36 results about "B+ tree" patented technology

The invention discloses a video storage and retrieval method, device and equipment based on a B + tree, and relates to the field of data processing. In the method, a video data stream is divided into a plurality of preset space grids; recognizing a dynamic target which continuously moves in the video data stream; determining a target space grid where the dynamic target is located, and generating a spatio-temporal trajectory fragment; constructing a B + tree index structure; receiving a retrieval request, and executing range query by using the B + tree index structure and the compound key to obtain a candidate fragment set; traversing the target spatio-temporal trajectory fragments in the candidate fragment set, determining a corresponding target identifier based on the query dynamic target, and accumulating all overlapping durations generated by the query dynamic target to obtain a total effective stay duration; and screening out the query dynamic targets of which the total effective staying duration is greater than or equal to a preset duration threshold value, and outputting the query dynamic targets as a final retrieval result. By implementing the technical scheme provided by the invention, the retrieval efficiency of the video data is improved.

The invention relates to a data recovery method and device, computer equipment, a storage medium and a program product. The method comprises the steps that a current repairing task of a hybrid tree is determined, the hybrid tree is a hybrid storage structure of a log merging tree LSM tree and a B tree, the hybrid tree comprises multiple layers, the multiple layers comprise the first layer of the LSM tree and all layers of the B tree, ordered character string tables of the first layer of the LSM tree are merged and sunk into the B tree, and each node in the B tree stores at least one ordered character string table obtained through merging and sinking; under the condition that the current repair task indicates that the repaired damaged data storage object belongs to the B tree, obtaining a B tree index of the damaged data storage object from the current repair task; searching a damaged data storage object from the mixed tree according to the B tree index, wherein the damaged data storage object is an ordered character string table or a node; and under the condition that the damaged data storage object is found, repairing the damaged data storage object. By adopting the method, the data recovery efficiency can be improved.

The invention relates to a block backup indexing method, system and device based on a B + tree and a storage medium, and belongs to the technical field of disaster recovery backup. The method comprises the steps of intercepting a QEMU block write-in request and extracting an IO descriptor, writing block data into a disk and obtaining a physical offset, initializing a two-dimensional B + tree index structure, inserting the block IO descriptor and offset information into a two-dimensional B + tree, querying the position of the block data according to a logic offset and / or timestamp range, and deleting and updating index nodes according to a specified range. The system comprises a block data interception module, a block data writing module, a B + tree initialization module, a block data insertion module, a block data search module and a block data deletion module. According to the method, through a two-dimensional hierarchical index structure of the logic offset and the timestamp, joint efficient positioning of time and space dimensions is achieved, sharing of timestamp leaf nodes between local and global B + trees is supported, data redundancy is effectively avoided, and range query efficiency is improved.

The provided index server indexes two-dimensional (2D) location data included in the transaction data as a K-d-B tree, constructs partial Merkle trees for a group of leaf nodes in the K-d-B tree, and stores hierarchical hash information based on the partial Merkle trees in each of nodes of the K-d-B tree. When a query related to the location data is acquired via the interface module, the index server searches the K-d-B tree to find a query result corresponding to the acquired query, and generates verification data for verifying the searched query result on the basis of the partial Merkle trees and the hierarchical hash information. The index server then provides the query result and the verification data to the user terminal to verify the integrity of the query result using the partial Merkle trees and the hierarchical hash information by the user terminal.

The invention relates to the technical field of block chains, and provides an account state data concurrent updating method, electronic equipment and a computer program product. The method comprises the following steps: determining each to-be-updated account of which the state data needs to be updated; sorting the accounts to be updated according to a sequence rule of storing the account state data by each leaf node of the tree structure data; for each to-be-updated account, distributing a state updating operation of the to-be-updated account to an updating task of a to-be-updated leaf node corresponding to the sequence of the to-be-updated account in each leaf node; and concurrently executing the updating task of each to-be-updated leaf node so as to complete the state updating operation of each to-be-updated account. According to the method, the consistency of the updated account state data of each block chain node can be improved while the data updating efficiency is improved by adopting tree structures such as a B + tree and the like.

