148 results about "Cache hit rate" patented technology

The invention discloses a storage and calculation separation key value storage delay optimization method based on prediction driving, and aims to solve the problems of performance bottleneck and high tail delay caused by passive management and high network delay in a key value storage system in a storage and calculation separation scene. The method comprises the following steps: time load prediction: deploying a time sequence prediction model at a client, and predicting a future read-write request based on a historical access sequence; active cache prefetching: according to the predicted read request, actively preloading hotspot data from a server side to a client side for caching so as to improve the cache hit rate and hide network delay; active write-in optimization: according to the predicted write request, executing maintenance at a server side through predictive pre-insertion and active node splitting, and moving high index structure adjustment overhead out of a key request path to eliminate a write delay peak; and structure sensing batch synchronization: pre-fragmenting a local write buffer by using a server index top layer model of a client cache, and combining multiple independent remote insertion operations into one efficient batch update to reduce data synchronization overhead. Compared with an existing passive management system, the characteristics of model prediction and active cooperation are fully utilized, and the average delay and the tail delay of the system are reduced.

The invention discloses a multi-core scheduling system and method based on interrupt affinity and a storage medium. The technical problem that data locality guarantee and load balancing are difficult to consider at the same time is solved by constructing a self-adaptive optimization closed loop of perception-decision-execution-feedback, and the method comprises the steps that a routing interruption configuration mechanism provides a basis for dynamic adjustment; an interrupt guide task placement mechanism ensures that an interrupt service program and an associated task are executed in the same CPU core, and the cache hit rate is increased; the execution overhead of an interrupt service program is brought into core load statistics through task context switching and a'task + ISR 'two-dimensional precise load evaluation mechanism triggered by ISR inlet / outlet double nodes, and real load awareness is achieved; a load-driven dynamic interrupt routing mechanism is combined with dual control of a hysteresis threshold and cooling time, so that a ping-pong effect is avoided. According to the method, dynamic load balancing is realized on the premise of keeping data locality, and the real-time response performance and throughput of the multi-core system are remarkably improved.

The invention discloses a front-end cache management method, system and device for a lightweight large-model dialogue state and a medium, belongs to the technical field of front-end cache management of a large-model dialogue system, and aims at solving the technical problem of how to overcome the defects that in a traditional scheme, long context cache is low in efficiency, storage redundancy and insufficient in dynamic semantic adaptation capacity, and the large-model dialogue state cannot be managed easily. In order to realize dialogue context volume compression, improve semantic similar request hit rate and reduce cross-end synchronization delay, the adopted technical scheme is as follows: data acquisition and preprocessing: capturing user interaction behaviors in real time through front-end burying points, and performing preprocessing operation on the acquired user behavior data; semantic normalization processing: performing embedded vector conversion and semantic clustering on the text input by the user to generate a unique semantic identifier and a context vector; querying and updating the multi-level cache; and dynamic collaborative updating: dynamically adjusting the cache based on the cache hit rate, the response delay and the user feedback, and optimizing the cache effect in real time.

The invention discloses a private cloud platform-based micro-service application remote main and standby disaster recovery rapid switching method, relates to the technical field of computers, and is used for solving the problems of long remote main and standby disaster recovery switching time, cold start storm, insufficient capacity evaluation and discharge strategies and complex back switching. The switching process is decomposed into controllable steps by combining real-time state feedback through a clear sequential arrangement and batch reconnection strategy of a data surface and a control surface, so that the switching time is remarkably shortened, the target of minute-level continuous availability is met, and meanwhile, the switching efficiency is greatly improved through heat set construction and standby side quasi real-time preheating before switching. Core hotspot data and indexes are loaded to the standby side in advance, it is ensured that the standby side has the high cache hit rate and index availability during master raising, and therefore the performance bottleneck caused by access of a large amount of cold data is avoided.

