUM traffic in the ingress VTEP, thereby, saving hardware buffer resources and avoiding unnecessary flooding of frames to a set of non-forwarding egress VTEPs, ultimately, reducing the load on the egress VTEP and freeing up packet processing resources.

公开(公告)号：US20210135903A1

公开(公告)日：2021-05-06

申请号：US16669985

申请日：2019-10-31

Applicant: Dell Products L.P.

Inventor： Avinash Natarajan ,  Damodharan Sreenivasagaperumal ,  Ramasubramani Mahadevan ,  Karthik Krishnamurthy

IPC: H04L12/46 ,  H04L12/751 ,  H04L12/715 ,  H04L12/761 ,  H04L12/947

Abstract: A VXLAN multi-tenant inter-networking device packet forwarding system includes a first aggregated networking device coupled to a first host device and a second aggregated networking device that is coupled to second host devices. The first aggregated networking device receives a data packet from the first host device and, in response, identifies a virtual network associated with the first host device. Based on a first and second portion of a virtual network identifier that identifies the virtual network, the first aggregated networking device generates respective first and second packet forwarding identifiers. The first aggregated networking device then provides the first and second packet forwarding identifiers in the data packet, and forwards the data packet to the second aggregated networking device. The second aggregated networking device may then forward the data packet to one of the second host devices based on the first and second packet forwarding identifiers in the data packet.

公开(公告)号：US09590889B2

公开(公告)日：2017-03-07

申请号：US14509918

申请日：2014-10-08

Applicant: DELL PRODUCTS L.P.

Inventor： Ramasubramani Mahadevan ,  Shivakumar Sundaram

IPC: H04L12/761 ,  H04L12/705 ,  H04L12/721 ,  H04L12/18

CPC classification number: H04L45/12 ,  H04L12/18 ,  H04L12/185 ,  H04L12/1877 ,  H04L45/16 ,  H04L45/18

Abstract: A method of routing traffic for multi-cast routing through a node of a network that utilizes loop-free alternative paths is presented. The method includes receiving a join in the node from a second node in the network on a shortest path between the node and the second node; retrieving loop-free alternative paths to the second node; and adding an entry to a multi-cast routing table at the node that is based on a group that includes the shortest path to the second node and the loop-free alternative paths to the second node as cost equivalent paths.

Abstract translation: 提出了一种通过利用无环路替代路径的网络节点路由多播路由流量的方法。 该方法包括：在节点和第二节点之间的最短路径上从网络中的第二节点接收节点中的连接; 检索到第二节点的无循环替代路径; 并且在基于包括到第二节点的最短路径的组的节点和到第二节点的无循环替代路径的节点处将多个入口路由表添加到成本等效路径中。