43 results about "Server farm" patented technology

The invention relates to the field of data processing, in particular to a server fault diagnosis method and system for a server group, and the method comprises the steps: obtaining and preprocessing server operation state data; constructing a running state isolated tree by using an isolated forest algorithm, analyzing over-limit and low-limit conditions of data of each dimension, determining a neighbor monitoring sequence and constructing a threshold offset index; based on the index, analyzing association characteristics among the dimension data to construct a fault association index; further analyzing data anomaly differences between leaf nodes and child nodes in the isolated tree, and constructing an intra-layer isolated index; and calculating an abnormal isolated weight and a significant abnormal weight, improving an isolated forest algorithm, and realizing accurate diagnosis of server faults. According to the method, the significant abnormal weight is introduced to improve the isolated forest algorithm, and the threshold offset index and the fault correlation index are constructed, so that the server fault diagnosis precision is significantly improved, false alarm and missing alarm are reduced, operation and maintenance personnel are helped to accurately position faults, and stable operation of the server is guaranteed.

The invention relates to the technical field of databases, and discloses a structural data synchronization method and system of a relational database. Data users are distributed to candidate data center servers deployed with the relational database; acquiring deployment state information and communication state information of a plurality of data center servers in a target structure data synchronization period, and solving an optimal relational database synchronization strategy of a server group by taking a synchronization target demand, a synchronization data volume demand and a synchronization frequency demand of each data user in different minimum service periods as a drive; therefore, the structure data synchronization between the relational databases is controlled. Therefore, by realizing the structure data synchronization between the relational databases under the scene of uneven regional communication load, the period consistency of the metadata of the databases is kept, the selection for the source data center server and the rationality and scientificity of the structure data synchronization path planning are improved, and the reliability of the structure data synchronization path planning is improved. The stability of the structure data synchronization period is ensured as much as possible.

The invention provides a teaching evaluation method and device, a storage medium and a program product, and relates to the technical field of teaching quality evaluation.The method comprises the steps that a first issuing instruction sent by a first user in a teaching classroom through a first server is received, and a first classroom questionnaire set is sent to a second user group and a second server group in the teaching classroom; acquiring a second classroom questionnaire set uploaded by a second server group; a second issuing instruction sent by the first user through the first server is received, and the first extracurricular questionnaire set is sent to a second user group and a third server group of the non-teaching classroom; obtaining a second extracurricular questionnaire set uploaded by a third server group; performing data fusion on the second classroom questionnaire set and the second extracurricular questionnaire set to obtain a target archive data set; and the target archive data set performs teaching evaluation on the target user group. The technical problem of low reliability of teaching evaluation in related technologies is solved, and the technical effect of improving the reliability of teaching evaluation is achieved.

The invention relates to the technical field of computer load balancing, in particular to a large model reasoning load balancing method and system based on a dynamic resource grading threshold value. The method comprises the following steps: performing resource consumption estimation on a user reasoning request to obtain a request score value; the servers are divided into two server groups with different computing capabilities; differentiated scheduling is carried out based on the request score value and the current dynamic resource grading threshold value; and collecting the load state of the server cluster, calculating a load difference value, and dynamically adjusting the dynamic resource grading threshold value through a feedback control algorithm to realize load balancing. According to the method, the resource consumption of the reasoning request of the user is accurately estimated, and the efficient load balancing of the large-model reasoning task is realized in combination with the hierarchical management of the server, the real-time load state evaluation of the server cluster and the dynamic threshold adjustment mechanism.

The invention discloses a data center dynamic energy-saving scheduling method and system based on AI prediction, electronic equipment and a storage medium, and the method comprises the steps: collecting load data of a data center server and external service flow data in real time to construct an original data set, and processing the original data set through employing a long short-term memory network to obtain a load change trend; according to the load change trend, determining distribution characteristics of the thermal load trend in a future time window, extracting idle computing resource data from the server group based on the distribution characteristics, and computing a computing resource distribution proportion needing to be adjusted; carrying out iterative optimization on the calculation resource distribution proportion by adopting a genetic algorithm to obtain an optimized server load distribution scheme; through a server load distribution scheme, obtaining associated refrigeration system parameter data, determining an adjustment vector of a refrigeration strategy, fusing real-time energy consumption data of a power supply module, and judging a correction demand of power supply scheduling; and obtaining the operation configuration of the overall energy efficiency optimization according to the correction demand.