The application relates to a B+ tree index construction method based on ZNS SSD cold and hot data dynamic placement, which comprises a persistent memory PM, a dynamic random access memory DRAM and a ZNS SSD hard disk; the persistent memory PM and the dynamic random access memory DRAM are directly connected with a CPU memory bus and have the ability of being accessed by the CPU by bytes; internal nodes of the B+ tree are stored in the dynamic random access memory DRAM, wherein the internal nodes only store index information for positioning leaf nodes; pseudo leaf nodes of the B+ tree are stored in the persistent memory PM, wherein the pseudo leaf nodes only store key index information of the leaf nodes; leaf nodes for actually storing key value pairs in the B+ tree are divided into hot leaf nodes and cold leaf nodes according to data access frequencies, wherein the hot leaf nodes are stored in the persistent memory PM, and the cold leaf nodes are stored in the ZNS SSD hard disk. The application utilizes the DRAM to quickly position the leaf nodes, reduces query time, stores the hot leaf nodes in the PM to meet real-time business read-write requirements, stores the hot data in the high-speed DRAM and the PM, reduces random read-write delay and performs well in big data analysis and other scenes.

The application discloses a read-write performance optimization method based on a persistent memory B+ tree index and belongs to the field of data storage, and comprises the following steps: establishing a B+ tree, non-leaf nodes of the B+ tree are stored in DRAM, and leaf nodes are stored in PM; introducing a concurrent hash table for storing hot data on the DRAM, transferring the access of the hot data from the PM to the DRAM, and relieving the read-write interference problem on the PM; based on the characteristic that there are hot spots in the work load under a real scene, a lightweight hot data identification and replacement strategy is also implemented, a weight is set for each storage element based on the access frequency, and the dynamic change of the hot data in the work load is dealt with; a low-overhead write operation is realized by combining an additional update mechanism and a self-verification insertion mechanism in the B+ tree, the data consistency guarantee overhead is greatly reduced, and a semi-sequenced leaf node is adopted, so that the query efficiency of the leaf node is improved, and meanwhile, the write efficiency is not affected. The application can provide high performance for read operation and write operation simultaneously.

Metadata of a file system is maintained in a key value store. The metadata includes a namespace represented as a B+ tree having pages written to the key value store. Identifiers to the pages include a snapshot identifier and a page number. The identifiers are stored as keys in the key value store. The snapshot identifiers are used to track pages that are shared between first and second snapshots. Upon a request to write to an existing page, a determination is made from the snapshot identifiers as to whether the existing page belongs to the first or second snapshot. If the existing page belongs to the second snapshot, the write to the existing page is allowed. If the existing page belongs to the first snapshot, a copy on write (COW) is performed to generate a new page for the write.

A key value store is maintained in which the key value store includes pages of a B+ Tree structure holding file metadata of a logical partition of a filesystem. The pages are associated with keys through which the pages are retrieved from the key value store. The keys are named. A clone is created of the logical partition, the logical partition now being a parent and the clone now being a child. The naming of the keys is used to allow the pages stored in the key value store to be read by any of the parent or child. The naming of the keys is used to allow any of the parent or child to make updates to the pages, where updated pages are owned by a respective parent or child that made the updates, and pages not updated remain accessible by any of the parent or child.

A method and a device for data migration of a blockchain. The method includes generating a state tree based on state data of accounts in the blockchain, in which the state tree is a B+ tree based on a copy-on-write technology, and each leaf node of the state tree is configured to store the state data; determining a target state tree to be migrated from the generated state tree when data migration is triggered; sending state data of the target state tree to an extended node of a master node; if migration completion information is received from the extended node, updating a data mapping table stored in the master node, in which the data mapping table is configured to represent a mapping relationship between state data stored in the extended node and the extended node; and deleting the target state tree in the master node.

A video storage and retrieval method, device and equipment based on B+ tree, relate to the field of data processing. In the method, the video data stream is divided into a plurality of preset spatial grids; a dynamic target continuously moving in the video data stream is identified; a target spatial grid where the dynamic target is located is determined, and a space-time trajectory segment is generated; a B+ tree index structure is constructed; a retrieval request is received, and a range query is performed by using the B+ tree index structure and a composite key to obtain a candidate segment set; a target space-time trajectory segment in the candidate segment set is traversed, a corresponding target identifier is determined based on a query dynamic target, and all overlapping time lengths generated by the query dynamic target are accumulated to obtain a total effective stay time length; a query dynamic target with the total effective stay time length greater than or equal to a preset duration threshold is screened out as a final retrieval result output. The technical scheme provided in the application improves the retrieval efficiency of video data.

