CN114138750B - AI consultation database based cluster building method and system - Google Patents
AI consultation database based cluster building method and system Download PDFInfo
- Publication number
- CN114138750B CN114138750B CN202111470286.6A CN202111470286A CN114138750B CN 114138750 B CN114138750 B CN 114138750B CN 202111470286 A CN202111470286 A CN 202111470286A CN 114138750 B CN114138750 B CN 114138750B
- Authority
- CN
- China
- Prior art keywords
- cluster
- enterprise
- node
- time
- building
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/211—Schema design and management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/14—Error detection or correction of the data by redundancy in operation
- G06F11/1402—Saving, restoring, recovering or retrying
- G06F11/1446—Point-in-time backing up or restoration of persistent data
- G06F11/1458—Management of the backup or restore process
- G06F11/1464—Management of the backup or restore process for networked environments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/27—Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5083—Techniques for rebalancing the load in a distributed system
Abstract
The invention discloses a database cluster building method and system based on AI consultation, which relate to the technical field of cluster building and solve the technical problem that the cluster built by an enterprise cannot be analyzed and detected in the prior art, so that the working efficiency of the cluster is reduced; the load balancing unit is used for analyzing the nodes in the enterprise cluster, performing task planning on each node according to the node response processing process, and performing distributed processing according to the planning result, so that the problems that the data processing efficiency is low due to overlarge load of a single node, the operation of the nodes in the cluster is unbalanced, the data resources are uncoordinated, and the cluster cannot normally operate for a long time are easily caused are solved.
Description
Technical Field
The invention relates to the technical field of cluster building, in particular to a cluster building method and system based on an AI consultation database.
Background
With the development of internet technology, mass data access makes the traditional database management system unable to meet business requirements, and the database is used as the core of an application system, the position in an IT system of an enterprise is very important, and the risk of using an independent server traditionally lies in that once a system failure occurs, not only the normal operation of enterprise business is seriously affected, but also economic loss is caused to the enterprise, and along with the growth of the enterprise, the access amount and data flow of the system are rapidly increased while the business volume is increased, the processing capacity and the calculation intensity of the system are correspondingly increased, so that a single device cannot be born at all, under the condition, if the existing device is thrown away to do a large amount of hardware upgrade, the waste of the existing resources is caused, and if the next business volume is increased, the high cost investment of the hardware upgrade is caused again, and even the device with excellent performance cannot meet the requirements of the current business volume;
however, in the prior art, it cannot be accurately determined whether an enterprise needs to establish a cluster, and meanwhile, the cluster cannot be accurately analyzed after being established, and the cluster cannot be subjected to distributed processing, so that load imbalance and low efficiency are caused; in addition, the performance of the cluster cannot be analyzed, and the cluster cannot be maintained or tidied in advance, so that the safety performance of a large amount of data in the cluster is reduced, and the convenience brought to an enterprise by the cluster cannot be guaranteed;
in view of the above technical drawbacks, a solution is proposed.
Disclosure of Invention
The invention aims to provide a database cluster building method and system based on AI consultation, which are used for building database clusters for each enterprise, thereby improving the efficiency of business consultation and inquiry of the enterprise and effectively preventing the later-stage improvement of data storage and processing cost of the enterprise; and performing task planning on each node according to the node response processing process, and performing distributed processing according to the planning result, so that the problems that the data processing efficiency is low due to overlarge load of a single node, the data resources are uncoordinated due to unbalanced operation of the nodes in the cluster, and the cluster cannot normally operate due to long-term easiness are solved.
The purpose of the invention can be realized by the following technical scheme:
a cluster building system based on AI consultation database comprises a cluster building platform; the cluster building platform is provided with a server, the server is in communication connection with an enterprise analysis unit, a cluster building unit and a plurality of nodes, and the cluster building unit is in communication connection with a load balancing unit, an analysis maintenance unit and a mirror fault tolerance unit;
the method comprises the steps that a database cluster is built for each enterprise through a cluster building platform, a server generates enterprise analysis signals and sends the enterprise analysis signals to an enterprise analysis unit, and the enterprise analysis unit analyzes the operation of the enterprise after receiving the enterprise analysis signals; generating a cluster building signal according to the analysis, sending the cluster building signal to a server, and building an enterprise cluster through a cluster building unit after the server receives the cluster building signal; after the cluster construction unit completes the cluster construction of the corresponding enterprise, the cluster construction unit detects and maintains the enterprise cluster, generates a load balancing signal and sends the load balancing signal to the load balancing unit; analyzing the nodes in the enterprise cluster through a load balancing unit; analyzing the operation of the enterprise cluster through an analysis maintenance unit; and carrying out data backup on the enterprise cluster through a mirror fault-tolerant unit.