The invention belongs to the technical field of computers and artificial intelligence, and discloses a priority differential cache-based MoE batch reasoning method and system, and the method comprises the steps: recognizing global hot experts with higher activation frequency in each layer, enabling the global hot experts to remain in a high-priority cache during initialization, and building a priority list of the experts in each layer; in the reasoning stage, selecting an expert set needing to be activated in the current layer; the expert priority score is dynamically updated according to the activation and reuse behaviors of the experts, and the experts are migrated and replaced among the high-level cache, the middle-level cache and the low-level cache according to the priority score and a cache threshold value; according to a current layer gating result and a historical activation track, predicting an expert set which may be activated in a next layer, and asynchronously loading the expert set to a low-priority cache during current layer calculation so as to realize time overlapping of calculation and communication. According to the method, the expert migration frequency can be reduced, the cache hit rate and the throughput rate are improved, the video memory utilization rate is remarkably optimized, and good expansibility and universality are achieved.

The invention discloses a performance monitoring system and method for expansion and upgrading of a big data platform, and relates to the technical field of big data analysis. According to the performance monitoring system and method, key information of newly-added Web servers is collected to form a standardized configuration data set, and node deployment normalization and cluster compatibility in the expansion stage are improved; operating load logs before capacity expansion are directionally pulled and screened, and a historical cluster server operating load set is constructed through structural analysis; first dynamic weight configuration is completed by combining two types of load sets, the cache types of the Web servers are distinguished according to cache configuration identifiers, and secondary weight adjustment is performed on the shared distributed cache Web servers based on the hit rate, so that the defects of existing weight configuration are overcome, accurate weight adaptation is realized, the cache hit rate is increased, the frequency of requesting to penetrate through a database is reduced, and the service life of the Web servers is prolonged. The database load is reduced, the response delay is shortened, resource imbalance and service instability are avoided, and the performance of platform expansion upgrading operation is guaranteed.

The invention relates to the technical field of database query optimization and distributed system data processing, and discloses an intelligent routing optimization method and system for multi-table association query, and the method comprises the steps: a query topology analysis unit analyzes an object relation mapping configuration file to construct a query topological graph; the environment parameter detection unit collects real-time state data of the server cluster based on the query topological graph, wherein the real-time state data comprises a pre-estimated cache hit rate, associated sparseness and a target table data line number estimation value; the routing strategy decision-making unit compares the real-time state data with a preset threshold value, and selects a cache-driven step-by-step query strategy, a column-type compression transmission strategy or an asynchronous intermediate table unloading strategy; and the dynamic execution control unit executes data acquisition and assembly operation according to a selected strategy. According to the method, the problems of low database computing resource utilization rate and application server memory overflow caused by the adoption of a static strategy in a traditional framework are solved, and the system response speed and the service availability of multi-table association query are effectively improved.

The invention relates to the technical field of server testing, in particular to a server multi-core computing processor load optimization testing method, device and equipment and a medium, and the method comprises the following steps: automatically identifying hardware configuration information of a server; generating a multi-dimensional pressure test plan based on the hardware configuration information; executing the test plan, and monitoring the load rate, the memory bandwidth, the cache hit rate and the I / O performance index of each CPU core in real time; dynamically adjusting a task allocation strategy based on a set load difference threshold to realize task dynamic optimization, and if the difference between the maximum core load rate and the minimum core load rate is monitored to exceed the threshold, triggering a task migration operation; system performance data are collected again, performance indexes before and after optimization are compared and analyzed, and system performance bottlenecks are recognized; and automatically generating a performance test report. The utilization rate and the overall task throughput of the multi-core processor are improved, and the problem that a static task allocation strategy is difficult to adapt to dynamic load changes is solved.

The invention discloses a data warehouse intelligent construction and management method and system, and belongs to the technical field of data warehouse construction and management, and the data warehouse intelligent construction and management method comprises the steps: obtaining an enterprise data source, and extracting the feature parameters of the data source; constructing a data warehouse logic model; self-adaptive ETL configuration parameters are defined; constructing a three-layer physical architecture comprising a data access layer, a data storage layer and a data service layer; intelligent cache strategy configuration is executed based on a deep learning algorithm, a three-level cache architecture is established, and a pre-cache rule and a failure strategy are defined; and running management optimization is implemented through a multi-objective optimization function. According to the method, the construction automation level is remarkably improved, the cache hit rate is increased, the average query response time is shortened, the data consistency and the system performance are effectively balanced, the elastic adaptive capacity of the system is enhanced, the operation and maintenance complexity and cost are reduced, and meanwhile efficient sharing of cache resources is achieved on the premise that data security isolation is guaranteed.

