Ciena Corporation

Emc Corporation

Data General Corporation

Plexxi Inc.

Equipe Communications Corporation

Peter B Everdell

Brian Branscomb

Robert C Solomon

Brian K Bailey

David J Husak

James W Espy

Richard Barry

Scott J Bleiweiss

Derek E Spock

Nicholas A Langrind

Ruoding Li

Chris R Noel

Corey Simons

Elizabeth H Reeves

Jeffrey S Kimmel

Colin Bell

Larry B Manor

Matthew William Morgan

Roland T Provencher

Daniel I Westerberg

Robert Yates

Data center network architectures, systems, and methods that can reduce the cost and complexity of data center networks. Such data center network architectures, systems, and methods employ physical optical ring network and multi-dimensional network topologies and optical nodes to efficiently allocate bandwidth within the data center networks, while reducing the physical interconnectivity requirements of the data center networks. The respective optical nodes can be configured to provide various switching topologies, including, but not limited to, chordal ring switching topologies and multi-dimensional chordal ring switching topologies.

Optical architecture and channel plan employing multi-fiber configurations for data center network switching

Data center network architectures, systems, and methods that can reduce the cost and complexity of data center networks. Such data center network architectures, systems, and methods employ physical optical ring network topologies, optical nodes, and optical junction nodes to efficiently allocate bandwidth within the data center networks, while reducing the overall physical interconnectivity requirements of the data center networks.

Optical junction nodes for use in data center networks

The present invention provides a method and apparatus for improving transmission of control information within a network device and between multiple connected network devices. Specifically, a control path is included within a network device that is independent of the data path and dedicates control path resources to each distributed processor within the network device. Dedicating resources insures that each processor has sufficient bandwidth on the control plane to transmit control information at high frequencies. This may prevent starvation of data transmissions during periods of high control information transfers and may also reduce the likelihood or further spreading of control information storms when one or more network devices in a network experiences a failure.

Network device including dedicated resources control plane

The present invention reduces the cost of a minimally configured network device by providing a network device with a distributed switch fabric. Such a network device locates a portion of the switch fabric functionality on each forwarding card allowing the minimal network device configuration to include less than the entire switch fabric functionality. The cost of the minimal configuration is, therefore, reduced allowing network service providers to more quickly recover the initial cost of the network device. As new services are requested, additional functionality, including both forwarding cards and universal port cards may be added to the network device to handle the new requests, and the fees for the new services may be applied to the cost of the additional functionality. Consequently, the cost of the network device more closely tracks the service fees received by network providers.

Network device including central and distributed switch fabric subsystems

The present invention provides a network device that fully utilizes the available space in a standard telco rack through the use of multiple mid-planes. Full utilization of available space allows a high switching capacity network device including both physical layer switch/router subsystems and upper layer switch/router subsystems to be fit in one telco rack. Inter-mid-plane connections may be provided by connecting switch fabric cards and/or control processor cards to each of the mid-planes. Providing a multi-layer network device in one telco rack allows for intelligent layer 1 switching (for example, dynamic network connection set up), allows for one network management system to control both layer 1 and upper layer networks and eliminates grooming fees. Compared with separate layer 1 and upper layer network devices or a multi-layer network device occupying multiple telco racks, a single network device saves valuable telco site space and reduces expenses by sharing overhead such as the chassis, power and cooling.

Network device including multiple mid-planes

The present invention provides a network device, such as a network switch or a router, having a high degree of modularity and reliability. The network device includes a data plane and a control plane. The data plane relays datagrams between a pair of receive and transmit network interface ports. The control plane runs management and control operations, such as routing and policing algorithms which provide the data plane with instructions on how to relay cell/packets/frames. Further, the control plane includes an internal control device that is primarily responsible for managing the internal resources of the network device, and a separate external control device that is primarily responsible for operations relating to the interfacing of the network device with an external environment.

Functional separation of internal and external controls in network devices

A high performance data path for performing XOR on the fly. A first memory is connected to a first bus and a second memory is connected to a second bus selectively coupled to the first bus. Logic for performing an XOR can be switched into connection with the first and second bus for XORing data in a buffer with the data passed from one of the memories to the other memory. The result is replaced into the buffer to permit successive XORing. When reading from an interrelated group of disks such as a RAID 3 group, the data path permits an N-1 and go mode in which a read does not wait for data from the last disk to retrieve its data sector. If the last disk contains data (as opposed to parity) the data is obtained from the XORed data in the XOR buffer of the high performance data path. For writing data, the XOR on the fly generates the parity sector for writing at the completion of a write to an interrelated group of disks.

High performance data path with XOR on the fly

A recovery method for use in a data storage system in which a plurality of data storage devices are connected to each of two communication paths. A switch may route requests to either of the two communication paths. With one data storage controller in communication with the data storage devices over a first path and a second data storage controller in communication with the data storage devices over a second path, upon detection of a malfunction on the first path, a data storage controller may request access to the other path. Thereafter, the controller may be switched into connection with the remaining operable path so as to share that path with the other controller.

Fail-over switching system

A recovery method and fail-over switch for use in a data storage system in which a plurality of data storage devices are connected to each of two communication paths. The switch may route requests to either of the two communication paths. Switching may be accomplished by two fail-over switches connected to each other and each in connection with one of the two communication paths. With one data storage controller in communication with the data storage devices over a first path and a second data storage controller in communication with the data storage devices over a second path, the fail-over switches may be used upon detection of a malfunction on one path to switch a controller into connection with the remaining operable path so as to share that path with the other controller.

A multiprocessor cache control uses ping-pong bits to reduces the number of invalidate cycles on a shared system bus in a multiprocessor system with a plurality of caches when data is being updated by multiple CPUs. The ping-pong bit is used predict when sharing is taking place and convert read-shared requests into read-exclusive requests.

Growing community of inventors

Littleton, MA, United States of America

Peter B Everdell