The invention provides a data restoration method based on version vector, and a server. Both a client side and a server side store a version tag of each backed file and a version tag of each restored file. When the client side requests for data restoration, the server side determines whether requested data is stored on the client side or not according to the version tag provided by the client side, deletes the data which are stored on the client side, and transmits only an indicator of repeating data to the client side. The repeating data deletion method based on version requires only transmission of the version vector of the restored data block, the method is more efficient, network bandwidth is economized, and restoration time is reduced.

Data restoration method based on version vector, and server

The invention discloses a transaction recovery method in a database system, and a database management system. Storage equipment of the database system supports copy on write. The method comprises the following steps of: determining a transaction to be recovered in the database system, wherein the transaction to be recovered is not submitted; and the transaction to be recovered performs the update operation of data in the database system; obtaining a physical address of an old value of data in the storage equipment from an update operation log for recording that the transaction to be recovered performs the update operation of the data, wherein the old value is a value before the data is updated; replacing a physical address of a new value of the data by utilizing the physical address of the old value, setting the physical address of the new value to be invalid, such that a logic address of the data points to the physical address of the old value, wherein the new value is a value after the data is updated. Compared with the traditional technology, the transaction recovery method in the database system, and the database management system disclosed by the invention have the advantages that: write operation can be effectively reduced when the transaction is recovered; therefore, erasing operation of the storage equipment is reduced; and thus, the service life of the storage equipment can be prolonged.

Transaction Recovery Method in Database System, and Database Management System

Systems and methods are disclosed for performing file backup and restore in a computing system. Certain embodiments provide a communication interface for communicating with a non-volatile memory and a controller configured to determine that a file stored in the non-volatile memory has been modified, identify a chunk of the file that has been modified, determine a new chunk associated with the modified chunk, the new chunk reflecting the modification, generate a separate chunk file including the new chunk and a filename, and storing the chunk file in the non-volatile memory using the communication interface.

Data deduplication using chunk files

The invention discloses a secure deletion method for data in a solid-state disk on the basis of a flash memory. The characteristic that multiple data channels in the solid-state disk can be operated in parallel is utilized, an improved secrete sharing scheme is adopted to conduct conversion processing on data, and encoded data is distributed to all the data channels; on one hand, data redundancy is guaranteed through encoding, the converted data is ciphertext, and the reliability and the security of the data are improved; on the other hand, the error-resilient encoding characteristic is utilized, covering on the whole data is not needed any more for deletion of the data, partial data is deleted, the data integrity is destroyed, and the data cannot be recovered. Even though an attacker obtains partial encoded data, the attacker cannot obtain plaintext, and the purpose of securely deleting the data is achieved. By means of the secure deletion method for the data in the solid-state disk on the basis of the flash memory, the data storage security problems that storage of the solid-state disk is unreliable, and files cannot be deleted securely are solved.

Secure deletion method for data in solid-state disk on basis of flash memory

One embodiment of the invention discloses a method for data backup. The method comprises following steps: obtaining a first backup instruction; wherein the first backup instruction is used for indicating to perform physical backup on a first data block at the first backup time; executing the first backup instruction; a data block which has been modified before the first backup time is determined at the first data block; performing physical backup on the first data block to obtain corresponding first backup data block. The embodiment of the invention also discloses a device for data backup, a method and device for data recovery.

Method and device for data backup and recovery

The invention provides a repeated data delete method based on causal relationship. The method eliminates the backup and recovers the repeated data in the task according to the causal relationship among the data files caught, thus quickening backup and recovered data transmission, improving data backup and recovery speed in a cloud backup system and leading the cloud backup system to be practical and usable.

Repeated data delete method based on causal relationship

The invention relates to a method for avoiding data migration in a hardware affair memory system. The method comprises the following steps: redirecting the write operation in the affair, which operates in the affair memory system, into a free address space by using the address redirection form; in the redirection form, recording the original address and the redirected address in a forecast addition mode, wherein the original address is used for storing data values before the affair operation, and the redirected address is used for storing data values after the affair operation is modified; when the affair is submitted normally, redirecting the visit to the original address to the redirected address; and if the affair comes into collision and needs to be cancelled, clearing the related items which are added into the redirection form, and directly visiting the original address in the subsequent operations. The invention is applicable to any hardware affair memory system, and can greatlyenhance the system performance.

Method for avoiding data migration in hardware affair memory system

As one of the most potential solution to improve thread level parallelism and reduce most ordinary programmers' burden on parallel programming, transactional memory (TM) systems have attracted a great deal of attention from both industry and academic since the notion was proposed in 1993. Since then, various designs and implementations are proposed to improve the performance while reducing the overheads. However, recent investigations of the high-contention and coarse-grained workloads reveal various pathologies that will offset the performance benefits. In this paper, we analysis the advantages and disadvantages of existing conflict management and version management schemes, make a case study in the interplay of them to learn its impact on performance. In particular, we apply a runtime environment to recognize application's dynamic behaviors and resolve transactional conflicts to make up the gap between the upper layer application's diversity and the underline hardware's capability. Throughout the comprehensive evaluation, we find that our proposal can obtain a significant performance improvement across the applications selected from the STAMP benchmark suite on DynTM, which is regarded as one of the latest progress in HTM systems.

Yujuan Tan

Papers

Repeated data delete method based on causal relationship

Method for avoiding data migration in hardware affair memory system

Using A Runtime to Overcome The Pathologies in Hardware Transactional Memory Systems