The embodiment of the application discloses a data synchronization management system, method, device, medium and product, the system comprises: at least two server groups; each of the server groups comprises a data synchronization process and at least one business server; each of the business servers is associated with the data synchronization process; for each server group, the business server is used for generating a data synchronization message containing to-be-synchronized business data when having a data cross-group synchronization demand, and sending the data synchronization message to the associated data synchronization process; the data synchronization process is used for determining a target server group to be cross-group synchronized according to the received data synchronization message, and transmitting the to-be-synchronized business data to a target business server of the target server group through a generated synchronization data packet. The flexibility and effectiveness of application business data synchronization under a large-scale Internet architecture are better ensured, and the generality of data synchronization under the large-scale Internet architecture is better embodied.

The application discloses a kind of server fault diagnosis method and system for server group, it is related to server operation and maintenance technical field, present and propose the following scheme, including data acquisition unit, data preprocessing unit, data training unit, data diagnostic analysis unit, scheme generation unit, scheme verification trial unit, data storage unit and data early warning unit, each unit is connected by communication bus or network link and forms closed loop diagnosis system.The application constructs full-link fault diagnosis system, through multidimensional collection hardware, software, network data, improves data quality by scientific pretreatment, adopts rule matching+machine learning hybrid model, combined with self-learning mechanism, improve diagnostic accuracy and coverage, generate personalized repair scheme and be verified by simulation / real machine, avoid secondary failure, full-process data safety storage and double early warning mechanism, realize fault tracing and prevention and control in advance, greatly improve server group operation and maintenance efficiency, guarantee business continuity.

A server farm with at least one hybrid computing module operating at clock speed optimally matching intrinsic clock speed of a related semiconductor die related thereto.

The invention discloses a control method of refrigeration equipment and related equipment, and the method comprises the following steps: determining a heating information set according to a first operation information set of a server cluster and current weather forecast information according to the control method and system of the refrigeration equipment, electronic equipment, a storage medium and a program product; determining a target refrigeration server according to a preset threshold set and the heating information set; according to the target refrigeration server, the current weather forecast information and the heating rule model, target refrigeration equipment and a refrigeration parameter set are determined; and performing directional refrigeration on the target refrigeration server according to the target refrigeration equipment and the refrigeration parameter set. According to the embodiment of the invention, the target server can be determined according to the heating information of the server cluster, and the target refrigeration equipment is determined and called to perform directional refrigeration in combination with the heating rule model, the current weather forecast information and the target server, so that the refrigeration energy consumption is reduced, and the refrigeration precision is improved. The method can be widely applied to the technical field of data center heat management.

Systems, apparatus, articles of manufacture, and methods to distribute workloads in server farms based on temperature are disclosed. An example first compute device includes at least one programmable circuit to at least one of instantiate or execute machine readable instructions to: analyze temperature data indicative of a first temperature of a first compute device and a second temperature of a second compute device; and cause an adjustment in at least one of a first workload executed by the first compute device or a second workload executed by the second compute device based on the temperature data, the adjustment to reduce a difference between the first and second temperatures.

This invention discloses an automatic inspection method and system for server security self-testing. The method includes the following steps: S1, acquiring distributed node information and initial security configuration parameters of a server group, and establishing a global security inspection topology; S2, deploying an adaptive inspection agent on each server node, wherein the agent initializes a local inspection strategy based on the global security configuration parameters; S3, through a distributed computing architecture, each adaptive inspection agent independently performs local security checks and generates node security status data. This invention relates to the field of server security technology. This automatic inspection method and system for server security self-testing, through collaborative communication between agents, enables the system to share threat information, vulnerability remediation suggestions, and system status data in real time, constructing a global security data pool, and deriving a global security posture value based on weighted calculation. This allows the system to promptly understand the security status of the entire server group and provide accurate security posture awareness.

