公开(公告)号：US20170242767A1

公开(公告)日：2017-08-24

申请号：US15589856

申请日：2017-05-08

Applicant: HUAWEI TECHNOLOGIES CO., LTD.

Inventor： Daohui Wang ,  Feng Zhang ,  Xuyou Liu

IPC: G06F11/20 ,  G06F12/02 ,  G06F9/50 ,  G06F17/30 ,  G06F9/46

Abstract: A distributed storage and replication system includes a MDC module, multiple IO routing modules, and multiple OSD nodes. The MDC module is adapted to configure at least two partition, the IO routing module is adapted to route an IO request to an OSD node, and the OSD node is adapted to execute storage of data corresponding to the IO request. The MDC is configured to determine a faulty OSD node, update a partition view of a partition group that includes a partition on the faulty OSD node, and send an updating notification to a primary OSD node in the updated partition view. The primary OSD node is adapted to process replication of the data corresponding to the IO request. According to embodiments of the present disclosure, processing performance, fault tolerance, and availability of consistency replication are improved.

公开(公告)号：US20160357634A1

公开(公告)日：2016-12-08

申请号：US15173327

申请日：2016-06-03

Applicant: HUAWEI TECHNOLOGIES CO.,LTD.

Inventor： Daohui Wang ,  Feng Zhang ,  Wei Fan ,  Zhile Zhang ,  Yongqiang Zeng

IPC: G06F11/10 ,  G06F3/06

CPC classification number: G06F11/1076 ,  G06F3/0617 ,  G06F3/0619 ,  G06F3/0631 ,  G06F3/064 ,  G06F3/067 ,  G06F9/466 ,  G06F9/5011 ,  G06F11/00

Abstract: A data storage method is used to improve storage consistency of a distributed storage system. The method includes: a primary storage node performs EC coding on a to-be-stored data segment to obtain a target EC stripe; determines in a storage node group to which the primary storage node belongs, m+k target storage nodes used to store m+k target EC blocks of the target EC stripe; sends a preparation message to the target storage nodes; receives a response message sent by a target storage node; and sends an execution message to the target storage nodes to instruct the target storage nodes to write target EC blocks that are in preparation logs.

Abstract translation: 数据存储方法用于提高分布式存储系统的存储一致性。 该方法包括：主存储节点对待存储的数据段执行EC编码以获得目标EC条带; 在主存储节点所属的存储节点组中确定m + k个目标存储节点，用于存储目标EC条带的m + k个目标EC块; 向目标存储节点发送准备消息; 接收目标存储节点发送的响应消息; 并向目标存储节点发送执行消息以指示目标存储节点写入准备日志中的目标EC块。

公开(公告)号：US09417962B2

公开(公告)日：2016-08-16

申请号：US14714754

申请日：2015-05-18

Applicant: Huawei Technologies Co., Ltd.

Inventor： Wei Zhang ,  Feng Zhang ,  Xiaosong Lei ,  Rui Xiang

IPC: G06F11/10 ,  H03M13/15

CPC classification number: G06F11/1076 ,  G06F11/1008 ,  G06F11/1084 ,  G06F11/1096 ,  H03M13/154

Abstract: Technologies are described herein for recovering data in a storage device comprising a controller and a plurality of storage units. The controller receives a data stream, and divides the data stream into a plurality of data blocks, obtains a code block using the plurality of data blocks. When there is one or more blocks with damaged data in the plurality of data blocks and the code block, the controller obtains a sub-block from the Mth bit to the Nth bit of each block in the plurality of data blocks and the code block as a set, and reconstructs data in one or more sub-blocks with damaged data using other sub-blocks with undamaged data in the set.

Abstract translation: 这里描述了用于在包括控制器和多个存储单元的存储设备中恢复数据的技术。 控制器接收数据流，并将数据流划分为多个数据块，使用多个数据块获得代码块。 当在多个数据块和代码块中存在一个或多个具有损坏数据的块时，控制器从多个数据块中的每个块的第M位到第N位获取子块，并且代码块作为 一组，并且在具有损坏数据的一个或多个子块中使用其中具有未损坏数据的子块重构数据。

公开(公告)号：US20150293707A1

公开(公告)日：2015-10-15

申请号：US14752068

申请日：2015-06-26

Applicant: Huawei Technologies Co., Ltd.

