公开(公告)号：US10069947B2

公开(公告)日：2018-09-04

申请号：US15222753

申请日：2016-07-28

Applicant: Huawei Technologies Co., Ltd.

Inventor： Yongqiang Yang ,  Fen Dai ,  Liufei Wen

IPC: G06F15/16 ,  H04L29/06 ,  H04L29/08

Abstract: A method and an apparatus for processing a data packet based on parallel protocol stack instances, where lower-layer protocol processing is performed, using a first protocol stack instance. An associated second protocol stack instance is determined using a target socket after the target socket that is needed to perform upper-layer protocol processing on the data packet is determined, and the upper-layer protocol processing is performed, using the target socket and the second protocol stack instance. The second protocol stack instance that performs the upper-layer protocol processing is determined using the target socket, and hence, a technical problem that protocol processing cannot be performed on a data packet because a protocol stack instance specified by an application (APP) conflicts with a protocol stack instance specified by a network adapter is resolved.

公开(公告)号：US20170329558A1

公开(公告)日：2017-11-16

申请号：US15667858

申请日：2017-08-03

Applicant: Huawei Technologies Co., Ltd.

Inventor： Lin Chen ,  Xiangyang Xu ,  Liufei Wen

IPC: G06F3/06

CPC classification number: G06F3/0659 ,  G06F3/0605 ,  G06F3/061 ,  G06F3/0635 ,  G06F3/067 ,  G06F16/00 ,  G06F16/183

Abstract: A method and device for processing an input/output (I/O) request in a network file system (NFS) includes sending, by a NFS server, a request for parsing the unidentifiable NFS FH to a centralized controller when a NFS file handle (NFS FH) in an I/O request cannot be identified, receiving, by the NFS server, a file identifier that corresponds to the unidentifiable NFS FH from the centralized controller according to the parsing request, where the file identifier is determined according to a pre-stored correspondence between NFS FHs and file identifiers, and processing, by the NFS server, the I/O request according to the file identifier.

公开(公告)号：US20160239337A1

公开(公告)日：2016-08-18

申请号：US15139724

申请日：2016-04-27

Applicant: Huawei Technologies Co., Ltd.

Inventor： Qiang Gu ,  Liufei Wen

IPC: G06F9/455 ,  H04L12/931

CPC classification number: G06F9/45558 ,  G06F9/544 ,  G06F2009/45562 ,  H04L49/70 ,  H04L69/12 ,  H04L69/321

Abstract: A method and computing device for selecting a protocol stack for performing protocol processing on data is presented. The computing device is configured with a hypervisor for managing a first virtual machine. According to the method, when a socket creation instruction sent by the first virtual machine is received, a protocol stack instance is selected from the protocol stack instances provided by the computing device. Then, a socket is created in the selected protocol stack instance according to the socket creation instruction; and a creation result is transmitted to the first virtual machine. Therefore, in a virtualized environment, multiple virtual machines disposed in a same computing device can share a network protocol processing capability, and protocol stacks of the virtual machines achieve load balance.

Abstract translation: 提出了一种用于选择用于对数据执行协议处理的协议栈的方法和计算设备。 计算设备配置有用于管理第一虚拟机的管理程序。 根据该方法，当接收到由第一虚拟机发送的套接字创建指令时，从由计算设备提供的协议栈实例中选择协议栈实例。 然后，根据套接字创建指令在所选协议栈实例中创建套接字; 并将创建结果发送到第一虚拟机。 因此，在虚拟化环境中，设置在同一计算设备中的多个虚拟机可以共享网络协议处理能力，并且虚拟机的协议栈实现负载平衡。

公开(公告)号：US10241830B2

公开(公告)日：2019-03-26

申请号：US14936118

申请日：2015-11-09

Applicant: HUAWEI TECHNOLOGIES CO., LTD.

Inventor： Qiang Gu ,  Liufei Wen ,  Guangyu Shi

IPC: G06F9/46 ,  G06F9/48 ,  H04L29/06 ,  G06F9/54

Abstract: A data processing apparatus and a data processing method are provided. The apparatus includes M protocol stacks and at least one distribution service module, and the M protocol stacks separately run on different logic cores of a processor and are configured to independently perform protocol processing on a data packet to be processed. The distribution service module receives an input data packet from a network interface and sends the data packet to one of the M protocol stacks for protocol processing, and receives data packets processed by the M protocol stacks and sends the data packets outwards through the network interface. The present disclosure implements a function of parallel protocol processing by multiple processes in user space of an operating system in a multi-core environment by using a parallel processing feature of a multi-core system, thereby reducing resource consumption caused by data packet copying.

