791results about How to "Reduce performance loss" patented technology

A processing method for power down and abnormal writing of a solid state disc comprises the steps of step 100, scanning a flash array, obtaining stored management information, performing initialization after extracting the management information, generating a plurality of mapping areas, dynamically adjusting the mapping areas, and managing the conversion between a multi level cell (MLC) to a single level cell (SLC); step 200, performing data reading and writing, during reading, finding a corresponding physical address through the two mapping areas, and reading the data on a flash memory; step 300, when the data is written, if finding a first mapping area already has no spare low level page, then starting waste page recovery management and abrasion balance management, after the abrasion balance management preferably recovers the waste page according to information in a mapping list of the first mapping area and then distributes the spare page, writing the data into the first mapping area; and step 400, when the conditions of power down or abnormal writing happen, enabling the plurality of mapping areas to perform corresponding operations to recover the data.

A software method for obtaining an address of a thread-local variable while executing a computer program is described. The method designates a memory location to store an address of a function. The method, after the computer program starts to run, obtains the address of the thread-local variable by executing a resolver function whose address is stored in the memory location when a module that references the thread-local variable is loaded. The resolver can be a lazy resolver function or other resolver functions optimized for specific modules or variables.

The invention provides a high-performance reliable solid-state disk realizing method. The method includes: (1), dividing all flash memory chips in a solid-state disk into groups and forming an RAID (redundant array of independent disks) 4-level flash array in each group by N successive flash chips; (2), receiving and storing data through a cache; (3), judging whether the cache is filled up or not, if yes, entering the step (4), and if not, returning to the step (2); (4), extracting N-1 data blocks from the cache and computing check values of the N-1 data blocks; making up the N-1 data blocks and the check values into filled stripe data and writing back the flash array; returning to the step (2). The flash chips in the solid-state disk are used for establishing the RAID4-level physical array to assure data reliability. Faults at different levels including page level, block level, or even chip level can be processed. Besides, writing-in performance is improved by writing of filled stripes and sequence, and spatial and performance loss resulted from data errors can be reduced to the utmost.

The invention discloses a direction-of-arrival estimation method based on sample covariance matrix sparsity, and relates to the field of array signal processing. The method comprises the steps that firstly, dimensionality reduction is carried out on a covariance matrix, receiving signals, of a radar antenna array, and the covariance matrix, receiving the signals, of the radar antenna array is obtained after dimensionality reduction; secondly, a sparse vector cost function based on sparse constraints is built according to the covariance matrix after dimensionality reduction; thirdly, the sparse vector cost function based on the sparse constraints is constructed to be the mode suitable for convex programming package solving, and a sparse vector is solved according to the mode of convex programming package solving; fourthly, a non-zero element in the sparse vector is determined as a target azimuth angle, and the target azimuth angle is a target arrival direction. The method mainly solves the problems that in the prior art, known noise power is needed, and the calculated amount is large, and the method is mainly used for scenes of array signal processing.

The invention discloses a method and a device for server task scheduling. According to the type of tasks, the tasks are distributed to a pre-bound task group, and a preset task scheduling algorithm corresponding to the pre-bound task group is used for scheduling the tasks, wherein the pre-bound task group comprises a first preset quantity of CPUs (Central Processing Units), and is bound with a memory bank which contains a preset second quantity of memory nodes. Different types of tasks are distributed to different task groups, the task group contains a corresponding quantity of CPUs and memory nodes, and the CPUs and the memory nodes are grouped to execute of different types of tasks; and when task scheduling is carried out, the characteristics of the task are considered, and the hardware characteristics of the CPU and the memory node of the server are also considered. During task scheduling, the hardware characteristics of the server are considered, and performance loss brought by task switching can be reduced so as to improve the resource use ratio of the server and improve server performance. Obviously, the method is favorable for improving the server performance.

The invention discloses a method for preparing a 3D printing material by use of recycled waste high-density polyethylene plastics. The raw material adopted by method is the waste high-density polyethylene plastics which are prepared from the following components in parts by weight: 100 parts of high-density polyethylene, 10-40 parts of jute fibers, 10-30 parts of calcium lignosulfonate, 10-30 parts of calciumsilicate, 0.8-1.4 parts of a coupling agent, 10 parts of a compatilizer and 20-30 parts of a plasticizer. The method comprises the following steps: 1, breaking and cleaning the waste plastics; 2, preprocessing the jute fibers; 3, adding additives, and grinding at low temperature; 4, extruding and pelletizing. According to the method, the waste high-density polyethylene plastics are recycled, and the additives are added to modify the recycled plastics and enhance the toughness, tensile strength and impact strength of the plastics, so that the plastics can be accepted as a material meeting 3D printing requirements, and the cost of the 3D printing material is greatly reduced.

