公开(公告)号：US09996421B2

公开(公告)日：2018-06-12

申请号：US14583321

申请日：2014-12-26

Applicant: HUAWEI TECHNOLOGIES CO.,LTD.

Inventor： Jianlong Xiao ,  Peng Zhang

IPC: G06F11/00 ,  G06F11/14 ,  G06F11/16 ,  G06F11/20

CPC classification number: G06F11/1435 ,  G06F11/1441 ,  G06F11/1666 ,  G06F11/2015 ,  G06F11/2092

Abstract: In a backup method, a first controller of a storage system receives a file process request, and divides the file process request into a plurality of file operations. The first controller processes each of the file operations to generate multiple pieces of file operation information. And then, the first controller writes all of the multiple pieces of file operation information into a cache protected area. When the first controller fails, it stores the multiple pieces of file operation information to a disk protected area of a disk in the storage system. After the first controller is powered on again, the first controller copies the multiple pieces of file operation information from the disk protected area to the cache protected area. And then, it writes the file operation information into a write cache. At last, the first controller stores the file operation information into a disk of the storage system.

公开(公告)号：US20210278983A1

公开(公告)日：2021-09-09

申请号：US17239194

申请日：2021-04-23

Applicant: Huawei Technologies Co., Ltd.

Inventor： Jianlong Xiao ,  Feng Wang ,  Qi Wang ,  Chen Wang ,  Chunhua Tan

IPC: G06F3/06

Abstract: A node capacity expansion method in a storage system and a storage system, where the storage system includes a first node, and a data partition group and a metadata partition group are configured for the first node, where the data partition group includes a plurality of data partitions, the metadata partition group includes a plurality of metadata partitions, and metadata of data in the data partition group is a subset of metadata in the metadata partition group. When a second node is added to the storage system, the first node splits the metadata partition group into at least two metadata partition subgroups, and migrates a first metadata partition subgroup in the at least two metadata partition subgroups and metadata in the first metadata partition subgroup to the second node.

公开(公告)号：US11030048B2

公开(公告)日：2021-06-08

申请号：US15459984

申请日：2017-03-15

Applicant: Huawei Technologies Co., Ltd.

Inventor： Jianlong Xiao ,  Keji Huang

IPC: G06F16/16 ,  G06F11/14 ,  G06F16/178 ,  G06F16/23 ,  G06F16/27 ,  G06F16/2455 ,  G06F16/11 ,  G06F16/174

Abstract: A method for a source storage device to send a source file and a clone file of the source file to a backup storage device. The source storage device sends a data block to the backup storage device to be stored as a target file. Then, the source storage device determines that the source file is associated with the clone file by searching a cloning recorder with a source file ID of the source file. Based upon the determining, the source storage device sends a clone creating request including the source file ID to the backup storage device to instruct the backup storage device to create a clone file of the target file. Thus, the method is capable of replicating the clone file of the file without sending a large number of mapping data.

公开(公告)号：US20150121130A1

公开(公告)日：2015-04-30

申请号：US14583321

申请日：2014-12-26

Applicant: HUAWEI TECHNOLOGIES CO., LTD.

Inventor： Jianlong Xiao ,  Peng Zhang

IPC: G06F11/20 ,  G06F11/10

CPC classification number: G06F11/1435 ,  G06F11/1441 ,  G06F11/1666 ,  G06F11/2015 ,  G06F11/2092

Abstract: In a backup method, a first controller of a storage system receives a file process request, and divides the file process request into a plurality of file operations. The first controller processes each of the file operations to generate multiple pieces of file operation information. And then, the first controller writes all of the multiple pieces of file operation information into a cache protected area. When the first controller fails, it stores the multiple pieces of file operation information to a disk protected area of a disk in the storage system. After the first controller is powered on again, the first controller copies the multiple pieces of file operation information from the disk protected area to the cache protected area. And then, it writes the file operation information into a write cache. At last, the first controller stores the file operation information into a disk of the storage system.

Abstract translation: 在备份方法中，存储系统的第一控制器接收文件处理请求，并将文件处理请求分成多个文件操作。 第一控制器处理每个文件操作以生成多条文件操作信息。 然后，第一个控制器将所有多条文件操作信息写入缓存保护区。 当第一控制器发生故障时，将多条文件操作信息存储到存储系统中的磁盘的磁盘保护区域。 在第一个控制器再次通电后，第一个控制器将多条文件操作信息从磁盘保护区复制到缓存保护区。 然后，它将文件操作信息写入写入高速缓存。 最后，第一控制器将文件操作信息存储到存储系统的盘中。