Huawei Technologies Co., Limited

Fudan University

University of Science and Technology of China

Yuangang Wang

Jun Xu

Yong Qing Wang

Qiang Wang

Wei Yang

Junfeng Zhao

Qun Yu

Yinyin Lin

Kai Yang

Yarong Fu

Liang Shi

Guanyu Zhu

Dongfang Shan

Xing Hu

Long Sun

Jiwu Shu

Yongbing Huang

Qiao Li

Zhipeng Li

RenHua Yang

Fen Dai

Yi Cai

Hsing Mean Sha

Chun Xue

Hao Xiao

Yinlong Xu

Yong Wang

Yejia Di

Tianxiang Li

Kai c/o Huawei Administration Building Yang

Yinyin c/o Huawei Administration Building Lin

Yarong c/o Huawei Administration Building Fu

A storage system includes a management node and a plurality of storage nodes forming a redundant array of independent disks (RAID). When the management node determines that not all data in an entire stripe is updated based on a received write request, the management node sends an update data chunk obtained from to-be-written data to a corresponding storage node. The storage node does not directly update, based on the received update data chunk, a data block stored in a storage device of the storage node, but store the update data chunk into a non-volatile memories (NVM) cache of the storage node and send the update data chunk to another storage node for backup. According to the data updating method, write amplification problems caused in a stripe update process can be reduced, thereby improving update performance of the storage system.

Data updating technology

A file management method, a distributed storage system, and a management node are disclosed. In the distributed storage system, after receiving a file creation request sent by a host for requesting to create a file in a distributed storage system, a management node allocates, to the file, first virtual space from global virtual address space of the distributed storage system, where local virtual address space of each storage node in the distributed storage system is corresponding to a part of the global virtual address space. Then, the management node records metadata of the file, where the metadata of the file includes information about the first virtual space, and the information about the first virtual space is used to point to local virtual address space of a storage node that is used to store the file. Further, the management node sends, the information about the first virtual space to the host.

File management method, distributed storage system, and management node

A storage system includes a management node and a plurality of storage nodes forming a redundant array of independent disks (RAID). When the management node determines that not all data in an entire stripe is updated based on a received write request, the management node sends update data chunk obtained from to-be-written data to corresponding storage node. The storage node do not directly update, based on the received update data chunks, data block stored in storage device of the storage node, but store the update data chunk into non-volatile memories (NVM) cache of the storage node and send the update data chunk to another storage node to backup. According to the data updating method, write amplification problems caused in a stripe update process can be reduced, thereby improving update performance of the storage system.

An access request processing method is performed by a computer device that includes a processor, a dynamic random-access memory (DRAM), and a non-volatile memory (NVM). When receiving a write request, the processor may identify an object cache page according to the write request. The processor obtains the to-be-written data from a buffer according to a buffer pointer in the write request, the to-be-written data including a new data chunk to be written into the object cache page. The processor then inserts a new data node into a log chain of the object cache page, where the NVM stores data representing the log chain of the object cache page. The new data node includes information regarding the new data chunk of the object cache page. The computer device provided in this application can reduce system overheads while protecting data consistency.

Access request processing method and apparatus, and computer device

A data access method, a routing apparatus, and a storage system are provided. The method is applied to a storage system including a first storage device, a second storage device, and a routing apparatus. A logical unit in each storage device includes at least one first-type logical block and at least one second-type logical block. According to the method, when sending access requests to the storage devices in the storage system according to a preset rule, the routing apparatus sends access requests corresponding to same target logical blocks to one of the storage devices according to a preset rule. This reduces network overheads between the storage system and the application server, and improves efficiency of processing the access requests by the storage system.

Data access method, routing apparatus, and storage system

A data migration method for a storage system after expansion and a storage system are provided. After an iexpansion is performed on the storage system, data migration is performed by using an auxiliary balanced incomplete block design. Because a quantity of tuples including any element in the auxiliary balanced incomplete block design is identical, and each migration unit includes an identical quantity of parity chunks, a data migration amount after the expansion is minimized. In this way, time required for data migration after the expansion is significantly reduced, and a delay in a response to a user request that is caused because a data migration operation needs to be performed after the expansion is also reduced.