The invention discloses a privacy protection smart medical service recommendation method based on weight matching. The method comprises the steps that a medical management center generates materials needed by a security key and sends the materials to a patient user and a server cluster; the medical management center carries out doctor registration service and patient user registration service through copy secret sharing and the server group; after registration is completed, each server in the server cluster obtains a pair of shared values; screening doctors by calculating a similarity vector between the user demand vector and the doctor attribute vector; and after the doctor service is completed, calculating the weight of each dimension of the score vector of the user, and updating the score of the doctor based on each dimension value of the score vector and the corresponding weight. According to the method, the most suitable doctor is recommended by calculating the similarity among the patient demand, the doctor attribute and the score vector; multi-party calculation of data is realized through a function secret sharing technology, and the data security is enhanced. Calculation and auxiliary servers are introduced, so that the online calculation cost is reduced, and malicious scoring is effectively resisted.

A multiphase cooling system for electronics racks, server farms, and data centers is described. An inlet of a three-way flow control valve (FCV) is coupled to a primary coolant source. A first outlet of the FCV is coupled to a single-phase cooling system, and a second outlet of the FCV is coupled to a two-phase cooling system. The FCV is configured to adjust the amount of coolant between the single-phase cooling system and the two-phase cooling system. Upon detecting a rise in vapor pressure in a return line of the two-phase cooling system, the FCV can be adjusted to direct more coolant to the two-phase cooling system and less coolant to the single-phase system. The FCV can continuously monitor the vapor pressure and adjust the amount of coolant to each cooling system accordingly.

The present invention relates to the technical field of liquid cooling system management, and particularly relates to an energy consumption data management platform for a liquid cooling system based on artificial intelligence. The energy consumption data management platform for the liquid cooling system includes a data acquisition module for acquiring the actual energy consumption data of each server in a server group under each historical liquid cooling cycle; a first partitioning module for obtaining initial clustering clusters; a second partitioning module for reassigning the energy consumption data points in the initial clustering clusters to obtain respective target clustering clusters; and a liquid cooling system management module for reassigning and managing the servers controlled by the CDU of the liquid cooling system according to the servers corresponding to the energy consumption data points in the target clustering clusters. Moreover, the present invention can reduce the energy consumption waste during the cooling of a centralized liquid cooling system, improve the energy consumption utilization rate and energy utilization rate of the centralized liquid cooling system, and realize the balanced management of the energy consumption and resources of the centralized liquid cooling system.

Server farm power management system.

The present disclosure relates to a method for edge level training of a model by one or more processors (202) The method includes adding one or more edge servers (104) to a cluster of centralized servers. Further, the method includes receiving, at the one or more edge servers (104), data from one or more edge devices (102). Further, the method includes training, at the one or more edge servers (104), a model with the data. Further, the method includes deploying the trained model onto at least one of, the one or more edge devices (102) or the one or more edge servers (104). Further, the method includes synchronizing the trained model with the cluster of centralized servers.

The embodiment of the invention provides a server state monitoring system and method, and relates to the technical field of computers. The system comprises a monitoring center and at least one server group, the server group comprises a plurality of servers, each server comprises a central server and at least one edge server, and each server automatically detects own operating environment information, registers matched resource acquisition services according to the operating environment information, and transmits the matched resource acquisition services to the monitoring center. The edge servers perform multi-source heterogeneous data acquisition on monitoring objects of the edge servers to obtain monitoring data, the monitoring data are packaged into structured message bodies, the edge servers report the structured message bodies to the central server, and the central server receives the structured message bodies from the edge servers and sends the structured message bodies to the server. And locally storing the structured message bodies of the edge servers and the structured message bodies of the edge servers and reporting the structured message bodies to a monitoring center. Therefore, the problem that a traditional monitoring scheme is poor in adaptability in a heterogeneous environment is effectively solved.