The invention discloses an optimization method for improving knowledge base image data caching processing, a platform can dynamically adjust a caching strategy according to a data access mode and access time change by utilizing a self-adaptive caching strategy combined with access time change and through replacement between two levels of caching strategies, so that the hit rate of caching is improved, and the caching efficiency is improved. According to the method, the access to a back-end database is reduced, the effective utilization of cache resources is ensured, and the platform can respond to the change of a user access mode in real time by dynamically adjusting a cache strategy, so that smoother and more efficient online transaction experience is provided; wherein the first-level cache stores frequently-used and frequently-accessed data, it is ensured that the data can be rapidly accessed, the time for obtaining the data from a rear-end database is shortened, the response speed of the system is improved, meanwhile, the second-level cache serves as a supplement of the first-level cache, more kinds of data are stored, and the coverage rate and efficiency of the cache are further improved.

This invention discloses a cache adjustment method, apparatus, device, and computer-readable storage medium, comprising: collecting access characteristic data from the client layer, object storage daemon layer, and device layer respectively; determining the object access mode of the system based on the access characteristic data; and adjusting the cache priority of each access task, the cache capacity ratio between the client layer and the object storage daemon layer, and the data residence time in the cache layer according to the object access mode; wherein the cache layer includes the client layer and the object storage daemon layer. This invention improves cache hit rate and overall read / write throughput performance, and reduces access latency.

The application discloses a kind of blockchain data traceability query optimization methods, by introducing the method of cache optimization, utilize cache to reduce the number of disk IO in traceability, improve the efficiency of traceability search, while multi-level cache structure is designed, the problem that cache hit rate is not high under the condition that the memory resource of full node is limited is solved, that is, while giving consideration to the consumption of memory resource in the improvement of cache hit rate, reduce the burden of full node.A kind of blockchain data traceability query optimization method first by the node in network to full node of blockchain initiates traceability query request, then in full node query cache, finally carries out consistency check to full node, realizes blockchain data traceability query.The application improves the traceability query efficiency of blockchain system, query credibility, with certain practicality.

The invention discloses an internet of things edge computing dynamic load prediction data processing method and system, and belongs to the technical field of edge computing. According to the method, edge node operation, network link quality and Internet of Things task characteristic parameters are collected, a time sequence prediction model is adopted to complete edge node load prediction, and related parameters are integrated to construct a task dynamic decision state space; computing task routing allocation is completed based on the state space, and a heterogeneous resource scheduling multi-objective optimization model is established and solved to obtain a scheduling scheme; according to the scheme, the original data of the Internet of Things is subjected to mixed compression, and a prefetching rule is constructed by combining compressed data access statistics and time-space association attributes, so that data prefetching and cache management are completed. According to the method, dynamic adaptation of the computing tasks and reasonable utilization of resources can be realized, data transmission bandwidth occupation is reduced, the cache hit rate of the edge nodes and the system operation stability are improved, and the real-time processing requirement of the edge computing scene of the Internet of Things is met.

The application discloses a kind of high concurrency time series data processing method and device for anti-UAV system.The application uses lock-free hash table to cache time series data in cache space, lock-free cache design and CAS atomic operation reduce lock overhead, batch write throughput is high, conflict probability is low, support multi-thread high concurrency read-write, completely eliminate thread competition and blocking caused by traditional cache mechanism, improve concurrent processing capability;At the same time, batch write reduces the overhead of dynamic memory allocation, improves memory utilization;Through composite key fast positioning, hot data index structure priority query and lock-free reading mechanism, the query delay of hot data is extremely low, data preloading mechanism further reduces the delay of future related query, improves real-time response capability;Based on the dynamic elimination mechanism of elimination score, hot data / recent data can be preferentially retained, ensure resource utilization, significantly improve cache hit rate, greatly reduce access frequency to TSDB.

