Huawei Technologies Co., Limited

International Business Machines Corporation

Rui Feng

Lijun Wei

Shuang Shuang Jia

Da Fei Shi

Jun Liu

Guang Chen

Dingguo Yang

Shengqian Jia

Guangyou Xiang

A garbage collection technology for a storage medium that includes a plurality of blocks, where each storage unit has a collection parameter related to data in one of the blocks, and where each block includes a plurality of pages. A group of blocks is selected from the plurality of blocks based on the collection parameter. A difference between values of collection parameters of any two blocks in the group of blocks is not greater than a preset value. Data in a first valid page and a second valid page in the group of blocks is replicated to a same destination block, to facilitate garbage collection.

Garbage collection method for storage medium, storage medium, and program product

Sensitive data is protected in a software product. A source file of the software product is compiled to generate an object file, in which the source file includes at least one piece of sensitive data marked with a specific identifier. The object file has a secure data section for saving storage information of the at least one piece of sensitive data at compile-time and run-time. The object file is linked to generate an executable file. The executable file updates the secure data section at run-time. Sensitive data is also protected when a core dump is generated.

Protecting sensitive data in software products and in generating core dumps

A technique for processing a lock request is provided. A first lock server is a takeover lock server of a second lock server. The first lock server enters a silent state after learning that a fault occurs in the second lock server, where a silent range is a resource for which the second lock server has assigned permission. The first lock server receives an acquire lock request that is originally sent to the second lock server, and the first lock server assigns lock permission for a corresponding resource according to the acquire lock request if the second lock server has not assigned resource for the resource. By means of this solution, an impact range of a fault occurring in a lock server can be reduced, and stability of a lock management system is improved.

Method for processing acquire lock request and server

The present disclosure discloses an arbitration processing solution when brain split occurs in cluster. The solution includes: receiving, by a quorum storage apparatus within a first refresh packet detection period, first master quorum node preemption requests sent by at least two quorum nodes in the cluster; sending, by the quorum storage apparatus, a first master quorum node preemption success response message to the initial master quorum node indicating that the initial master quorum node succeeds in master quorum node preemption when the first master quorum node preemption requests received within the first refresh packet detection period comprise the master quorum node preemption request sent by the initial master quorum node.

Arbitration processing method after cluster brain split, quorum storage apparatus, and system

The present disclosure discloses an arbitration processing solution when brain split occurs in cluster. The solution includes: receiving, by the quorum storage apparatus, first master quorum node preemption requests within a first master quorum node preemption detection period sent by the at least two quorum nodes when brain split occurs in the cluster; sending, by the quorum storage apparatus, a first master quorum node preemption success response message to the initial master quorum node when the first master quorum node preemption request sent by the initial master quorum node is received; where the first master quorum node preemption success response message indicates the initial master quorum node succeeds in master quorum node preemption.

A method for processing a fault of a lock server in a distributed system is disclosed, where the distributed system includes m lock servers, which locally store same lock server takeover relationship information. Lock servers in the distributed system that are not faulty receive a notification message, which carries information about a fault of a first lock server; after receiving the notification message, a second lock server determines that it is a takeover lock server of the first lock server according to the lock server takeover relationship information, and the takeover lock server enters a silent state; after receiving the notification message, a third lock server in the distributed system determines that it is not the takeover lock server of the first lock server according to the lock server takeover relationship information. After receiving a locking request, the third lock server allocates lock permission information according to the locking request.

Growing community of inventors

Beijing, China

Rui Feng