The application relates to the technical field of databases, and discloses a relational database structure data synchronization method and system, wherein data users are distributed to candidate data center servers in which relational databases are deployed, the deployment state information and the communication state information of a plurality of data center servers in a target structure data synchronization period are acquired, the synchronization target demand, the synchronization data volume demand and the synchronization frequency demand of each data user in different minimum service time periods are taken as driving, the optimal relational database synchronization strategy of a server group is solved, and the structure data synchronization between relational databases is controlled. Therefore, the structure data synchronization between relational databases in a regional communication load uneven scene is realized, the periodic consistency of database metadata is maintained, the rationality and the scientificity of the selection of a source data center server and the structure data synchronization path planning are improved, and the stability of the structure data synchronization period is ensured as much as possible.

The invention provides a data processing method and system based on edge cloud and a storage medium, and belongs to the technical field of data processing.The method specifically comprises the steps that standby servers in servers are determined on the basis of the distance between the servers and different common access sites, the standby servers are freely combined to obtain a server group, and the server group is stored in the server group; determining a matched server group in the server groups according to access processing delays and residual storage spaces of standby servers in different server groups on different dates, and determining the distance data between the different standby servers in the matched server group and a common access place and the access data of the matched access user according to the distance data between the different standby servers in the matched server group and the common access place and the access data of the matched access user. According to the method and the device, the data processing sequence of the standby servers corresponding to the different common access sites is determined, and when the access processing delay of the standby server does not meet the requirement, the next standby server in the data processing sequence is automatically switched, so that the data processing efficiency is improved.

The application belongs to the technical field of information matching and data evaluation, and proposes a full asset wealth protection and inheritance risk evaluation method and system, specifically: first, a full asset network identification benchmark server and a redistribution server are identified, which are respectively recorded as a benchmark object and an evaluation object, a benchmark word adaptive portrait is constructed through an asset database of each evaluation object, then inheritance execution quality time sequence analysis is performed by using the benchmark word adaptive portrait to obtain an inheritance execution score of each evaluation object, and finally, inheritance risk evaluation is performed in combination with the inheritance execution score of each evaluation object. The coordination between the redistribution server and the related database of the full asset wealth needing to perform inheritance processing and the benchmark server can be effectively quantified, and the risk of data tampering, loss or individual asset data read / write failure of the server group in the inheritance task is reduced.

The invention provides an asset vulnerability repairing method and device, equipment, a storage medium and a program product, and relates to the field of financial science and technology or other related fields. The method comprises the following steps: acquiring a vulnerability scanning report of a server cluster, and standardizing the vulnerability scanning report to obtain vulnerability scanning data of a target vulnerability; according to a vulnerability knowledge base and an asset fingerprint database, determining an asset identifier corresponding to the vulnerability identifier and an asset feature parameter corresponding to the asset identifier, the vulnerability knowledge base including a mapping relationship between the vulnerability identifier, the vulnerability type, the vulnerability influence range and the asset identifier, and the asset fingerprint database including a mapping relationship between the asset identifier and the asset feature parameter; and determining a target repair scheme of the target vulnerability according to the asset feature parameters, the vulnerability type and the vulnerability influence range, and repairing the target vulnerability according to the target repair scheme. According to the method, the problem of low decision efficiency caused by vulnerability and asset splitting is solved, the vulnerability scanning efficiency is improved, and accurate vulnerability positioning is realized.

This disclosure describes a system of distributed, compact, low height, and modular cooling systems that can potentially achieve significant reduction in energy usage by data centers, server farms and other facilities. Discrete cooling ecosystem configurations for each group of server rack, each server rack, each GPU tray containing multiple GPUs within a rack, or each of the GPUs within a GPU tray are disclosed.

A server farm ventilation assembly adapted to separate an interior of a building from an exterior while allowing air to pass from the exterior to the interior, comprising of a frame defining an air filter opening, and an air filter subassembly comprising a plurality of air filters. A plurality of removable vertical brackets and horizontal brackets are used to retain the plurality of air filters within the frame thereby allowing filtered air to pass through the air filter opening.

