International Business Machines Corporation

Lenovo (singapore) Pte. Ltd.

Nandakumar Peethambaram

Keshav G Kamble

Vijoy A Pandey

Dar-Ren Leu

Sushma Anantharam

Nirapada Ghosh

Chandarani J Mendon

Dayavanti G Kamath

Nilanjan Mukherjee

Stephan Benny

Alexander Philip Campbell

Abhijit Prabhakar Kumbhare

Simarbir Singh

Kumbhare P Abhijit

Nilanjan Mukharjee

A distributed device architecture includes a master device and one or more member devices. A simple network management protocol (SNMP) agent of a master device receives an SNMP request from a managing device. Where the SNMP request pertains to a given member device, and where the SNMP request requires involvement of the given member device to fulfill the SNMP request, the master device generates a non-SNMP request corresponding to the SNMP request and transmits the non-SNMP request to the given member device. A non-SNMP agent of the given member device processes the non-SNMP request and transmits processing results back to the master device. The master device generates an SNMP response corresponding to the processing results, and the SNMP agent of the master device transmits the SNMP response back to the managing device.

SNMP request processing within distributed device architecture

A process of diagnosing and debugging a network element remotely may employ sending of performance data to a third party evaluator for processing of the data. A mobile device may interface with the network element and request evaluation of the performance by the third party evaluator. Large memory loads may be sent to the third party evaluator where intensive processing may be performed. The results of the processing by the third party evaluator may be sent to the mobile device for viewing by a user.

Network element diagnostic evaluation

A distributed system includes first-tier entities, and a master entity in communication with each first-tier entity. The master entity provides a single access point through which an administrator can submit commands to manage all entities. The master entity maintains a table of virtual slots. Each virtual slot points to one of the first-tier entities, and each first-tier entity is pointed to by at least one virtual slot. The processor runs an RPC (remote procedure call) client to submit RPC requests to the first-tier entities, and determines a destination first-tier entity for a given RPC request in response to which virtual slot the administrator submits a command. The distributed system can include second-tier entities, each indirectly communicating with the master entity through a first-tier entity. The table has a virtual slot for each second-tier entity, which points to the first-tier entity acting as proxy for the second-tier entity.

Remote procedure call for a distributed system

Management of a network may be provided by moving management functions from each switch in a network cluster into a locally controlled server. The management functions may be organized into an Ethernet distributed fabric switching protocol (E-DFP) based management plane. Control functions of each switch may also be organized into a control plane. The management plane and control plane of the network may be moved from each switch and integrated into the server. Efficiency in network switching may be provided by releasing the switches of their management and control responsibilities and centralizing these responsibilities to the server. In some embodiments, the server may be outside the cluster.

Management of distributed network switching cluster

A distributed fabric system comprises a plurality of independent network elements interconnected by inter-switch links and assigned to a same group. Each network element includes a switching chip, a processor, and memory storing program code that is executed by the processor. The program code of each network element includes a device configuration (DC) stacking module and a switch discovery protocol (SDP) module. The SDP module of each network element, when executed, discovers each other network element in the group and elects one of the network elements as a master network element. The SDP module of the master network element, when executed, sends messages to the DC-stacking module of the master network element. Each sent message identifies one of the network elements in the group. The DC stacking module of the master network element, when executed, maintains a record of all network elements that are currently members in the group.

Distributed fabric management protocol

A distributed fabric system has distributed line card (DLC) chassis and scaled-out fabric coupler (SFC) chassis. Each DLC chassis includes a network processor and fabric ports. Each network processor of each DLC chassis includes a fabric interface in communication with the DLC fabric ports of that DLC chassis. Each SFC chassis includes a fabric element and fabric ports. A communication link connects each SFC fabric port to one DLC fabric port. Each communication link includes cell-carrying lanes. Each fabric element of each SFC chassis collects per-lane statistics for each SFC fabric port of that SFC chassis. Each SFC chassis includes program code that obtains the per-lane statistics collected by the fabric element chip of that SFC chassis. A network element includes program code that gathers the per-lane statistics collected by each fabric element of each SFC chassis and integrates the statistics into a topology of the entire distributed fabric system.

Diagnostics in a distributed fabric system

A distributed fabric system has distributed line card (DLC) chassis and scaled-out fabric coupler (SFC) chassis. Each DLC includes a network processor and fabric ports. Each network processor of each DLC includes a fabric interface in communication with the fabric ports of that DLC. Each SFC includes at least one fabric element and SFC fabric ports. A fabric communication link connects each SFC fabric port to one DLC fabric port. Each fabric communication link includes cell-carrying lanes. Each fabric element of each SFC detects connectivity between each SFC fabric port of that SFC and one DLC fabric port over a fabric communication link. Each SFC includes program code that reads connectivity matrix from fabric element chips and sends connection information corresponding to the detected connectivity from that SFC to a central agent. A network element includes the central agent, which, when executed, constructs a topology of the distributed fabric system from the connection information sent from each SFC.

Management of a distributed fabric system

A switch includes network ports and a network processor with a fabric interface that provides SerDes (Serializer/Deserializer) channels. The network processor divides each packet received over the network ports into cells and distributes the cells across the SerDes channels. Fabric ports of the switch communicate with the fabric interface to transmit cells to and receive cells from the fabric interface. The switch is selectively configurable as a standalone switch by connecting each fabric port of the switch to another of the fabric ports of the switch, as a member of a switch stack by connecting each fabric port of the switch to a different other switch through one fabric port of that other switch, or as a member of a distributed fabric system by connecting each fabric port of the switch to a different scaled-out fabric coupler (SFC) chassis by an SFC fabric port of that SFC chassis.

Multi-role distributed line card

A network includes a first switching domain having a distributed fabric comprised of interconnected standalone switches. The standalone switches communicate with each other in accordance with a packet-based distributed fabric protocol. A second switching domain has a plurality of cell-based switches in communication with a cell-based switch fabric. The cell-based switches communicate with each other through the cell-based switch fabric in accordance with a cell-based distributed fabric protocol. One of the cell-based switches is coupled by a communication link to one of the standalone switches of the first switching domain. The second switching domain includes a server device coupled to one of the cell-based switches. The server device is configured with logic to process control packets for the standalone switches in accordance with the packet-based distributed fabric protocol and control packets for the cell-based switches in accordance with a protocol that is different from the packet-based distributed fabric protocol.

Centralized control and management planes for different independent switching domains

A distributed fabric system has distributed line card (DLC) chassis and scaled-out fabric coupler (SFC) chassis. Each DLC includes a network processor and fabric ports. Each network processor includes a fabric interface in communication with the fabric ports of that DLC. Each SFC includes at least one fabric element and SFC fabric ports. A fabric communication link connects each SFC fabric port to one DLC fabric port. Each fabric communication link includes cell-carrying lanes. Each fabric element detects connectivity between each SFC fabric port of that SFC and one DLC fabric port over a fabric communication link. Each SFC reads a connectivity matrix from fabric element chips and sends connection information corresponding to the detected connectivity from that SFC to a central agent. A network element includes the central agent, which, when executed, constructs a topology of the distributed fabric system from the connection information sent from each SFC.

A communication protocol in a layer two (L2) network switch comprises, in response to a service request by a source node, registering the source node for packet communication service. The protocol further comprises forwarding one or more packets from the registered source node to one or more destination nodes. The protocol further comprises receiving packets from one or more destination nodes and forwarding each received packet to a corresponding registered node.

Growing community of inventors

Santa Clara, CA, United States of America

Nandakumar Peethambaram