The application provides a read cache optimization method and device in an AI model training scene, comprising: during AI model training, setting each Epoch reading training set data process as two rounds of iterations; in the first round of iteration, each Mini-Batch queries S*N items of data to the cache in random order; S items of data are read from the data in the cache after the query for the current Mini-Batch, and the remaining data is marked; after all the training set data is traversed once, the first round of iteration ends, the second round of iteration starts, and the marked data in the first round of iteration is traversed; each Mini-Batch queries S items of data to the cache, the data is directly read in the cache, or the data is read from a remote storage node. The application improves the read cache hit rate when loading the training set, thereby reducing the time delay of reading data, improving the utilization rate of GPU and other scarce computing power devices, saving computing power, and improving the model training efficiency.

The application discloses a Tensor data processing method and device and related equipment. The method comprises the following steps: acquiring prior information of Tensor data, wherein the prior information comprises a dimension parameter and a storage address mapping rule of the Tensor data; determining each target element based on the dimension parameter of the Tensor data; converting the virtual storage address of each target element into a physical storage address by using the storage address mapping rule, wherein the physical storage addresses are continuous; storing each target element into the corresponding physical storage address; and performing a preset operation on each target element stored in the physical storage address by using a preset algorithm to obtain a processing result. Thus, the cache hit rate is improved.

This application provides a thread scheduling method, apparatus, electronic device, and storage medium, relating to the field of hardware testing technology. The specific solution is as follows: Obtaining the importance representation information of each target LLC to a target process, where the target LLC is the LLC occupied by threads running in the target process; determining the recommended LLC corresponding to the target process from the obtained importance representation information; in response to scheduling the target threads in the target process, determining the CPU core based on the load state of the CPU cores sharing the recommended LLC, and scheduling the target threads to the determined CPU cores for execution. Applying the solution provided in this application can improve the cache hit rate for scheduling threads in the same process to CPU cores sharing the same LLC and improve the operating efficiency of multi-core CPUs.

This invention relates to the field of bioinformatics processing and medical laboratory data analysis, specifically a method for classifying hepatitis B antibody patterns based on plasma protein profiles. The method includes: acquiring host hardware information of the execution environment and extracting central processing unit (CPU) cache parameters; acquiring a high-dimensional sparse one-dimensional array and extracting non-zero feature indices; performing mapping calculations using a locality-sensitive hashing (LSH) algorithm to reconstruct the data into a locally dense two-dimensional matrix; dynamically segmenting the data into independent sub-blocks according to cache parameters and initial segmentation dimensions; performing low-rank tensor decomposition on the independent sub-blocks to extract local latent feature vectors; concatenating the sub-blocks, weighting them through a single-layer attention network, and inputting the result into a classifier function to output a target classification pattern vector; and generating dimension update instructions based on a preset dynamic adjustment mechanism to adjust subsequent segmentation dimensions. This invention achieves lower memory peaks, higher cache hit rates, and stable multi-label pattern output capabilities.

The invention discloses a cache optimization method and device based on distributed storage, and relates to the technical field of distributed storage systems. The method comprises the following steps: dynamically calculating a weight factor of a cache file, attenuating the weight factor based on the access frequency and timeliness of the file, and compensating and adjusting according to the size of the file; the maximum file threshold value set by the system is dynamically adjusted according to the size distribution of the files in the cache, and the maximum file threshold value is determined based on a preset quantile range so as to eliminate interference of extremely large files on weight calculation; when the weight factor of the cache file is lower than a preset threshold value, the file is removed from the cache, and the cache space is released; when the cache capacity is insufficient, the file with the minimum weight is selected for replacement based on the weight factor of the current cache file, and the high-value file is preferentially reserved in the cache. According to the method, the cache hit rate of small files in the distributed storage system is effectively increased, cache space waste and metadata operation overhead are reduced, and system performance and resource utilization rate are remarkably improved.

The present application belongs to the technical field of computers. Disclosed are a cache management method and apparatus. The method is applied to a cache of a computer system, and the computer system further comprises a processor in communication with the cache. The method comprises: grouping memory access requests from a processor; applying different allocate policies to two groups of memory access requests; comparing performance costs consumed by processing the two groups of memory access requests after the corresponding allocate policies are applied to the two groups of memory access requests; and on this basis, determining a target allocate policy for the remaining groups of memory access requests other than the two groups of memory access requests. By means of the method, a better allocate policy can be determined for the remaining groups of memory access requests, and the better allocate policy can achieve caching of data that is highly likely to be accessed again by the processor in the future, thereby improving subsequent cache hit rates of memory access requests in the remaining groups, that is, improving the performance of the processor.