The invention discloses a data communication network, an intelligent control system and a control method thereof, the data communication network comprises a first data communication network and a first server cluster, and the first data communication network comprises a first core switch and a first service network switch; the first core switch is connected with the first service network switch; the first service network switch is connected with a control network; the first service network switch is connected with the first server cluster; and the control network is connected with the DCS system. Communication between the server and the DCS system can be realized based on the first data communication network, so that the burden of the DCS system is reduced, and the operation stability of the DCS system is improved.

To facilitate new residents to move in homes, thereby promoting the new residents to settle down in the homes.SOLUTION: A home service providing system includes a communication terminal for each resident at a plurality of homes and a server group. The server group comprises a coupon code database for storing coupon codes for obtaining tickets to enjoy prescribed services in the neighborhood to which the homes belong, and a server-side processor. The communication terminal comprises a terminal-side processor and a display. A CPU of the server executes processing for distributing a coupon code in the coupon code database to a resident (S513). A CPU of a mobile terminal receives input of the coupon code distributed to the resident (S111) and displays a ticket corresponding to the received coupon code on a display of the mobile terminal (S132).SELECTED DRAWING: Figure 6

An automatic card meter copying device, a charge center server sends out inquiry messages, when the card meter is to be charged, the IC card of the card meter is inserted into the IC card reader, then the computer runs the collection program, when the collection program collects the inquiry messages sent out by the charge center server, the channel structure message of the response inquiry message is transmitted to the charge center server, to prevent the channel structure messages transmitted by several computers from colliding due to the same transmission time or a very short interval, using the anti-collision method of the invention, after the inquiry message is collected, a delay of any length is followed, then the inquiry message is transmitted, here, by using the any length program to calculate any length one, the probability of collision of the channel structure messages transmitted by the computer can be reduced to the minimum.

The present invention provides a method and a system for processing session data in an interactive process, a server group, and a computer readable storage medium. A memory module and a protocol configuration module are configured in a gateway server, keywords can be directly extracted by the protocol configuration module from a session of a client and a feedback session of an interactive logic server, and then the keywords can be directly forwarded, so that the throughput of session data processing is improved. The corresponding data is stored, extracted and forwarded by using the memory module, so that the delay of data transmission is reduced. Because the memory module is arranged in the gateway server, the occurrence rate of interface instability can be reduced.

A fully integrated gyrator with a high speed stack and networks of server farms and telecommunication modes incorporating the gyrator.

The application relates to a big data driven network security situation monitoring and visualization method, and relates to the field of network security.The method comprises the following steps: using a big data network element to perform each training operation on a feedforward neural network to obtain a network security situation monitoring model; using the network security situation monitoring model to intelligently judge the attack virus number of a set server group in a target time interval based on the duration of each time interval, the configuration data of the set server group and the past attack virus number of the set server group. Through the application, the network security situation of a server group composed of multiple servers in a future time interval can be predictively predicted, monitored and responded to by using the network security situation monitoring model after corresponding training, so that the probability of successful virus attack is reduced, and the network crisis of data leakage or system paralysis is avoided.

The application provides a network virus propagation defense method, device and equipment based on reinforcement learning. The method comprises the following steps: abstracting a target network into a two-dimensional space Graph form as training data for storage; constructing a reinforcement learning training model, and defining a training environment of the reinforcement learning training model; the training environment of the reinforcement learning model comprises a training environment rule, a server state, an invader attack rule and a training end condition; training the reinforcement learning training model to obtain a reinforcement learning defense model; and deploying the reinforcement learning defense model to a management server, inputting a server state in the target network into the reinforcement learning defense model, and performing offline on a server in an output result. In this way, the reaction speed of both attack and defense sides can be considered without considering the defense or attack strength on the premise that the virus attack can be detected, the server group is protected from a macro perspective, and the whole can still run when a small part of computers in a complex network is invaded.