Further, the analysis process of the enterprise analysis unit is as follows:
acquiring the quantity of business orders of an enterprise in real time, marking the quantity of the business orders of the enterprise acquired in real time as SL, acquiring the business access quantity corresponding to the enterprise according to the quantity of the business orders of the enterprise, and marking the business access quantity corresponding to the enterprise as FW; acquiring data traffic generated by enterprise business through the quantity of business orders and the business access quantity, and marking the data traffic generated by the enterprise business as LL;
obtaining an enterprise real-time cluster building analysis coefficient X through analysis, and comparing the enterprise real-time cluster building analysis coefficient X with a cluster building analysis coefficient threshold value:
if the real-time cluster building analysis coefficient X of the enterprise is larger than or equal to the cluster building analysis coefficient threshold value, judging that the corresponding enterprise needs to carry out cluster building, generating a cluster building signal and sending the cluster building signal to a server; and if the real-time cluster building analysis coefficient X of the enterprise is less than the cluster building analysis coefficient threshold, judging that the corresponding enterprise does not need to build a cluster, generating a cluster non-building signal and sending the cluster non-building signal to the server.
Further, the balancing process of the load balancing unit is as follows:
collecting each node in the enterprise cluster, and setting the node with the labels i, i =1,2, …, n, n being a natural number greater than 1; acquiring a demand instruction received by each node, analyzing according to keywords in the demand instruction, and judging whether the demand instruction is of the same type;
marking the nodes receiving the same type of demand instructions as the same order nodes, and marking the same order nodes as u, wherein u belongs to i; acquiring the response time of the demand instruction processed by the same-order node, and marking the response time of the demand instruction processed by the same-order node as XYu; acquiring the receiving frequency of the repeated instruction after the demand instruction of the same order node responds, and marking the receiving frequency of the repeated instruction after the demand instruction of the same order node responds as PLu; the repeated instruction in the application is represented as the same demand instruction received by the same order node; obtaining a response matching coefficient Zu of the same-order node through analysis;
comparing the response matching coefficient of the concordance node with a response matching coefficient threshold:
if the response matching coefficient of the corresponding order node is not less than the response matching coefficient threshold, judging that the processing efficiency of the corresponding order node is high, and marking the corresponding order node as a high-efficiency order node; if the response matching coefficient of the corresponding syntype node is less than the response matching coefficient threshold value, judging that the processing efficiency of the corresponding syntype node is low, and marking the corresponding syntype node as a low-efficiency syntype node;
sequencing the high-efficiency concordant nodes according to the sequence of the corresponding response matching coefficient values from large to small, matching the first high-efficiency concordant node with the corresponding demand instruction, marking the first high-efficiency concordant node as an exclusive node of the corresponding demand instruction, and simultaneously marking the second high-efficiency concordant node as a standby node of the corresponding demand instruction; marking the low-efficiency syntype node as a no-response node of the corresponding demand instruction, and if the no-response node receives the corresponding demand instruction, transmitting the corresponding demand instruction and the demand instruction sending terminal to the exclusive node together; and sending the exclusive node, the standby node, the no-response node and the corresponding demand instruction to a cluster construction unit for storage.