The invention provides a CPU end DLRM reasoning dynamic optimization method based on load pre-calculation, and belongs to the technical field of deep learning and computer system optimization. Comprising the following steps: step 1, dynamic core distribution: S1, constructing a heterogeneous load quantification model of an embedded search task and a Bottom-MLP task, S2, carrying out load sampling and dynamic adjustment, and S3, modifying a deep learning framework thread pool mechanism, and distributing and binding a CPU physical core for the two types of tasks to realize parallel scheduling; and step 2, dynamic prefetching: T1, determining timeliness matching and accurate prefetching principles and defining a prefetching distance as a key parameter according to embedded search three-level indirect addressing characteristics, T2, recording an iteration period, separating and aggregating calculation time and DRAM access time in an embed dingbag operator core cycle, and T3, dynamically calculating the prefetching distance and inserting a prefetching instruction to execute prefetching. According to the method, the problems of unbalanced heterogeneous task core distribution and irregular memory access prefetching failure are effectively solved, the DLRM reasoning end-to-end delay is reduced, and the CPU resource utilization rate and the cache hit rate are improved.

The invention discloses a data multi-level cache collaborative acceleration method and system, and mainly relates to the technical field of data cache acceleration. Comprising the steps of obtaining data access features and inputting a model fusing sliding window statistics and exponential decay to calculate a dynamic heat value; dynamically scheduling the data to a high-speed memory, a distributed sharing or a local file cache level according to the heat value; optimizing a cross-level query path based on the intelligent routing model; the multi-level cache consistency is maintained by adopting an asynchronous version control mechanism; executing hierarchical collaborative cleaning and elastic capacity expansion according to the real-time popularity and the capacity utilization rate; and realizing hot data preloading by utilizing the prediction model. The method has the beneficial effects that through intelligent scheduling and collaborative optimization, the cache hit rate and the system throughput are remarkably improved, the access delay and the storage cost are reduced, and the stability and the data consistency under high concurrency are enhanced.

The invention discloses a fair resource management method and system for multi-user shared large language model reasoning, and the method is characterized in that the method employs a cold data recognition elimination mechanism and a fair cache distribution mechanism, carries out the unified quantification of the resource consumption of a request in GPU calculation, decoding and cache transmission, and carries out the fair scheduling based on the accumulated CPI value of each user; the system comprises a user request management module, a CPI fair scheduling module, a model reasoning execution module, a resource monitoring module and a cache management module. Compared with the prior art, the method has the advantages that users can share computing and caching resources fairly, efficient operation performance of the system is guaranteed, users with high cache hit rate are prevented from excessively occupying GPU execution opportunities due to low apparent computing overhead, the overall reasoning efficiency and the cache hit rate of the system are remarkably improved, and the user experience is improved. Unified and fair distribution of computing resources and cache resources is achieved, the effect is prominent especially in a multi-tenant high-load scene, and the method has good application prospects and commercial development value.

The application relates to the field of computer architecture and parallel computing technology, and discloses an address mapping method and system for GPU convolution acceleration, which is applied to the process of performing convolution operation based on general matrix multiplication by a GPU, dynamically remaps the memory access address of a workspace matrix, and comprises the following steps: identifying a data repetition mode in the workspace matrix; based on the data repetition mode, remapping the original address of a current memory access request to a target address by using an address mapping strategy; performing a data access operation by using the target address; if the target data is cached in a cache level of the GPU, directly returning the same value in the original access address; otherwise, loading data from a global memory and storage. The application avoids frequent global memory and storage access caused by scattered storage of repeated data in a traditional method, significantly improves the cache hit rate, and further breaks through the bottleneck of low cache utilization of an existing strategy.

