Cisco Technology, Inc.

Deepak Srinivas Mayya

Lalit Kumar

Amit Surendra Singh

Ravikiran Kaidala Lakshman

Rajesh L G

Tanjore K Suresh

Srividhya Nagarajan

Sagar Borikar

David Walker

Wei-Min Cheng

Ranganathan Rajagopalan

Narendra Jayawant Gathoo

Shrikant Jay Vaidya

Prashant Pathak

Saket Jain

Ranga Rajagopalan

Simon Hsu

Presented herein are techniques enable existing hardware input/output resources, such as the hardware queues (queue control registers), of a network interface card to be shared with different hosts (i.e., each queue mapped to many hosts) by logically segregating the hardware I/O resources using assignable interfaces each associated with a distinct Process Address Space Identifier (PASID). That is, different assignable interfaces are created and associated with different PASIDs, and these assignable interfaces each correspond to a different host (i.e., there is a mapping between a host, an assignable interface, a PASID, and a partition of a hardware queue). The result is that that the hosts can use the assignable interface to directly access the hardware queue partition that corresponds thereto.

Network interface card resource partitioning

The disclosed technology relates to a process for general-purpose ring buffer handling in a network controller. Specifically, general purpose ring buffers are used in software queue on both the host domain and the network controller to carry out user-defined protocols. The protocols define the relationship of the ring buffers in the host domain and the network controller domain (e.g. how data is moved and processed). The protocols minimize interruptions to the network controller. Furthermore, the implementation of ring buffers in the network controller domain also provides additional resources to the host domain for carrying out the user-defined protocols.

General purpose ring buffer handling in a network controller

Presented herein are methodologies for increasing effective throughput on a network. A method includes receiving a command request via a communication bus, the command request including a command ID, determining, based on the command ID, whether data in the command request is to be joined with data from other command requests having the same command ID, when it is determined, based on the command ID, that the data in the command request is to be joined with other data from other command requests having the same command ID, writing the data to a selected buffer in which the other data is already stored, and causing the data and the other data in the buffer to be sent as a payload of a single packet across a communications fabric.

Increasing throughput of non-volatile memory express over fabric (NVMEoF) via peripheral component interconnect express (PCIe) interface

Example embodiments may include a method for configuring an interface that includes determining information for a configuration of an interface of a first device including a plurality of SERDES slices having a plurality of connections to a second device over the interface; and configuring a back channel layer associated with the first device to form a back channel path to carry a message between a transmitter and a receiver of the first device based on the configuration of the plurality of connections to the second device. The transmitter can be in a first SERDES slice of the plurality of SERDES slices and the receiver is in a second SERDES slice of the plurality of SERDES slices.

Back channel support for systems with split lane swap

Traffic between a first device and a second device across a plurality of communication lanes of a communication link is stalled. A first device side of the communication lane is powered-down on at least one of the plurality of communication lanes to deactivate the at least one communication lane. A power-down request is sent from the first device to the second device to power-down a second device-side of the at least one communication lane. Traffic between the first device and the second device is restarted over the active communication lanes.

Dynamic communication link scaling

In one embodiment, a method is provided for alleviating congestion in a network system. In this method, the receipt of data packets destined for a destination apparatus is detected. Flow control signals are also received with each flow control signal corresponding to a data packet. Various time periods are tracked with each time period being between the detection of the receipt of a data packet and the receipt of tracked corresponding flow control signal. An average of the time periods is calculated and this average is compared to a threshold. One or more data packets are dropped in reference to the comparison.

Congestion alleviation in network systems

Various systems and methods for using credits to determine cable length are described herein. A number of available credits is determined at a transmitter, the credits being received at the transmitter from a receiver, and each of the credits being used for the transmitter to transmit a packet to the receiver. An average number of credits for a period is determined. The cable length between the transmitter and the receiver is determined based on the average number of credits for the period.

Method and apparatus for using credits to determine cable length

Techniques are provided to allow for efficient data packing under certain coding schemes, e.g., when an enhanced 64/66 encoding scheme is employed. The traditional 64/66 encoding scheme uses idle bytes to pad packets to obtain a packet of fixed size, thereby adding idle overhead to transmitted packets. Accordingly, the techniques eliminate padding bytes and provide for receiving packets by way of a first data link at forwarding device. The packets are transmitted using a scheme that pads the packets to achieve a predetermined packet length. The padding is removed from the received packets to obtain unpadded packets. Data segments of the unpadded packets are concatenated for forwarding in a continuous data stream with a packet delimiter separating each of the unpadded packets. The continuous data stream is encoded using enhanced physical coding sub layer for forwarding over a second data link.

Fabric link physical coding sublayer packing scheme

An example system and method for multicast flow reordering in a network is provided and includes receiving a data packet at an egress of a node, the data packet associated with a descriptor including at least one parameter; determining the at least one parameter of the descriptor; and processing the data packet for reordering based on the determining of the at least on parameter of the descriptor.

Multicast flow reordering scheme

A number of ports are configured in a linecard in a network device as dedicated ports and a remaining number of ports as shared ports. A total bandwidth allocated to the dedicated ports is computed. It is determined that available bandwidth at a central crossbar is greater than the total bandwidth allocated to the dedicated ports. The total data sent the central crossbar is rate limited to less than the available bandwidth at the central crossbar. First data associated with the dedicated ports is scheduled to the central crossbar using a first priority. Second data associated with the shared ports is scheduled using a second priority. A shared port data is scheduled based on a ration of a bandwidth requirement for the shared port as a fraction of overall bandwidth requirement for the shared ports.

Growing community of inventors

Fremont, CA, United States of America

Deepak Srinivas Mayya