Further, the analysis maintenance process of the analysis maintenance unit is as follows:
acquiring the time of establishing the enterprise cluster, and acquiring the real-time establishment duration of the enterprise cluster according to the time of establishing the enterprise cluster and the current time; acquiring the average fault time length and the minimum interval time length of adjacent faults in the real-time construction time length of the enterprise cluster, and respectively marking the average fault time length and the minimum interval time length of adjacent faults in the real-time construction time length of the enterprise cluster as PSC and DSC; the reliability coefficient SS of the enterprise cluster is obtained through analysis, and is compared with a reliability coefficient threshold value:
if the reliability coefficient SS of the enterprise cluster is larger than or equal to the reliability coefficient threshold, judging that the reliability of the enterprise cluster is qualified, generating a cluster reliable signal and sending the cluster reliable signal to a cluster construction unit; if the reliability coefficient SS of the enterprise cluster is smaller than the reliability coefficient threshold value, judging that the reliability of the enterprise cluster is unqualified, generating a cluster unreliable signal and sending the cluster unreliable signal to a cluster construction unit;
acquiring the average maintenance time of the faults in the real-time construction duration of the enterprise cluster, and marking the average maintenance time of the faults in the real-time construction duration of the enterprise cluster as WSC; acquiring the times of completing maintenance in a short time according to the average maintenance time of the faults, and marking the times of completing maintenance in the short time of the faults as CWS; obtaining a maintainability coefficient KW of the enterprise cluster through analysis;
comparing the maintainability coefficient KW of the enterprise cluster with a maintainability coefficient threshold: if the maintainability coefficient KW of the enterprise cluster is not less than the maintainability coefficient threshold, judging that the maintainability of the enterprise cluster is qualified, generating a maintainability signal and sending the maintainability signal to the cluster construction unit; if the maintainability coefficient KW of the enterprise cluster is smaller than the maintainability coefficient threshold, judging that the maintainability of the enterprise cluster is unqualified, generating an unremainable signal and sending the unremainable signal to a cluster construction unit;
calculating the sum of the average maintenance time of faults and the average time of faults in the real-time construction time of the enterprise cluster, calculating the ratio of the average time of faults to the corresponding sum, marking the corresponding ratio as an availability coefficient, and comparing the availability coefficient of the enterprise cluster with an availability coefficient threshold value: if the availability coefficient of the enterprise cluster is larger than or equal to the availability coefficient threshold value, judging that the availability of the enterprise cluster is qualified, generating an available signal and sending the available signal to a cluster construction unit; and if the availability coefficient of the enterprise cluster is less than the availability coefficient threshold value, judging that the availability of the enterprise cluster is unqualified, generating an unavailable signal and sending the unavailable signal to the cluster construction unit.
Further, the backup process of the mirror fault tolerant signal is as follows:
monitoring business data and instruction responses of an enterprise cluster, arranging a sub-server in the enterprise cluster, storing and counting the business data and the instruction responses of the enterprise cluster by the sub-server, deleting the sub-server in a delayed manner when the server in the enterprise cluster deletes the data, generating a judgment misoperation instruction by a sending terminal of the misoperation instruction if the data is deleted to be instruction misoperation, sending the judgment misoperation instruction to the server, and copying corresponding data in the sub-server after the data is received by the server; and simultaneously generating a mirror image completion signal and sending the mirror image completion signal to the cluster building unit.
Further, the cluster building method based on the AI consultation database comprises the following steps:
step one, enterprise analysis, namely analyzing the enterprises in real time, acquiring an enterprise real-time cluster building coefficient according to the business access amount, the business order number and the corresponding data flow, judging whether the enterprises carry out cluster building or not according to the enterprise real-time cluster building coefficient, and entering a step two if the enterprises carry out cluster building;
step two, establishing a cluster, namely setting each department in an enterprise as each node, carrying out communication connection on a server and each node, and carrying out network coverage on the server and each node by a local area network in the enterprise after the server is in communication connection with each node to construct an enterprise cluster;
thirdly, load balancing detection, namely dividing each node in the enterprise cluster, dividing the node into an exclusive node, a standby node and a no-response node according to a real-time demand instruction, and performing instruction distribution on each node according to a division type;
analyzing, maintaining and detecting the cluster, performing analysis, maintenance and detection after the enterprise cluster is constructed, and judging the real-time running state of the enterprise cluster by analyzing the reliability, maintainability and availability of the enterprise cluster;
and fifthly, setting cluster mirror image fault tolerance, namely setting a sub-server according to a server in the enterprise cluster, transmitting data in the server to a sub-server for storage, and setting a delayed deletion instruction for the sub-server, if the data in the server is deleted and judged to be in misoperation, generating a misoperation judgment instruction through a misoperation sending terminal, sending the misoperation judgment instruction to the server, and restoring the data through the sub-server by the server.