Embodiments of the present application provide a data page read-ahead method, an electronic device, and a storage medium. In the embodiments of the present application, a target node and sibling nodes thereof are located by means of parent nodes on various tree index structures such as a B+ tree or a B-tree, and a plurality of data pages of the target node and the sibling nodes thereof are read ahead into a memory in batches from a disk. Therefore, a new read-ahead mechanism for a tree index structure is provided, the I / O wait time of an I / O request for accessing a database system or a file system is greatly shortened, and the searching performance of the database system or the file system can be improved.

The invention belongs to the technical field of databases, and discloses a persistent memory-oriented load-aware dynamic mixed B + tree index structure. Comprising a dynamic hybrid B + tree and a load awareness prediction module. The dynamic hybrid B + tree is responsible for storage, query, insertion and deletion of all data, internal nodes are in pure ordered layout, and leaf nodes are in dynamic hybrid layout; the load sensing prediction module comprises a prediction model and a load sensing module; the load sensing module collects a read-write proportion, and when the load sensing module judges that the read proportion exceeds a dynamic threshold value, the prediction model is operated, and the capacity proportion of the ordered key value pair area and the disordered key value pair area of the leaf nodes in the dynamic mixed B + tree is updated. The core problem that the performance of an existing persistent memory index structure is seriously limited under the dynamic working load is solved, self-adaptive real-time adjustment of the ordered storage / unordered storage proportion at the node level is achieved, and the approximate optimal comprehensive throughput capacity is continuously maintained under the dynamic load.

A system can maintain a copy-on-write b+ tree for a data storage system that is configured to store data as chunks, and a mapping between tree page identifiers of the b+ tree and first in-use chunk identifiers. The system can perform reference count validation, comprising iterating over the mapping, wherein the iterating over the mapping comprises, based on a first tree page identifier of the tree page identifiers being present in the mapping, loading second in-use chunk identifiers that correspond to the first tree page identifier directly from the mapping; and based on the first tree page identifier of the tree page identifiers being absent, loading corresponding content, extracting third in-use chunk identifiers from object metadata of the content, and updating the mapping based on the third in-use chunk identifiers, to produce an updated mapping. The system can perform garbage collection of the data storage system based on the validation.

The present application provides a data management method and electronic device based on a B+ tree, relating to the field of data processing technology. The method comprises: in response to a first write operation on node data of a B+ tree, writing the node data to an original page of a disk; in response to a first write operation on a target entry in the original page, creating a first incremental record in an incremental page based on the original page; the first incremental record is used to record first data corresponding to the first write operation, and the first pointer of the target entry indicates first address information of the first incremental record; in response to a second write operation on the target entry, creating a second incremental record in the incremental page; the second incremental record is used to record second data corresponding to the second write operation; the first pointer of the target entry indicates second address information of the second incremental record, and the second pointer of the second incremental record indicates the first address information of the first incremental record. Embodiments of the present application can reduce disk space waste and lower read and write complexity.

This invention provides a method for reading data stored in a tree data structure, such as a B+ tree, using near data processing (NDP) in a cloud-native database. According to an embodiment, the expected LSN is used for NDP page reads on a primary compute node (e.g., a primary SQL node). When the primary compute node (e.g., a primary SQL node) reads a regular page, the maximum expected LSN of that regular page (e.g., the latest page version number) is used. The embodiment utilizes the expected LSN and page locking features, where the correct version of a page can be obtained by combining the expected LSN associated with the page with the page locking, and consistent tree structure readings can be achieved, resulting in good read / write concurrency.

The application belongs to the field of data storage, and particularly relates to a segmented B+ tree index structure based on PM. The HGB+ tree improves the performance of the index structure through a 'hourglass' leaf node structure and four mechanisms. The constructed HGB+ tree index structure is directly stored on the PM, and only the PM is used to complete the creation and use of the index structure. When the tree is created, the memory pool is first created or opened through the management library in the PM. Then, the necessary global metadata structure is created for the HGB+ tree in the PM, and subsequent operations are performed on the HGB+ tree. The internal nodes and leaf nodes of the HGB+ tree are deployed on the PM. The application designs a unique 'hourglass' leaf node structure scheme, and improves the performance of the storage system through the optimization of the basic operation strategy of the index structure and four beneficial mechanisms.

