Abstract: A method and an apparatus for determining a structured query language (SQL) execution plan are provided to optimize determining of the SQL execution plan and improve execution efficiency of the SQL execution plan. The SQL execution plan corresponds to at least one relation table. During an Nth iteration, the method includes obtaining a first iteration parameter generated after a first plan tree is executed on the at least one relation table during an (N?1)th iteration, where N is a natural number greater than 1, establishing a second plan tree according to the first iteration parameter, and determining the first plan tree or the second plan tree as the SQL execution plan when a difference between the second plan tree and the first plan tree is not greater than a first threshold.

Abstract: A database cluster architecture based on dual port solid state disk (SSD) is provided. The database cluster includes a plurality of nodes and a plurality of dual port solid state disks (SSDs). Each dual port SSD of the plurality of dual port SSDs is connected to a unique pair of nodes and each node in the unique pair of nodes is connected to a unique pair of dual port SSDs. A first node of the plurality of nodes writes a transaction log to a first dual port SSD, a second node connected to the first dual port SSD acquires the transaction log from the first dual port SSD and operates a database process to recover data according to the transaction log when the first node crashes.

Abstract: An object encoding and computation method in a database system, where a global object sample space is constituted based on correlation between a plurality of object sample spaces in a database. A global dictionary table is created in the global object sample space to provide a globally unique encoding value for each object, set the encoding value as an object identifier (ID) of the object, and ensure that the object ID and object description information are in a one-to-one mapping relationship. The database object is encoded and compressed based on the global dictionary table. Further, the object ID is directly used for database query computation.

Abstract: A database cluster architecture based on dual port solid state disk (SSD) is disclosed. The database cluster including a plurality of nodes and a plurality of dual port solid state disks (SSDs). Each dual port SSD of the plurality of dual port SSDs is connected to a unique pair of nodes and each node in the unique pair of nodes is connected to a unique pair of dual port SSDs. A first node of the plurality of nodes writes a transaction log to a first dual port SSD, a second node connected to the first dual port SSD acquires the transaction log from the first dual port SSD and operates a database process to recover data according to the transaction log when the first node crashes.

Abstract: The present disclosure provides a data management method for a database cluster including a first dual port SSD, a second dual port SSD, a first node, a second node, and a third node, where the first dual port SSD is connected to the first node and the second node, and the second dual port SSD is connected to the second node and the third node; and the first node writes a transaction log to the first dual port SSD, when the first node crashes, the second node acquires the transaction log from the first dual port SSD, and operates, according to the transaction log, data that is stored in the first node, so that, the second node or the third node can use a dual port SSD to read log information of the crashed node, and replace the first node to provide a service externally.

Abstract: A method and an apparatus for determining a structured query language (SQL) execution plan are provided to optimize determining of the SQL execution plan and improve execution efficiency of the SQL execution plan. The SQL execution plan corresponds to at least one relation table. During an Nth iteration, the method includes obtaining a first iteration parameter generated after a first plan tree is executed on the at least one relation table during an (N?1)th iteration, where N is a natural number greater than 1, establishing a second plan tree according to the first iteration parameter, and determining the first plan tree or the second plan tree as the SQL execution plan when a difference between the second plan tree and the first plan tree is not greater than a first threshold.

Abstract: The present disclosure provides a data management method for a database cluster including a first dual port SSD, a second dual port SSD, a first node, a second node, and a third node, where the first dual port SSD is connected to the first node and the second node, and the second dual port SSD is connected to the second node and the third node; and the first node writes a transaction log to the first dual port SSD, when the first node crashes, the second node acquires the transaction log from the first dual port SSD, and operates, according to the transaction log, data that is stored in the first node, so that, the second node or the third node can use a dual port SSD to read log information of the crashed node, and replace the first node to provide a service externally.

Abstract: Techniques for resolving single source shortest path for database processing are provided. Graph data for nodes having weights for edges of a database network are iterated producing a new message table and results table in each iteration. The results table stores the minimum path weight. For each iteration the graph data and message table are joined to produce a state of a node and outgoing edge messages. The message table and results table are co-grouped to generate a new message for a next node to process. When done the single source shortest path for the network is produced.

Abstract: A database management method and system. The method includes dividing, by a database system, a data partition into a physical partition layer and a logical partition layer, wherein the logical partition segment and the physical partition segment are mapped to each other using a mapping relationship; receiving, by the database system, a data operation instruction of a user, and determining a logical partition segment to which a data record that is requested for operation by the data operation instruction belongs; determining, according to the data operation instruction and the mapping relationship, a physical partition segment that is corresponding to the determined logical partition segment and at the physical partition layer to perform an operation on the data record that is requested for operation by the data operation instruction.

Abstract: A database system lock operation method is provided that includes: acquiring, by a first processing node, a lock operand of a target lock; using, by the first processing node, the target lock to perform a lock acquiring operation on a target data object; when the lock operand meets a condition for a successful lock acquiring operation, determining, by the first processing node, that the lock acquiring operation is successful, and updating a value of a data bit that is in the lock operand and used to indicate a state of the target lock on the first processing node to a value that is used to indicate that the state of the target lock on the first processing node is lock held. Accordingly, the embodiments of the present invention further disclose a node device. The embodiments of the present invention can improve performance of a database system.