公开(公告)号：US20160337482A1

公开(公告)日：2016-11-17

申请号：US15222741

申请日：2016-07-28

Applicant: Huawei Technologies Co., Ltd.

Inventor： Qiang Gu ,  Liufei Wen

IPC: H04L29/06 ,  H04L29/08

CPC classification number: H04L69/162 ,  H04L67/10

Abstract: A method and an apparatus for processing a data packet based on parallel protocol stack instances and the method includes determining a distribution policy of a data packet according to distribution policy information of a network adapter, determining a first protocol stack instance according to the distribution policy of the data packet, and creating a target socket in the first protocol stack instance such that when the data packet is distributed to the first protocol stack instance, the first protocol stack instance performs protocol processing on the data packet using the target socket. Hence, a case in which a protocol stack instance that is specified for the target socket by an application conflicts with a protocol stack instance specified by a network adapter is avoided, and a technical problem that protocol processing cannot be performed on the data packet is resolved.

Abstract translation: 一种用于基于并行协议栈实例处理数据分组的方法和装置，所述方法包括根据网络适配器的分发策略信息来确定数据分组的分配策略，根据分发策略确定第一协议栈实例 数据包，并在第一协议栈实例中创建目标套接字，使得当数据包分发到第一协议栈实例时，第一协议栈实例使用目标套接字对数据包执行协议处理。 因此，避免了由应用程序为目标套接字指定的协议栈实例与由网络适配器指定的协议栈实例冲突的情况，并解决了对数据包不能执行协议处理的技术问题 。

公开(公告)号：US20140129510A1

公开(公告)日：2014-05-08

申请号：US14153257

申请日：2014-01-13

Applicant: Huawei Technologies Co., Ltd.

Inventor： Kopylov Vladislav ,  Liufei Wen ,  Guangyu Shi

IPC: G06N5/04

CPC classification number: G06N5/048 ,  G06F17/30011 ,  G06K9/6218

Abstract: A parameter inference method to solve a problem that precision of a Latent Dirichlet Allocation model is poor is provided. The method includes: calculating a Latent Dirichlet Allocation model according to a preset initial first hyperparameter, a preset initial second hyperparameter, a preset initial number of topics, a preset initial count matrix of documents and topics, and a preset initial count matrix of topics and words to obtain probability distributions; obtaining the number of topics, a first hyperparameter, and a second hyperparameter that maximize log likelihood functions of the probability distributions; and determining whether the number of topics, the first hyperparameter, and the second hyperparameter converge, and if not, putting the number of topics, the first hyperparameter, and the second hyperparameter into the Latent Dirichlet Allocation model until the optimal number of topics, an optimal first hyperparameter, and an optimal second hyperparameter that maximize the log likelihood functions of the probability distributions.

Abstract translation: 提供了一种参数推理方法来解决潜在的Dirichlet分配模型的精度差的问题。 该方法包括：根据预设的初始第一超参数，预设的初始第二超参数，预设的初始数量的主题，预设的文档和主题的初始计数矩阵以及主题的预设初始计数矩阵来计算潜在Dirichlet分配模型，以及 词获得概率分布; 获得最大化概率分布的对数似然函数的主题数量，第一超参数和第二超参数; 以及确定主题数量，第一超参数和第二超参数是否收敛，如果不是，将主题数量，第一超参数和第二超参数放入潜在Dirichlet分配模型中，直到最优数量的主题， 最优的第一超参数，以及最大化概率分布的对数似然函数的最优的第二超参数。

公开(公告)号：US10127067B2

公开(公告)日：2018-11-13

申请号：US15139724

申请日：2016-04-27

Applicant: Huawei Technologies Co., Ltd.

Inventor： Qiang Gu ,  Liufei Wen

IPC: G06F9/455 ,  H04L12/931 ,  G06F9/54 ,  H04L29/06 ,  H04L29/08

Abstract: A method and computing device for selecting a protocol stack for performing protocol processing on data is presented. The computing device is configured with a hypervisor for managing a first virtual machine. According to the method, when a socket creation instruction sent by the first virtual machine is received, a protocol stack instance is selected from the protocol stack instances provided by the computing device. Then, a socket is created in the selected protocol stack instance according to the socket creation instruction; and a creation result is transmitted to the first virtual machine. Therefore, in a virtualized environment, multiple virtual machines disposed in a same computing device can share a network protocol processing capability, and protocol stacks of the virtual machines achieve load balance.

公开(公告)号：US09985926B2

公开(公告)日：2018-05-29

申请号：US15213911

申请日：2016-07-19

Applicant: Huawei Technologies Co., Ltd.