The present invention provides a method for batch sequential deletion of B+ trees, including the following steps: Divide a B+ tree into multiple ones; Define a linked list, and store the root node pointer and the maximum value of the tree into the linked list; Define a p_tree_head pointer for each B+ tree, and the p_tree_head pointer is used to mark the position of the first valid node after sequential deletion; When inserting and querying, first traverse the linked list to find the B+ tree to be inserted, and sequentially compare the maximum values of the trees in the linked list. If it is smaller than this value, query or insert into this B+ tree; For the deletion operation, find the p_tree_head pointer of the first node from the linked list, sequentially fetch the B+ trees backward, and then move the pointer backward; Until the key value corresponding to the p_tree_head pointer is the maximum value of this B+ tree; Delete the entire B+ tree and simultaneously delete the corresponding node in the linked list. On the premise of batch sequential deletion, the present invention can greatly improve the efficiency of the deletion operation.

This invention belongs to the field of database technology and discloses a load-aware dynamic hybrid B+ tree index structure for persistent memory. It includes a dynamic hybrid B+ tree and a load-aware prediction module. The dynamic hybrid B+ tree is responsible for all data storage, querying, insertion, and deletion. Its internal nodes have a purely ordered layout, while the leaf nodes have a dynamically hybrid layout. The load-aware prediction module includes a prediction model and a load-aware module. The load-aware module collects the read / write ratio. When the load-aware module determines that the read ratio exceeds a dynamic threshold, it runs the prediction model and updates the capacity ratio of ordered and unordered key-value pair areas in the leaf nodes of the dynamic hybrid B+ tree. This invention solves the core problem of the severely limited performance of existing persistent memory index structures under dynamic workloads, achieving adaptive real-time adjustment of the ordered / unordered storage ratio at the node level, and continuously maintaining near-optimal overall throughput under dynamic loads.

The application discloses an online incremental learning method and system based on adaptive B+ tree index, relates to the technical field of artificial intelligence, and comprises the following steps: constructing an extended adaptive B+ tree index structure, inserting a new sample and calculating a track disturbance value; combining the disturbance value and a node state to generate a disturbance score, and adding the new sample and associated node samples to an incremental learning scheduling queue when a condition is met; obtaining an affected node and updating the structure code of the affected node when a node splitting, merging or reconstructing event occurs; monitoring sample label track changes, adjusting a label confidence marker when a threshold is exceeded, and limiting the participation frequency according to a screening mechanism; and performing a forgetting compression operation on a leaf node with low access frequency and small label fluctuation degree. Through the construction of the adaptive B+ tree index, the disturbance-driven scheduling, the structure change response and the forgetting compression strategy, the sample management is refined, the learning update is efficient, and the structure maintenance is controllable, so that the stability and resource utilization efficiency of the online incremental learning are improved.

The application provides a non-unique index structure updating method and device based on PostgreSQL. The method comprises the following steps: obtaining a non-unique index, wherein the data organization structure of the non-unique index is a B+ tree structure, the leaf node in the B+ tree comprises a TID pointer structure pointing to a data tuple, the TID pointer structure is provided with a count value representing the number of repeated key values and an address list; and in response to an updating operation on the non-unique index, the count value and the address list are synchronously updated. The application optimizes the index structure under the premise of maintaining the B+ tree, combines the repeated key value rows in the leaf tuple, creates an address list for each group of key values, stores the physical addresses of each row of the repeated key values, saves the storage space of the non-unique index, and improves the scanning efficiency of the non-unique index.

The invention relates to a B + tree merging method and device, equipment, a storage medium and a program product. The method comprises the steps of obtaining at least two to-be-merged B + trees; key value pairs are extracted from the first leaf nodes of the at least two B + trees to be combined, and the key value pairs are sorted according to the sequence of values of keys from small to large; and sequentially writing the key value pairs into a second leaf node of the combined B + tree based on a pre-acquired structure parameter, and generating an upper layer node of the second leaf node to obtain the combined B + tree. According to the method, the key value pair is extracted from the first leaf node of the to-be-merged B + tree and written into the second leaf node of the merged B + tree, and the upper layer node of the second leaf node is generated to obtain the complete merged B + tree, so that when data on a plurality of B + trees are integrated, the data on the plurality of B + trees can be integrated; the change of a large number of nodes and tree depth caused by inserting the nodes into the B + tree in batches is avoided, the consumption of computing resources is reduced, and meanwhile, the processing efficiency is improved.

A system for searching for pages of a memory device based on a B+ tree node structure. The system includes a memory device and a controller including a cache memory and a search accelerator. The search accelerator receives a key-value pair associated with a target page among a plurality of pages; and searches for the target page from the cache memory based on the key-value pair, using a tree structure including B+ tree nodes mapped to the plurality of pages. The key-value pair includes a searched key and a value indicating a pointer to a tree node mapped to the target page, among the B+ tree nodes.

