Alibaba Group Holding Limited

Hangzhou Hikvision Digital Technology Co., Ltd.

Advanced New Technologies Co., Ltd.

Hangzhou Hikvision System Technology Co., Ltd.

Peng Lin

Qiqian Lin

Weichun Wang

Min Ye

Fanli Zeng

Shiliang Pu

Xin Jiang

Sansan Ding

Shuanbao Yao

Conghui Zhang

Yunfeng Li

Dawei Chen

Shaoqing Liu

A data processing method and device, and a distributed storage system are described. The method and device are applied in a dual-control storage server. The dual-control storage server comprises two controllers. If one controller fails due to abnormalities, the other controller determines the storage device managed by the failed controller, and the other controller scans the data in the determined storage device to obtain the metadata of the storage device, and uses the metadata to read the data stored in the determined storage device. It can be seen that in this solution, if one controller is abnormal, the other controller will take the place of the failed controller to provide external services. This improves the stability of the data storage by the dual-control storage server.

Data processing method, device and distributed storage system

Implementations of this specification provide a method and an apparatus for changing a master node in a blockchain system. An example method performed by a backup node in the blockchain system includes determining that an epoch change is to occur in a blockchain system, generating a random number, and in response to determining that the random number satisfies an epoch change trigger condition, sending an epoch change request message to multiple network nodes other than the backup node in the blockchain system. The backup node receives a request acknowledgement message from at least one of the multiple network nodes, the request acknowledgement message indicating that the backup node is to serve as the new master node, and in response to receiving a number of request acknowledgement messages that is greater than a predetermined number of messages, the backup node sends an epoch change acknowledgement message to the multiple network nodes.

Changing a master node in a blockchain system

The embodiments of the present disclosure disclose a method and apparatus for reading or writing disk data. The disk includes a primary metadata block storing metadata, a backup metadata block for backing up metadata, and at least one object block group. Each object block group includes a data block storing object data, and an index block storing an index of the object data. The primary metadata block is located at a head of the disk; and the backup metadata block is located at a tail of the disk. The at least one object block group is located between the primary metadata block and the backup metadata block. In reading or writing object data, the index in the index block, the metadata in the primary metadata block and the metadata backed up in the backup metadata block can be separately operated. By applying the embodiments of the present disclosure, the life and security of the disk are increased.

Disk data reading/writing method and device

Embodiments of the present application provide a method and apparatus for displaying objects. In the method, data segments of to-be-stored data are stored in storage objects. A storage object is a minimum unit for data storage. The difference between the number of data segments stored in a storage object and that of another storage object is no more than a first preset threshold. Data segments are equally stored, to the greatest extent, into storage objects. The case that data having segments lost due to failure of a storage object cannot be restored is avoided.

Data storage, distribution, reconstruction and recovery methods and devices, and data processing system

A method of performing a change of a primary node in a blockchain network includes a backup node of the blockchain network determining that an epoch change needs to be performed, determining a respective weight of the backup node associated with each of three phases of a consensus process in a current epoch, determining a weight sum for the backup node based on the respective weights, sending an EPOCH_CHANGE message to the other network nodes to apply for a new primary node in a new epoch, receiving NEW_EPOCH messages from the other network nodes, determining whether a number of valid NEW_EPOCH messages exceeds a second predetermined threshold, and determining the backup node to be the new primary node in the new epoch in response to determining that the number of valid NEW_EPOCH messages exceeds the second predetermined threshold.

Performing a change of primary node in a distributed system

Implementations of the present specification include a computer-implemented method for achieving a consensus among a number of network nodes of a blockchain network. The blockchain network includes a primary node and one or more backup nodes. The method includes receiving a transaction request by the primary node, sending a number of first messages to the backup nodes by the primary node, receiving second messages from the backup nodes by the primary node, reconstructing the transaction request based on data in the second messages by the primary node, sending a third message to the backup nodes by the primary node, and executing the transaction request in response to receiving a predetermined number of third messages.

Achieving consensus among network nodes in a distributed system

A method for performing a recovery process of a network node of a blockchain network includes broadcasting a state request message by the network node to the other network nodes of the blockchain network to recover a target transaction of a target sequence number, receiving state reply messages that each includes a sequence number from the other network nodes, identifying the target sequence number based on the sequence numbers in the state reply messages, sending a requesting message to the other network nodes to request an ECHO message from each of the other network nodes, determining a number of valid ECHO messages that are sent by the other network nodes, recovering the target transaction based on the number of valid ECHO messages, and sending a message to the other network nodes indicating that the network node has been recovered.

Performing a recovery process for a network node in a distributed system

The present application discloses a storage device and a block storage method based on the storage device, wherein disk space of the storage device includes a data area having a plurality of end-to-end adjacent data block units, each of which includes a metadata block and a data block adjacent to each other; wherein the data block with a first preset capacity is configured to store target data, and the metadata block with a second preset capacity is configured to store information of the target data. With the storage device, metadata is stored in a distributed manner, wherein each of the metadata blocks and the data blocks are stored as a whole. The address offset of updating metadata after writing data is thus reduced, and the metadata update efficiency, as well as the overall storage performance of the storage device are improved. In addition, with the metadata stored in this distributed manner, the probability of data unavailability due to continuous damage of a plurality of metadata blocks is greatly reduced, and the defects in the prior art that the intact data blocks cannot store data normally due to continuous damage of a plurality of metadata is overcome.

Storage device and block storage method based on the storage device

A method of storing data includes: determining a number of copies to be initially stored for a data block x after receiving a request for storing the data block x, and storing the data block x according to the number of copies to be initially stored; and periodically determining a current importance level of the data block x, determining whether a number of copies corresponding to the importance level is identical to a number of currently stored copies of the data block x, and adjusting the number of currently stored copies of the data block x to be consistent with the number of copies corresponding to the importance level if the number of copies corresponding to the importance level is not identical to the number of currently stored copies of the data block x. The present disclosure also provides a data storage managing server.

Growing community of inventors

Hangzhou, China

Peng Lin