This invention discloses a method for testing inbreak, which obtains the information of software behavior characters of a program, a structure access model and acts it on the program to prevent abnormal behavior so as to protect the entire network server system to avoid network inbreak including the following basic processes: collecting and processing a behavior security box to monitor the software behavior timely and a protection system suitable for the above mentioned method is also disclosed to realize the close coupling control to the behavior of the applied program.

The invention discloses a driving separation system inside a virtual machine and a method and relates to a computer fault tolerant technology. The system, which is simple, effective and capable of being compatible with existing operation systems and a driving program framework and isolating a driving program inside the virtual machine based on a full virtualization technology, is achieved, the driving program isolating system inside the virtual machine comprises a driving isolation uploading module which runs in a user mode of the virtual machine, a memory information monitoring module in a kernel mode of the virtual machine and a memory access control module inside a manger of the virtual machine, and memory information monitoring module in the kernel mode of the virtual machine comprises a packaging function library, a isolation control submodule and an information communication submodule.

The invention discloses a method and a device using a vector instruction to process a file index in a parallel mode. The method comprises the following steps of calculating and sorting index key values of data in a data table, distributing needed storing space of all leaf nodes, using an SIMD (single instruction multiple data) instruction to load a plurality of index key values once, sequentially filling the data in all the leaf nodes, distributing the needed storing space of the internal nodes of the upper layer, building the internal nodes, continuously building the internal nodes of the upper layer until the node number of the upper layer is one, regarding the node as a root node, and finishing the building of an index tree; building a buffer pool in internal storage, using an inquiring queue to temporarily store all received inquiring requests, taking the inquiring requests out of the buffering pool when the length of the inquiring queue reaches a preset threshold value, sorting the inquiring requests, and forming a sorted to-be-inquired queue; sequentially obtaining a to-be-inquired term in the sorted inquiring queues, sequentially searching from the root node of the built index tree, and using the SIMD method to search the to-be-inquired term in a parallel mode inside the internal nodes and the leaf nodes.

The embodiment of the invention discloses a configuration linkage method, a system, a terminal and a network equipment management server, belonging to the communication field. The method comprises the steps of: receiving a configuration linkage trigger message and obtaining current configuration data stored in the terminal in an incremental way according to prestored historical changing serial-numbers; and updating corresponding configuration data in an incremental way according to the current configuration data. The method comprises the steps of receiving the current configuration data, updating corresponding configuration data and storing the current configuration data. The terminal comprises a receiving module, a configuration updating module and a memory module. The network equipment management server comprises a trigger information receiving module, an acquiring module and an updating module. A unified network equipment management system comprises a network configuration server, a network equipment management server and a terminal. The embodiment of the invention leads the performance loss and the user waiting time of the network equipment management server to be reduced greatly.

The invention discloses a disk array caching method for a virtual SSD and an SSD isomerous mirror image. The disk array caching method comprises the steps that a disk is logically divided into two parts, wherein the upper half part forms an RAID0 or RAID5 structure to serve as the virtual SSD so as to back up dirty data in a real SSD, the virtual SSD is in a log writing mode, a ram writing cache is set, and the lower half portion is assembled to be of different RAID structures according to needs of users, and is used as a data array so as to store cold data; an address mapping Hash locating table between a data disk and the cache of the real SSD is built; a timer is set so as to regularly write data cached in the ram writing cache of the virtual SSD into the virtual SSD; a request from an upper layer file system is received; for a reading request, when the reading request is not hit in the real SSD, data are read from the data array. The disk array caching method gives consideration to performance, reliability and cost.

The embodiment of the invention discloses a container-technology-based video monitoring cloud service platform realization method and system. The method comprises: an IVM receives a video service task issued by a user by a VS and a correlated parameter of the task; the IVM generates a resource demand according to the video service task and applies for resources to a resource management device according to the resource demand; the resource management device selects one calculation node server, satisfying the resource demand, among calculation node server as a host machine of the video service task and sends the resource demand and the correlated parameter to the host machine; and the host machine distributes resources to the video service task and establishes a C-IVU based on a video service image and the C-IVU operates the video service and executes the video service task according to the correlated parameter of the video service task. According to the invention, compared with the traditional virtual-machine-based video monitoring cloud realization method, the provided method enables the calculation performance loss to be reduced and the resource utilization rate to be improved.