The B+ tree-based dynamic adaptive multi-window tiling method belongs to the field of graphical user interface and human-computer interaction technology. To address the poor user experience caused by multiple windows crowding each other in tiling windows, the key is that when the window manager receives a command to open or close a window, if the number of windows to be displayed in the available display area is greater than the first maximum number of windows but not greater than the second maximum number of windows, the window manager sets the current tree-structure tiling window management mode to the second tree-structure tiling window management mode. If the number of windows to be displayed in the available display area is greater than the second maximum number of windows, the window manager determines the windows to be closed based on the number of user clicks and the most recent usage time, and closes them, setting the current tree-structure tiling window management mode to the second tree-structure tiling window management mode. This achieves efficient management and adaptive layout of multiple windows.

A B+-tree-based data management method includes writing, when a first write operation on a node data of a B+ tree is detected, the node data to a base page of a disk, wherein the base page contains a pointer field configured to indicate an associated delta page; creating a first delta record in a delta page based on the base page, when a first write operation on a target entry in the base page is detected; creating a second delta record in the delta page when a second write operation on the target entry is detected, the second delta record is used to record second data corresponding to the second write operation. Accordingly, a waste of disk space may be reduced, and complexity of read-write operations may be lowered.

The present application relates to the technical field of storage. Specifically disclosed are a data storage method, apparatus and device, and a computer-readable non-volatile storage medium. On the basis of the size of a data block of a device in which a B+ tree is located and the size of a unit of data to be stored, the order of the B+ tree to be created is determined, so as to call a B+ tree creation script to create a B+ tree of a reasonable order. The utilization of node resources of the B+ tree is monitored, and when the B+ tree meets a node splitting condition, whether a sibling node of a node to be split meets a migration condition for receiving data of the node to be split is queried first. If the migration condition is met, the B+ tree creation script is prohibited from executing a node splitting task, and the data of the node to be split is migrated to the sibling node; and if the migration condition is not met, the B+ tree is allowed to execute the node splitting task, such that the utilization rate of node resources stored in the B+ tree is improved by means of delaying the node splitting of the B+ tree and preventing unnecessary node splitting of the B+ tree, thereby improving the utilization rate of storage resources when the B+ tree is used for data storage.

The application relates to a B+ tree-based block backup index method, system, device and storage medium, and belongs to the technical field of disaster recovery backup. The method comprises the following steps: intercepting a QEMU block write request and extracting an IO descriptor, writing block data into a disk and obtaining a physical offset, initializing a two-dimensional B+ tree index structure, inserting the block IO descriptor and offset information into the two-dimensional B+ tree, querying a block data position according to a logical offset and / or a timestamp range, and deleting and updating an index node according to a specified range. The system comprises a block data interception module, a block data writing module, a B+ tree initialization module, a block data insertion module, a block data searching module and a block data deleting module. Through the two-dimensional hierarchical index structure of the logical offset and the timestamp, the application realizes joint efficient positioning in the time and space dimensions, supports sharing of the timestamp leaf node between the local and global B+ trees, effectively avoids data redundancy, and improves the range query efficiency.

The embodiment of the invention provides a data page pre-reading method, electronic equipment and a storage medium, in the embodiment of the invention, a target node and brother nodes thereof are found through father nodes on various tree index structures such as a B + tree or a B tree, and a plurality of data pages of the target node and the brother nodes thereof are pre-read from a disk into a memory in batches, so that the data page pre-reading efficiency is improved, and the data page pre-reading efficiency is improved. A new pre-reading mechanism aiming at the tree index structure is provided, the I / O waiting time of an access I / O request for accessing a database system or a file system is greatly shortened, and the search performance of the database system or the file system can be improved.

A system can maintain a copy-on-write b+ tree for a data storage system that is configured to store data as chunks, and a mapping between tree page identifiers of the b+ tree and first in-use chunk identifiers. The system can perform reference count validation, comprising iterating over the mapping, wherein the iterating over the mapping comprises, based on a first tree page identifier of the tree page identifiers being present in the mapping, loading second in-use chunk identifiers that correspond to the first tree page identifier directly from the mapping; and based on the first tree page identifier of the tree page identifiers being absent, loading corresponding content, extracting third in-use chunk identifiers from object metadata of the content, and updating the mapping based on the third in-use chunk identifiers, to produce an updated mapping. The system can perform garbage collection of the data storage system based on the validation.