Techniques for dynamically assigning jobs to processors in a cluster using local job tables
    31.
    发明授权
    Techniques for dynamically assigning jobs to processors in a cluster using local job tables 有权
    使用本地作业表将作业动态分配到群集中的处理器的技术

    公开(公告)号:US09396021B2

    公开(公告)日:2016-07-19

    申请号:US12336329

    申请日:2008-12-16

    IPC分类号: G06F9/46 G06F9/48 G06F9/50

    摘要: A technique for operating a high performance computing cluster includes monitoring workloads of multiple processors. The high performance computing cluster includes multiple nodes that each include two or more of the multiple processors. Workload information for the multiple processors is periodically updated in respective local job tables maintained in each of the multiple nodes. Based on the workload information in the respective local job tables, one or more threads are periodically moved to a different one of the multiple processors.

    摘要翻译: 用于操作高性能计算集群的技术包括监视多个处理器的工作负载。 高性能计算集群包括多个节点,每个节点包括多个处理器中的两个或多个。 多个处理器的工作量信息在维护在多个节点中的每个节点的相应本地作业表中周期性地更新。 基于各个本地作业表中的工作负载信息,将一个或多个线程周期性地移动到多个处理器中的不同的一个。

    Checksum verification accelerator
    33.
    发明授权
    Checksum verification accelerator 失效
    校验和验证加速器

    公开(公告)号:US08726132B2

    公开(公告)日:2014-05-13

    申请号:US13302688

    申请日:2011-11-22

    IPC分类号: H03M13/00

    摘要: Disclosed is a method and system for validating a data packet by a network processor supporting a first network protocol and a second network protocol and utilizing shared hardware. The network processor receives a data packet; identifies a network packet protocol for the data packet; and processes the data packet according to the network packet protocol comprising: updating a first register with a first partial packet length specific to the first network protocol; updating a second register with a second partial packet length specific to the second network protocol; and updating a third register with a first checksum computed from fields independent of the network protocol. The system produces a second checksum utilizing a function that combines values from the first register, the second register, and the third register. The system validates the data packet by comparing the data packet checksum to the second checksum.

    摘要翻译: 公开了一种用于通过支持第一网络协议和第二网络协议的网络处理器来验证数据分组并利用共享硬件的方法和系统。 网络处理器接收数据包; 识别数据包的网络包协议; 并根据网络分组协议对数据分组进行处理,包括:以第一网络协议特有的第一部分分组长度更新第一寄存器; 用第二网络协议特有的第二部分分组长度更新第二寄存器; 以及用独立于网络协议的字段计算的具有第一校验和的更新第三寄存器。 该系统利用组合来自第一寄存器,第二寄存器和第三寄存器的值的函数产生第二校验和。 系统通过将数据包校验和与第二个校验和进行比较来验证数据包。

    Bidirectional packet flow transformation
    34.
    发明授权
    Bidirectional packet flow transformation 失效
    双向分组流转换

    公开(公告)号:US08619782B2

    公开(公告)日:2013-12-31

    申请号:US13325624

    申请日:2011-12-14

    摘要: A network packet includes a packet key that includes one or more source-destination field pairs that each include a source field and a destination field. For each selected source-destination field pair, first and second sections are selected in the packet key. A source field value is extracted from the source field and a destination field value is extracted from the destination field. For each source bit of the source field value: a destination bit is selected from the destination field; an OR logic function is applied to the source bit and the destination bit to generate a first resulting value is stored at the same bit position as the source bit in the first section; an AND logic function is applied to the source bit and the destination bit to generate a second resulting value stored at the same bit position as the source bit in the second section.

    摘要翻译: 网络分组包括分组密钥,其包括一个或多个源 - 目的地字段对,每个源对目的字段对都包括源字段和目的地字段。 对于每个选择的源 - 目的地字段对,在分组密钥中选择第一和第二部分。 从源字段提取源字段值,并从目标字段提取目的字段值。 对于源字段值的每个源位:从目标字段中选择目标位; 将OR逻辑功能应用于源位,并将目标位产生为第一结果值,存储在与第一段中的源位相同的位位置; 将AND逻辑功能应用于源位和目标位,以产生存储在与第二部分中的源位相同的位位置处的第二个结果值。

    Network Data Processsing System
    35.
    发明申请
    Network Data Processsing System 有权
    网络数据处理系统

    公开(公告)号:US20130132535A1

    公开(公告)日:2013-05-23

    申请号:US13299045

    申请日:2011-11-17

    IPC分类号: G06F15/177

    摘要: A method and apparatus for processing of data packets by a data processing component comprising a plurality of processing resources. A metric value for a current configuration of the processing resources that are processing the data packets is identified. A new configuration of the processing resources is selected using the metric value. The current configuration of the processing resources is changed to the new configuration and data packets are distributed to the new configuration for processing as the data packets are received.