Inventor： Feng Zhang

IPC: G06F3/06 ,  G06F12/02

CPC classification number: G06F3/0607 ,  G06F3/0629 ,  G06F3/0632 ,  G06F3/0644 ,  G06F3/067 ,  G06F12/0223 ,  G06F2212/1048 ,  G06F2212/152 ,  G06F2212/154 ,  G06F2212/254 ,  G06F2212/261 ,  G06F2212/263

Abstract: In a distributed storage system, a method for extending a number N of data node devices comprises a server receives an instruction of adding X new data nodes into the distributed storage system. Then, the server obtains a number M of the all partitions included in a hash ring and determines that M/(N+X) is lower than a preset threshold. The hash ring is organized according to a distributed hash table (DHT), and includes a plurality of partitions. Each partition is mapping to a data node. Based upon the determination, the server generates new partitions by multiplying partitions mapping to each data node, based upon the determination. After storing mapping relationship between the new partitions and the X new data nodes, the server adds the X new data nodes into the distributed storage system.

Abstract translation: 在分布式存储系统中，扩展N个数据节点设备的方法包括：服务器接收将X个新数据节点添加到分布式存储系统中的指令。 然后，服务器获得散列环中包括的所有分区的数量M，并确定M /（N + X）低于预设阈值。 散列环根据分布式哈希表（DHT）来组织，并且包括多个分区。 每个分区映射到数据节点。 基于确定，服务器基于确定，通过将映射到每个数据节点的分区相乘来生成新的分区。 在存储新分区和X新数据节点之间的映射关系之后，服务器将X个新数据节点添加到分布式存储系统中。

公开(公告)号：US20150178170A1

公开(公告)日：2015-06-25

申请号：US14639657

申请日：2015-03-05

Applicant: Huawei Technologies Co., Ltd.

Inventor： Feng Zhang ,  Fengtao Teng

IPC: G06F11/14 ,  G06F11/20

CPC classification number: G06F11/1451 ,  G06F11/1435 ,  G06F11/1448 ,  G06F11/2058 ,  G06F17/30073 ,  G06F17/30194 ,  G06F2201/84

Abstract: In a data recovery method, there are a server and a plurality of storage devices each storing a copy of a data block. The server divides each copy of the data block into N segments corresponding to a sequence of N partitions. And then, the server constructs a plurality of different trial data blocks each including N segments corresponding to the sequence of N partitions. After that, the server calculates a check code for each trial data block, and continues to identify a trial data block having a check code identical to a pre-stored standard check code of the data block. At last, the server replaces at least one of the copies of the data block with the identified trial data block having the check code identical to the pre-stored standard check code.

Abstract translation: 在数据恢复方法中，存在服务器和多个存储装置，每个存储装置存储数据块的副本。 服务器将数据块的每个副本分成与N个分区序列相对应的N个段。 然后，服务器构建多个不同的试验数据块，每个试验数据块包括对应于N个分区的序列的N个段。 之后，服务器计算每个试验数据块的校验码，并且继续识别具有与数据块的预先存储的标准校验码相同的校验码的试验数据块。 最后，服务器用具有与预先存储的标准校验码相同的校验码的所识别的试验数据块替换数据块的副本中的至少一个。

公开(公告)号：US08736656B2

公开(公告)日：2014-05-27

申请号：US13975862

申请日：2013-08-26

Applicant: Huawei Technologies Co., Ltd

Inventor： Feng Zhang

IPC: H04N7/14

CPC classification number: H04N7/152 ,  H04N5/2624

Abstract: A method and an apparatus for Multiple Control Unit (MCU) to optimize configuration of multiple pictures are disclosed. The solutions provided by embodiments of the present invention include: calculating, according to resolutions, bandwidths, frame rates, importance, and volume of received video images sent by N conferencing terminals, the area occupied by each of the video images in a picture presented on an MCU display screen; and dividing, according to the area occupied by each of the video images in the picture presented on the MCU display screen, the picture into N regions, and filling the video images sent by the N conferencing terminals into corresponding regions of the picture.

Abstract translation: 公开了一种用于优化多个图像的配置的多控制单元（MCU）的方法和装置。 根据本发明的实施例提供的解决方案包括：根据分辨率，计算由N个会议终端发送的接收视频图像的带宽，帧率，重要性和音量，图像中呈现的图像中每个视频图像所占据的区域 一个MCU显示屏; 并根据在MCU显示屏幕上显示的图像中的每个视频图像所占据的面积划分为N个区域，并将N个会议终端发送的视频图像填充到图像的相应区域中。