Inventor： Xiangyang Xu ,  Liufei Wen

IPC: H04L29/12 ,  H04L12/46 ,  H04L12/713 ,  H04L12/743

CPC classification number: H04L61/103 ,  H04L12/4633 ,  H04L45/586 ,  H04L45/64 ,  H04L45/7453 ,  H04L61/6022

Abstract: An address acquiring method includes receiving an address resolution request packet sent by a source host, where the address resolution request packet includes an Internet Protocol (IP) address of a destination host; determining another network virtualization edge (NVE) device, where the another NVE device stores a correspondence between the IP address of the destination host and a Media Access Control (MAC) address of the destination host and a correspondence between the IP address of the destination host and an IP address of a destination NVE device corresponding to the destination host; obtaining the MAC address of the destination host and the IP address of the destination NVE device corresponding to the destination host from the another NVE device according to the IP address of the destination host. The technical solutions provided in the present disclosure are intended to reduce processing pressure on a physical network.

公开(公告)号：US20150277993A1

公开(公告)日：2015-10-01

申请号：US14738436

申请日：2015-06-12

Applicant: Huawei Technologies Co., Ltd.

Inventor： Lin Gu ,  Zhiqiang Ma ,  Zhonghua Sheng ,  Liufei Wen

IPC: G06F9/52 ,  G06F9/455

CPC classification number: G06F9/52 ,  G06F9/455 ,  G06F9/45558 ,  G06F2009/45583

Abstract: A task processing method and virtual machine are disclosed. The method includes selecting an idle resource for a task; creating a global variable snapshot for a global variable; executing the task, in private memory space in the selected idle resource; after the execution of the task is complete, acquiring a new global variable snapshot corresponding to the global variable, and acquiring an updated global variable according to a local global variable snapshot and the new global variable snapshot; and determining whether a synchronization variable of a to-be-executed task in a task synchronization waiting queue includes the current updated global variable, and if the synchronization variable of the to-be-executed task in the task synchronization waiting queue includes the current updated global variable, putting the task into a task execution waiting queue.

Abstract translation: 公开了一种任务处理方法和虚拟机。 该方法包括为任务选择空闲资源; 为全局变量创建全局变量快照; 在所选择的空闲资源中的专用存储器空间中执行该任务; 在执行任务完成后，获取与全局变量相对应的新的全局变量快照，并根据本地全局变量快照和新的全局变量快照获取更新的全局变量; 并且确定任务同步等待队列中的待执行任务的同步变量是否包括当前更新的全局变量，并且如果任务同步等待队列中待执行任务的同步变量包括当前更新的全局变量 全局变量，将任务放入任务执行等待队列中。

公开(公告)号：US08527448B2

公开(公告)日：2013-09-03

申请号：US13722078

申请日：2012-12-20

Applicant: Huawei Technologies Co., Ltd.

Inventor： Kopylov Vladislav ,  Liufei Wen ,  Guangyu Shi

IPC: G06F9/44 ,  G06N7/02 ,  G06N7/06 ,  G06F15/18 ,  G06F7/00 ,  G06F17/30

CPC classification number: G06N5/02 ,  G06F17/2785 ,  G06F17/3071 ,  G06N5/003

Abstract: Embodiments of the present invention disclose a data processing method including: sending global initial statistical information to each slave node; merging received local statistical information of each slave node, to obtain new global statistical information; if Gibbs sampling performed by a slave node has ended, calculating a probability distribution between a document and topic and a probability distribution between the topic and a word according to the new global statistical information; according to the probability distributions obtained through calculation, establishing a likelihood function of a text set, and maximizing the likelihood function, to obtain a new hLDA hyper-parameter; and if iteration of solving for an hLDA hyper-parameter has converged, and according to the new hLDA hyper-parameter, calculating and outputting the probability distribution between the document and topic and the probability distribution between the topic and word.

Abstract translation: 本发明的实施例公开了一种数据处理方法，包括：向每个从节点发送全局初始统计信息; 合并接收每个从节点的本地统计信息，获得新的全局统计信息; 如果从节点执行的吉布斯抽样已经结束，则根据新的全局统计信息计算文档和主题之间的概率分布以及主题与单词之间的概率分布; 根据通过计算获得的概率分布，建立文本集的似然函数，并最大化似然函数，获得新的hLDA超参数; 并且如果求解hLDA超参数的迭代已经收敛，并且根据新的hLDA超参数，计算和输出文档和主题之间的概率分布以及主题与词之间的概率分布。