    摘要翻译: 一种由包括多个处理资源的数据处理组件处理数据分组的方法和装置。 识别处理数据分组的处理资源的当前配置的度量值。 使用度量值选择处理资源的新配置。 处理资源的当前配置被改变为新的配置,并且数据分组被分配到新配置,以便在接收到数据分组时进行处理。

    Dual scheduling of work from multiple sources to multiple sinks using source and sink attributes to achieve fairness and processing efficiency
    36.
    发明授权
    Dual scheduling of work from multiple sources to multiple sinks using source and sink attributes to achieve fairness and processing efficiency 失效
    使用源和宿属性从多个源到多个宿的双重调度工作,以实现公平性和处理效率

    公开(公告)号:US08295305B2

    公开(公告)日:2012-10-23

    申请号:US12650174

    申请日:2009-12-30

    IPC分类号: H04J3/22

    CPC分类号: H04L47/522 H04L47/6215

    摘要: Assigning work, such as data packets, from a plurality of sources, such as data queues in a network processing device, to a plurality of sinks, such as processor threads in the network processing device is provided. In a given processing period, a source is selected in a manner that maintains fairness in the selection process. A corresponding sink is selected for the selected source based on processing efficiency. If, due to assignment constraints, no sink is available for the selected source, the selected source is retained for selection in the next scheduling period, to maintain fairness. In this case, to optimize efficiency, a most efficient currently available sink is identified and a source for providing work to that sink is selected.

    摘要翻译: 提供了诸如数据分组的工作,诸如诸如网络处理设备中的数据队列的多个源到网络处理设备中的诸如处理器线程的多个接收器。 在给定的处理期间,以选择过程中保持公平的方式选择源。 基于处理效率为所选择的源选择相应的接收器。 如果由于分配限制,所选择的源没有可用的接收器,所选择的源被保留用于在下一个调度周期中进行选择,以保持公平性。 在这种情况下,为了优化效率,识别出最有效的当前可用的接收器,并且选择用于向该接收器提供工作的源。

    SELECTION OF RECEIVE-QUEUE BASED ON PACKET ATTRIBUTES
    37.
    发明申请
    SELECTION OF RECEIVE-QUEUE BASED ON PACKET ATTRIBUTES 失效
    基于分组属性选择接收队列

    公开(公告)号:US20120218885A1

    公开(公告)日:2012-08-30

    申请号:US13466914

    申请日:2012-05-08

    IPC分类号: H04L12/24

    CPC分类号: H04L47/6215

    摘要: According to embodiments of the invention, there is provided a method for operating a network processor. The network processor receiving a first data packet in a stream of data packets and a set of receive-queues adapted to store receive data packets. The network processor processing the first data packet by reading a flow identification in the first data packet; determining a quality of service for the first data packet; mapping the flow identification and the quality of service into an index for selecting a first receive-queue for routing the first data packet; and utilizing the index to route the first data packet to the first receive-queue.

    摘要翻译: 根据本发明的实施例,提供了一种用于操作网络处理器的方法。 网络处理器接收数据包流中的第一数据包和适于存储接收数据包的一组接收队列。 网络处理器通过读取第一数据分组中的流标识来处理第一数据分组; 确定第一数据分组的服务质量; 将流标识和服务质量映射到用于选择用于路由第一数据分组的第一接收队列的索引; 以及利用所述索引将所述第一数据分组路由到所述第一接收队列。

    Merging Result from a Parser in a Network Processor with Result from an External Coprocessor
    38.
    发明申请
    Merging Result from a Parser in a Network Processor with Result from an External Coprocessor 失效
    从具有外部协处理器结果的网络处理器中的解析器合并结果

    公开(公告)号:US20120204190A1

    公开(公告)日:2012-08-09

    申请号:US13365778

    申请日:2012-02-03

    IPC分类号: G06F9/46

    CPC分类号: G06F9/546 G06F9/544

    摘要: A mechanism is provided for merging in a network processor results from a parser and results from an external coprocessor providing processing support requested by said parser. The mechanism enqueues in a result queue both parser results needing to be merged with a coprocessor result and parser results which have no need to be merged with a coprocessor result. An additional queue is used to enqueue the addresses of the result queue where the parser results are stored. The result from the coprocessor is received in a simple response register. The coprocessor result is read by the result queue management logic from the response register and merged to the corresponding incomplete parser result read in the result queue at the address enqueued in the additional queue.