The invention provides a video file parallel transcoding method and system. The method includes the following steps of firstly, obtaining the number of currently-available transcoding machines and the number of target video files to be output with different code rates; secondly, segmenting a source video file according to the number of the currently-available transcoding machines and the number of the target video files to obtain multiple stages of running streams, wherein each stage of running stream comprises a plurality of audio and video stream data segments; thirdly, sequentially starting the multiple stages of running streams from top to bottom according to the code rates of the target video files so as to transcode the data segments; fourthly, combining the transcoded data segments of each stage of running stream according to the corresponding target format, and obtaining the target video files with different code rates. By means of the video file parallel transcoding method and system, under the condition that the coding quality of output targets is not reduced, the transcoded data size is reduced, and therefore performance losses in the transcoding process are effectively reduced, the transcoding efficiency is improved, and response time of video transcoding tasks is effectively shortened.

The invention discloses a data message processing method from a user mode to a kernel mode. The method comprises the steps that S1, a virtual network card is created in the kernel mode through a TUN/TAP driving manner; S2, a network card drive is loaded in the user mode, and network interface information in the user mode is sent to the virtual network card; S3, the virtual network card creates virtual interfaces according to the network interface information in the user mode, wherein the number and the state of the virtual interfaces are kept consistent with those of a network interface in the user mode; S4, when receiving a message needing to be processed by the kernel mode, the network interface in the user mode sends the message needing to be processed by the kernel mode to the virtual network card, and the virtual network card sends the message needing to be processed by the kernel mode to the kernel mode for processing. The method can be used for eliminating performance loss caused by loading a kernel network card drive and enabling the kernel mode to perceive the states of an actual physical network and a network interface.

The invention discloses a dynamic migrating method of a virtual machine based on performance prediction. The method comprises the following steps of obtaining a characteristic vector by using a singular value decomposing principle by extracting system performance data and analyzing the similarity of the singular value to obtain; then, obtaining the performance prediction value at the next moment by a reverse SVD (Singular Value Decomposition) algorithm so as to start a dynamic migrating mechanism of the virtual machine according to future performance prediction values. According to the method, the performance prediction mechanism can make accurate prediction value and the dynamic prediction mechanism can balance the system load. The system can effectively cope with sudden load by effectively analyzing current information of a physical machine and effectively predicting the future load, so that the performance loss caused by unnecessary migration is reduced.

This invention discloses a method for transmitting information including: the transmitting end codes information to be emitted, interlaces channels and matches rate, then interlaces the matched information sequence and check sequence integratedly and transmits them to the receiving end, which de-interlaces the integer of the sequence and then caries out de-rate match, de-channel interlace and decoding to resume the information sequence. This invention also provides a system for transmitting information including a transmitter and a receiver, in which, the transmitter includes a coding module, a channel interlacing module and a decoding module.

The invention discloses a method for classifying operated load in a virtual machine under cloud computing environment. The method includes firstly acquiring monitoring parameters in 5 minutes of load operation and subjecting the monitored parameters to normalization processing; classifying load monitored into four categories such as CPU (central processing unit)-intensive load, memory intensive load, I/O (input/output) intensive load and network intensive load by means of TSRSVM (training sets refresh SVM); providing corresponding customized optimizing strategies for operation systems which running the four categories of intensive loads, and monitoring operating state of the systems through a performance comparison device; indicating that classification strategies are correct if performance of the systems is improved, otherwise, indicating that the classification strategies are incorrect. By the method, accuracy of load classification is high and system performance loss is low.

The invention relates to an optimized deploying method based on virtual cluster online migration, and belongs to the cloud computing technical field; the method comprises the following steps: using a management module to connect with a management cluster node, wherein the management module is provided with an overall monitoring module, an algorithm forming module, a migration planning module and a migration driving module; using the overall monitoring module to receive the cluster node and virtual machine information data in the cloud computing environment, and sending the information to the algorithm forming module so as to predict future loads and build a long term load performance model; planning a deploying strategy for the virtual cluster, and sorting the virtual machines in the virtual cluster into different host nodes of the cluster, thus allowing the virtual machines to fully utilize the host resources in the cluster when the virtual cluster migration is finished. The novel method can solve the problems that the virtualization technology may easily cause resource fragments, and can solve the defects that the virtual machines cannot obtain the expected requesting resource quantity when the dynamic loading demands and the virtual machines are combined, thus reducing virtual machine performance loss in the cluster nodes, and improving resource utilization rate.