Abstract: A method, apparatus, and system for storing an execution plan, and relate to the field of databases, which may improve a cache hit rate of an execution plan and reduce system overheads. The method includes sending, by the client, a structured query language (SQL) request and generated location information to the database server, and according to the received SQL request and the received location information, generating and executing, by the database server, an execution plan corresponding to the SQL request, and saving the execution plan to a cache indicated by the location information. The location information may indicate a location of a cache of a client or a cache of a database server.

Abstract: A database management method and system. The method includes dividing, by a database system, a data partition into a physical partition layer and a logical partition layer, wherein the logical partition segment and the physical partition segment are mapped to each other using a mapping relationship; receiving, by the database system, a data operation instruction of a user, and determining a logical partition segment to which a data record that is requested for operation by the data operation instruction belongs; determining, according to the data operation instruction and the mapping relationship, a physical partition segment that is corresponding to the determined logical partition segment and at the physical partition layer to perform an operation on the data record that is requested for operation by the data operation instruction.

Abstract: Techniques for parallel frequent sequential pattern detection are provided. A sequence database is split into separate datasets and each node is given a specific dataset to resolve specific frequent items occurring in its specific dataset based on counts. Then, each node groups its item frequent items into “n” (varying) length sequences representing sequential patterns present in the original sequence database. The nodes process in parallel with one another and collectively produce a complete set of the sequential patterns defined in the original sequence database.

Abstract: A method for recording a transaction log, and a database engine, and pertains to the field of database technologies. The method includes receiving an structured query language (SQL) request; modifying, according to the SQL request, a data page corresponding to the SQL request; generating a corresponding transaction log for the modifying; partitioning the transaction log to obtain a plurality of log segments; and writing the plurality of log segments into a plurality of log buffers, and writing the log segments in the plurality of log buffers into a log file in parallel. According to the present invention, not only lock granularities of the log buffers of the transaction log are reduced, but also a speed at which a log file is written into a magnetic disk is greatly increased, thereby improving performance of a database system and enhancing user experience.

Abstract: Techniques for resolving single source shortest path for database processing are provided. Graph data for nodes having weights for edges of a database network are iterated producing a new message table and results table in each iteration. The results table stores the minimum path weight. For each iteration the graph data and message table are joined to produce a state of a node and outgoing edge messages. The message table and results table are co-grouped to generate a new message for a next node to process. When done the single source shortest path for the network is produced.

Abstract: A method for recording a transaction log, and a database engine, and pertains to the field of database technologies. The method includes receiving an structured query language (SQL) request; modifying, according to the SQL request, a data page corresponding to the SQL request; generating a corresponding transaction log for the modifying; partitioning the transaction log to obtain a plurality of log segments; and writing the plurality of log segments into a plurality of log buffers, and writing the log segments in the plurality of log buffers into a log file in parallel. According to the present invention, not only lock granularities of the log buffers of the transaction log are reduced, but also a speed at which a log file is written into a magnetic disk is greatly increased, thereby improving performance of a database system and enhancing user experience.

Abstract: A database system lock operation method is provided that includes: acquiring, by a first processing node, a lock operand of a target lock; using, by the first processing node, the target lock to perform a lock acquiring operation on a target data object; when the lock operand meets a condition for a successful lock acquiring operation, determining, by the first processing node, that the lock acquiring operation is successful, and updating a value of a data bit that is in the lock operand and used to indicate a state of the target lock on the first processing node to a value that is used to indicate that the state of the target lock on the first processing node is lock held. Accordingly, the embodiments of the present invention further disclose a node device. The embodiments of the present invention can improve performance of a database system.

Abstract: The present invention relates to the technical field of railroad passenger car, and more particularly relates to a sleeper compartment of a railroad passenger car. The sleeper compartment may be an open type or a private booth type, comprising: sidewalls, partition walls, and two columns of longitudinally arranged sleepers that are parallel with each other; each column of the longitudinally arranged sleepers is formed by no less than two sleepers that are joined together, the sleepers comprise an upper-sleeper layer and a lower-sleeper layer; one partition wall is provided between adjacent sleepers; the length direction of the longitudinally arranged sleepers is consistent with the length direction of the railroad passenger car; the longitudinally arranged sleepers at two sides are fixed to the sidewalls respectively, and a corridor is provided between two adjacent columns of longitudinally arranged sleepers.

Abstract: The present invention relates to the technical field of railroad passenger car, and more particularly relates to a sleeper compartment of a railroad passenger car. The sleeper compartment may be an open type or a private booth type, comprising: sidewalls, partition walls, and two columns of longitudinally arranged sleepers that are parallel with each other; each column of the longitudinally arranged sleepers is formed by no less than two sleepers that are joined together, the sleepers comprise an upper-sleeper layer and a lower-sleeper layer; one partition wall is provided between adjacent sleepers; the length direction of the longitudinally arranged sleepers is consistent with the length direction of the railroad passenger car; the longitudinally arranged sleepers at two sides are fixed to the sidewalls respectively, and a corridor is provided between two adjacent columns of longitudinally arranged sleepers.