    摘要翻译: 提供了一种机制,用于在来自解析器的网络处理器结果和来自提供由所述解析器请求的处理支持的外部协处理器的结果中合并。 结果队列中的机制排队,解析器结果需要与协处理器结果合并,并且不需要与协处理器结果合并的解析器结果。 使用一个附加队列来排列存储解析器结果的结果队列的地址。 协处理器的结果是在简单的响应寄存器中接收的。 协处理器结果由响应寄存器的结果队列管理逻辑读取,并被合并到在附加队列中排队的地址的结果队列中读取的相应的不完整解析器结果。

    Host Ethernet Adapter for Handling Both Endpoint and Network Node Communications
    39.
    发明申请
    Host Ethernet Adapter for Handling Both Endpoint and Network Node Communications 失效
    用于处理端点和网络节点通信的主机以太网适配器

    公开(公告)号:US20120192190A1

    公开(公告)日:2012-07-26

    申请号:US13011663

    申请日:2011-01-21

    IPC分类号: G06F9/46

    CPC分类号: G06F15/1735

    摘要: A host Ethernet adapter (HEA) and method of managing network communications is provided. The HEA includes a host interface configured for communication with a multi-core processor over a processor bus. The host interface comprises a receive processing element including a receive processor, a receive buffer and a scheduler for dispatching packets from the receive buffer to the receive processor; a send processing element including a send processor and a send buffer; and a completion queue scheduler (CQS) for dispatching completion queue elements (CQE) from the head of the completion queue (CQ) to threads of the multi-core processor in a network node mode. The method comprises operatively coupling an Ethernet adapter to a multi-core processor system via a processor bus, selectively assigning a first plurality of packets to a first queue pair for servicing in an endpoint mode, running a device driver on the multi-core processing system, the device driver controlling the servicing of the first queue pair by dispatching the first plurality of packets to only one processor core of the multi-core processor system, selectively assigning a second plurality of packets to a second queue pair for servicing in a network node mode; and the Ethernet adapter controlling the servicing of the second queue pair by dispatching the second plurality of packets to multiple processor threads.

    摘要翻译: 提供主机以太网适配器(HEA)和管理网络通信的方法。 HEA包括被配置为通过处理器总线与多核处理器进行通信的主机接口。 所述主机接口包括接收处理元件,所述接收处理元件包括接收处理器,接收缓冲器和用于从所述接收缓冲器向所述接收处理器分发分组的调度器; 包括发送处理器和发送缓冲器的发送处理元件; 以及用于从完成队列(CQ)的头部将网络节点模式中的多核处理器的线程调度完成队列元素(CQE)的完成队列调度器(CQS)。 该方法包括经由处理器总线可操作地将以太网适配器耦合到多核处理器系统,选择性地将第一多个分组分配到第一队列对以在端点模式下进行服务,在多核处理系统上运行设备驱动程序 所述设备驱动程序通过将所述第一多个分组分派到所述多核处理器系统的一个处理器核心来控制所述第一队列对的服务,选择性地将第二多个分组分配给第二队列对以在网络节点中进行服务 模式; 以及所述以太网适配器通过将所述第二多个分组分派到多个处理器线程来控制所述第二队列对的服务。

    SELECTION OF RECEIVE-QUEUE BASED ON PACKET ATTRIBUTES
    40.
    发明申请
    SELECTION OF RECEIVE-QUEUE BASED ON PACKET ATTRIBUTES 审中-公开
    基于分组属性选择接收队列

    公开(公告)号:US20120155267A1

    公开(公告)日:2012-06-21

    申请号:US13302285

    申请日:2011-11-22

    IPC分类号: H04L12/26

    CPC分类号: H04L47/6215

    摘要: According to embodiments of the invention, there is provided a method, a system, and a computer program product for operating a network processor. The network processor processing a received data packet by reading a flow identification in the data packet; determining a quality of service criteria (QoSC) for the data packet; mapping the flow identification and the QoSC into an index for selecting a receive-queue for routing the data packet; and utilizing the index to route the data packet to the receive-queue.

    摘要翻译: 根据本发明的实施例,提供了一种用于操作网络处理器的方法,系统和计算机程序产品。 网络处理器通过读取数据分组中的流标识来处理接收到的数据分组; 确定数据包的服务质量标准(QoSC); 将流标识和QoSC映射到用于选择用于路由数据分组的接收队列的索引; 并利用索引将数据包路由到接收队列。