The invention discloses a soft solution mapping method, and relates to the technical field of digital information transmission. The method comprises the following steps of: calculating an amplitude and a phase of a received sign; performing soft solution mapping on an amplitude bit in the received sign according to channel state information and a hard judgment boundary of the received sign, so as to obtain a log-likelihood ratio of the amplitude bit; and performing the soft solution mapping on a phase bit in the received sign according to the channel state information and the hard judgment boundary of the received sign, so as to obtain a log-likelihood ratio of the phase bit, wherein the amplitude bit is a bit which is only related to the amplitude in the received sign; and the phase bit is a bit which is only related to the phase in the received sign. According to the method, an Euclidean distance between each bit and the hard judgment boundary is calculated; the calculation complexity is O (log M); the operation complexity is simplified; the performance loss of a system is extremely low; the resource consumption of hardware is reduced; and the calculation speed is increased.

The invention discloses a decoding method of a PUCCH and an apparatus thereof. The method is characterized in that: a network side analyzes overall user data from a receiving data of the PUCCH and separates independent user data of the each user from the overall user data; a reference pilot frequency symbol is determined for the independent user data, the reference pilot frequency symbol is taken as a reference to calculate a frequency offset estimation value of the pilot frequency symbol and the frequency offset estimation value is used to carry out frequency offset compensation to the other pilot frequency symbols except the reference pilot frequency symbol; the reference pilot frequency symbol of the each user and the other pilot frequency symbols after the frequency offset compensation are used to calculate a channel estimation value of each data symbol, a phase shift of a data symbol is calculated according to the frequency offset estimation value of the pilot frequency symbol and the channel estimation value of the data symbol is used to compensate the corresponding phase shift. By using the method and the apparatus of the invention, under the condition of the large phase shift, decoding performance can be improved to a great extent.

The invention provides a thin provisioning method and apparatus. The method comprises the steps of judging whether data blocks need to be pre-allocated to a logic volume or not according to a preset policy; if yes, determining an amount of the data blocks needed to be pre-allocated to the logic volume, and allocating the data blocks with the amount to the logic volume; and when a writing request for the logic volume is received, obtaining the data block from the data blocks currently allocated to the logic volume according to a data size in the writing request, writing data in the writing request, and recording a physical address of the obtained data block in an address mapping list corresponding to a logic address in the writing request. According to the method and the apparatus, the data blocks are pre-allocated to the logic volume, and the allocated data blocks can be directly obtained to store the data under the condition of whether the granularity of the writing request is relatively large or a quantity is relatively large during allocation in writing, so that the performance loss caused by multi-time traversal query of an allocation bitmap during allocation in writing can be reduced and the writing speed is increased.

The invention provides a preparation process of a Ti2AlNb/TA15 bi-metal titanium alloy composite hollow structure part. The Ti2AlNb/TA15 bi-metal titanium alloy composite hollow structure wing type component is prepared through the steps of blank preparation, surface treatment, welding stop agent coating, superplastic forming/diffusion bonding, cooling, out of a furnace and the like. The superplastic forming/diffusion bonding process method of the dissimilar material hollow structure part breaks through the key technology of the multi-layer hollow structure prepared by combining a Ti2AlNb intermetallic compound and an ordinary titanium alloy, the high-temperature resisting characteristic of the Ti2AlNb intermetallic compound and the good superplasticity of the TA15 titanium alloy are givento full play, structure weight losing is achieved, the guided missile thrust-weight ratio is increased, the application of the Ti2AlNb intermetallic compound to types is promoted, the material application range is expanded, the material utilization rate is increased, and the cost is reduced.

The invention discloses a text big data-oriented Chinese word segmentation method, and belongs to the field of natural language processing. The method is characterized by comprising the following steps: (1) performing decomposition processing on local massive data files to form a data block; (2) performing Map treatment on the decomposed data block to obtain a <Key, Value> key value pair with offset of Key and text content of Value; (3) obtaining a final word segmentation result through a series of word segmentation processing and obtaining the <Key, Value> key value pair with offset of Key and word segmentation result of Value to serve as output of a Map function; (4) performing Reduce processing on the <Key, Value> key value pair obtained by the Map function, obtaining an index file of the <Key, Value> key value pair corresponding to an original file and a word segmentation result file, and summarily writing a final result into an HDFS (Hadoop Distributed File System). According to the method, the word segmentation accuracy is guaranteed under the condition of text big data, and meanwhile the throughput rate of the system and the Chinese word segmentation efficiency are greatly improved, so the method has an extremely high practical value.