The invention discloses a cache key generation method, an icon path acquisition method and electronic equipment, and the method comprises the steps: acquiring an icon information file, and extracting an application package name, a version number, an icon path and a priority type from the icon information file; and generating a composite cache key by using the application package name, the version number, the icon path and the priority type. According to the method, a simple file path is not used as a cache key any more, but the composite cache key is constructed by using the application package name, the version number, the icon path and the priority type, so that the cache hit rate of the icon is increased, and APK icon analysis cache can be executed efficiently, accurately and monitorably in a server environment.

The application discloses a kind of dynamic resource adaptation multilayer cache architecture systems, it is related to computer system architecture and data storage cache technical field.The system includes: resource state perception module, for real-time acquisition system performance index data and load characteristics in running, performance index data includes the access delay of each cache level, cache hit rate, data access frequency, cache medium read-write speed and system concurrent connection quantity.The application is through resource state perception module comprehensive acquisition each cache level performance index and load characteristics, combined with the deep reinforcement learning algorithm of cache strategy decision module dynamically generates cache strategy parameter, solves the problem that static configuration is difficult to adapt complex load in prior art;Data layout optimization module is through hot data identification, hot and cold data asynchronous migration and cross-level data block size adaptation, accurately match the characteristics of heterogeneous cache medium, make up the defect that traditional architecture is not enough to perceive medium difference.

The invention relates to an edge node collaborative scheduling and response method based on a clustering algorithm, and the method comprises the steps: analyzing a network topology structure and the load information of nodes through the clustering algorithm, and gathering the nodes with similar demands, similar access modes and similar loads into a same collaborative domain. In each cooperative domain, the nodes schedule and update the cache content by sharing cache resources, intercommunicating data and collaboratively optimizing updating and replacing strategies of the cache content in combination with real-time access requirements in the region, and ensure that the high-frequency access content can be preferentially stored in the node cache, so that the cache hit rate is increased, and the network delay is reduced. Through the mechanism, the nodes in the cooperative domain can efficiently share cache resources, so that the cross-region transmission is reduced, the burden of the central node is reduced, the flow pressure borne by the central node is reduced, and the network bottleneck is relieved.

This application provides a method, system, device, and medium for virtual terminal resource preloading and caching management based on the ISOBUS standard. This application innovatively constructs a multi-level caching architecture and combines a Markov chain-based resource prediction engine with adaptive memory management technology to achieve intelligent preloading and efficient caching of interface resources. Addressing the pain points of existing ISOBUS VT systems, such as high resource loading latency, low cache hit rate, and insufficient memory utilization, this invention provides a targeted technical solution. Experimental data verifies that this technical solution can reduce resource loading time by 50% to 70%, increase cache hit rate to over 92%, improve memory utilization by 40%, and reduce system crash rate by over 80%.

The present disclosure provides a blockchain contract processing method and device, electronic equipment and storage medium, comprising: loading N candidate smart contracts in batches from the to-be-processed smart contracts; in response to identifying that the candidate smart contract needs to perform a target external data read-write operation, judging whether the target external data exists in the local cache; in response to the target external data not existing in the local cache, performing saving processing on the first execution context data of the candidate smart contract, and determining a target contract from the to-be-processed smart contracts except the N candidate smart contracts; replacing the target contract with the candidate smart contract, and loading the target contract to the execution slot of the execution memory corresponding to the candidate smart contract, and executing the target contract. When it is found that the candidate smart contract needs to wait for the database to return, the context is saved and the execution is switched, avoiding the idling of the execution unit, improving the hardware utilization rate, and at the same time, by supporting batch preloading, the cache hit rate and execution continuity are improved.

The application provides a method, system, device and storage medium for quickly reading metadata, the method comprising: in response to receiving a metadata request for accessing a first file by a service, a metadata server only reading metadata information of the first file; responding to the metadata request issued by a client according to the read metadata information of the first file; asynchronously issuing the metadata request to a metadata disk and batch loading metadata information of other files in a directory where the first file is located; and in response to again receiving metadata information of a target file in the directory where the first file is located, directly reading metadata information of the target file from the metadata disk. The application improves the metadata cache reading performance of a distributed file storage system in a sequential reading scenario and a random reading scenario. The stored metadata cache space is more efficiently utilized, and the cache hit rate is improved.