Data migration method for a storage system after expansion and storage system

A cache management method, a cache controller, and a computer system are provided. In the method, the cache controller obtains an operation instruction; when a destination address in the operation instruction hits no cache line cache line in a cache of the computer system, and the cache includes no idle cache line, the cache controller selects a to-be-replaced cache line from a replacement set, where the replacement set includes at least two cache lines; and the cache controller eliminates the to-be-replaced cache line from the cache, and stores, in the cache, a cache line obtained from the destination address. According to the cache management method, system overheads of cache line replacement can be reduced, and cache line replacement efficiency can be improved.

Cache management method, cache controller, and computer system

A flash memory controller refreshes memory blocks in a flash memory device by setting different refresh cycles for individual memory blocks in the flash memory device. The flash memory controller records a number of erase operations performed on each memory block of the flash memory device. Upon detecting that a bit error rate of a memory block is greater than a preset threshold, the flash memory controller determines a refresh cycle for the memory block based on recorded number of erase operations performed on the memory block, and then refreshes the memory block according to the refresh cycle.

Method and apparatus for refreshing flash memory device

A method for accessing a flash memory device and a flash memory device. After receiving a write request for an address, a flash memory controller obtains an indicator of the address, where the indicator indicates a last access type of the address, which might be a write operation or a read operation. When determining the indicator indicates a write operation, which means the access type for the address is normally write operation, to save time, the flash memory controller perform a fast-write operation on the address, when the indicator indicates a read operation, which means there might be plenty of read operations on the address, to facilitate future read operation, the controller performs a slow-write operation on the address.

Method and apparatus for accessing flash memory device

A processing node, a computer system, and a transaction conflict detection method, where the processing node includes a processor and a transactional cache. When obtaining a first operation instruction in a transaction for accessing shared data, the processor accesses the transactional cache for caching shared data of a transaction processed by the processing node. If the transactional cache determines that the first operation instruction fails to hit a cache line in the transactional cache, the transactional cache sends a first destination address in the operation instruction to a transactional cache in another processing node. After receiving status information of a cache line hit by the first destination address from the other processing node, the transactional cache determines, based on the received status information, whether the first operation instruction conflicts with a second operation instruction executed by the other processing node.

Processing node, computer system, and transaction conflict detection method

A memory extensible chip () is provided. The chip () includes a substrate (), and a processor (), a first memory module set (), and a second memory module set () that are integrated on the substrate (). The processor () communicates with at least one memory module in the first memory module set () using a first communications interface (), and the processor () communicates with at least one memory module in the second memory module set () using a second communications interface (). A memory module in the first memory module set () communicates with a memory module in the second memory module set () using a substrate network, where the substrate network is a communications network located inside the substrate (). In this way, the processor () can access a memory module in the first memory module set () by using the second memory module set ().

Memory extensible chip

An access request processing method and apparatus, and a computer device are disclosed. The computer device includes a processor, a dynamic random-access memory (DRAM), and a non-volatile memory (NVM). When receiving a write request, the processor may identify an object cache page according to the write request. The processor obtains the to-be-written data from a buffer according to a buffer pointer in the write request, the to-be-written data including a new data chunk to be written into the object cache page. The processor then inserts a new data node into a log chain of the object cache page, where the NVM stores data representing the log chain of the object cache page. The new data node includes information regarding the new data chunk of the object cache page. The computer device provided in this application can reduce system overheads while protecting data consistency.

A file data access method and a computer system, where the method includes accessing a page global directory (PGD) of the process using PGD space when accessing first file data by a process, determining, based on access to the PGD and according to a first virtual address of the first file data in file system space, a first PGD entry in the PGD, linking a file page table of the process to the first PGD entry, where the file page table points to a physical address of the file data such that a processor retrieves a first physical address of the first file data in a memory according to the first virtual address using the PGD and the file page table, and accessing the first file data according to the first physical address.

Growing community of inventors

Shenzhen, China

Yuangang Wang