Compared with the prior art, the invention has the beneficial effects that:
according to the invention, the database cluster is built for each enterprise, so that the business consultation and query efficiency of the enterprise is improved, and the later-stage improvement of data storage and processing cost of the enterprise is effectively prevented; the method comprises the steps of performing task planning on each node according to a node response processing process, and performing distributed processing according to a planning result, so that the problems that the data processing efficiency is low due to overlarge load of a single node, the data resources are uncoordinated due to unbalanced operation of the nodes in a cluster, and the cluster cannot normally operate due to long-term easiness are solved;
whether the operation state of the enterprise cluster is normal or not is judged, so that the problem that the operation of the enterprise cluster is abnormal and cannot be detected in time, the service data is lost to block the service processing progress, and negative effects are brought to enterprise benefits; the data backup is carried out on the enterprise cluster, so that the data loss caused by network intrusion or instruction misoperation is prevented, and the irreparable loss is caused to the enterprise.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the prior art descriptions will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a schematic block diagram of the system of the present invention;
FIG. 2 is a flow chart of the method of the present invention.
Detailed Description
The technical solutions of the present invention will be described below clearly and completely in conjunction with the embodiments, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
As shown in fig. 1, a cluster building system based on an AI consultation database includes a cluster building platform and a cluster operating platform, both of which are in bidirectional communication connection, the cluster building platform is provided with a server, the server is in communication connection with an enterprise analysis unit, a cluster building unit, a plurality of nodes, and the cluster building unit is in communication connection with a load balancing unit, an analysis and maintenance unit, and a mirror fault tolerance unit;
the cluster building platform is used for carrying out database cluster to each enterprise and builds, the efficiency of business consultation of enterprise and inquiry has been improved, effectively prevent the enterprise later stage and improve data storage and processing cost, the server generates enterprise analysis signal and with enterprise analysis signal transmission to enterprise analysis unit, enterprise analysis unit receives enterprise analysis signal after, carries out the analysis to the operation of enterprise, judges whether the enterprise needs carry out the cluster and builds, concrete analytic process is as follows:
acquiring the quantity of business orders of an enterprise in real time, marking the quantity of the business orders of the enterprise acquired in real time as SL, acquiring the business access quantity corresponding to the enterprise according to the quantity of the business orders of the enterprise, and marking the business access quantity corresponding to the enterprise as FW; acquiring data traffic generated by enterprise business through the quantity of business orders and the business access quantity, and marking the data traffic generated by the enterprise business as LL;
by the formula
Acquiring an enterprise real-time cluster building analysis coefficient X, wherein a1, a2 and a3 are all preset proportionality coefficients, and a1 is more than a2 and a3 is more than 0; the enterprise real-time cluster building analysis coefficient is a numerical value used for judging the cluster building probability of the enterprise by carrying out normalization processing on real-time parameters of enterprise services; the larger the quantity of the service orders, the service access amount and the data flow which can be obtained through the formula, the larger the real-time cluster building analysis coefficient of the enterprise isThe larger the probability of the enterprise for constructing the cluster is;
comparing the real-time cluster building analysis coefficient X of the enterprise with a cluster building analysis coefficient threshold value:
if the real-time cluster building analysis coefficient X of the enterprise is larger than or equal to the cluster building analysis coefficient threshold value, judging that the corresponding enterprise needs to carry out cluster building, generating a cluster building signal and sending the cluster building signal to the server;
if the real-time cluster building analysis coefficient X of the enterprise is smaller than the cluster building analysis coefficient threshold, judging that the corresponding enterprise does not need to carry out cluster building, generating a cluster non-building signal and sending the cluster non-building signal to the server;
after receiving the cluster building signal, the server sets each department in the enterprise as each node, performs communication connection between the server and each node, and performs network coverage on the server and each node through a local area network in the enterprise corresponding to the cluster building unit after the server is in communication connection with each node to build an enterprise cluster; the enterprise cluster can perform functions of data storage, stored data query, data transmission and the like; in the application, the node comprises equipment to which a department belongs;
after the cluster construction unit completes the cluster construction of the corresponding enterprise, the enterprise cluster is detected and maintained, load balancing signals are generated and sent to the load balancing unit, the load balancing unit is used for analyzing the nodes in the enterprise cluster, task planning is carried out on each node according to the node response processing process, meanwhile, distributed processing is carried out according to the planning result, the single node is prevented from being overloaded, the data processing efficiency is low, the unbalanced operation of the nodes in the cluster is caused, the data resources are not coordinated, the cluster cannot normally operate for a long time, and the specific balancing process is as follows:
collecting each node in the enterprise cluster, and setting the node with the labels i, i =1,2, …, n, n as a natural number more than 1; acquiring a demand instruction received by each node, analyzing according to keywords in the demand instruction, and judging whether the demand instruction is of the same type, wherein the type of the demand instruction in the application is represented as a process required by a business process, such as inquiry of a business order, business negotiation reservation, business remittance inquiry and the like;
marking the nodes receiving the same type of demand instructions as the same order nodes, and marking the same order nodes as u, wherein u belongs to i; acquiring the response time length of the demand instruction processed by the same-order node, and marking the response time length of the demand instruction processed by the same-order node as XYu; acquiring the receiving frequency of the repeated instruction after the demand instruction of the same order node responds, and marking the receiving frequency of the repeated instruction after the demand instruction of the same order node responds as PLu; the repeated instruction in the application is represented as the same demand instruction received by the same order node; by the formula
Acquiring a response matching coefficient Zu of the same order of nodes, wherein c1 and c2 are preset proportionality coefficients, c1 is larger than c2 and larger than 0, and beta is an error correction factor and takes a value of 2.01; the response matching coefficient of the syntactical node is a numerical value used for judging the processing demand instruction efficiency of the syntactical node by carrying out normalization processing on the real-time parameter of the syntactical node; the larger the response time of the processing demand instruction is, the smaller the receiving frequency of the repeated instruction is, and the larger the response matching coefficient of the same-order node is, the better the efficiency of the same-order node for processing the demand instruction is;
comparing the response matching coefficient of the concordance node with a response matching coefficient threshold:
if the response matching coefficient of the corresponding order node is not less than the response matching coefficient threshold, judging that the processing efficiency of the corresponding order node is high, and marking the corresponding order node as a high-efficiency order node;
if the response matching coefficient of the corresponding syntype node is less than the response matching coefficient threshold value, judging that the processing efficiency of the corresponding syntype node is low, and marking the corresponding syntype node as a low-efficiency syntype node;
sequencing the high-efficiency syntactical nodes according to the sequence from large to small of the corresponding response matching coefficient values, matching the first sequenced high-efficiency syntactical node with the corresponding demand instruction, marking the first sequenced high-efficiency syntactical node as an exclusive node of the corresponding demand instruction, and simultaneously marking the second sequenced high-efficiency syntactical node as a standby node of the corresponding demand instruction; marking the low-efficiency same-order node as a no-response node of the corresponding demand instruction, and transmitting the corresponding demand instruction and a demand instruction sending terminal to the exclusive node together if the no-response node receives the corresponding demand instruction;
sending the exclusive node, the standby node, the no-response node and the corresponding demand instruction to a cluster construction unit for storage;
after the cluster construction unit receives the exclusive node, the standby node and the no-response node and the corresponding demand instruction, an analysis maintenance signal is generated and sent to the analysis maintenance unit, the analysis maintenance unit is used for analyzing the operation of the enterprise cluster, and whether the operation state of the enterprise cluster is normal or not is judged, so that abnormal operation of the enterprise cluster is prevented, timely detection cannot be performed, business data loss is caused, the progress of business handling is obstructed, negative effects are brought to enterprise benefits, and the specific analysis maintenance process is as follows:
acquiring the time of establishing the enterprise cluster, and acquiring the real-time establishment duration of the enterprise cluster according to the time of establishing the enterprise cluster and the current time; acquiring the average fault time length and the minimum interval time length of adjacent faults in the real-time construction time length of the enterprise cluster, and respectively marking the average fault time length and the minimum interval time length of adjacent faults in the real-time construction time length of the enterprise cluster as PSC and DSC; by the formula
Acquiring a reliability coefficient SS of the enterprise cluster, wherein d1 and d2 are preset proportionality coefficients, and d1 is greater than d2 and is greater than 0;
comparing the reliability coefficient SS of the enterprise cluster with a reliability coefficient threshold:
if the reliability coefficient SS of the enterprise cluster is larger than or equal to the reliability coefficient threshold, judging that the reliability of the enterprise cluster is qualified, generating a cluster reliable signal and sending the cluster reliable signal to a cluster construction unit; if the reliability coefficient SS of the enterprise cluster is smaller than the reliability coefficient threshold value, judging that the reliability of the enterprise cluster is unqualified, generating a cluster unreliable signal and sending the cluster unreliable signal to a cluster construction unit, and maintaining a node corresponding to the unreliable signal in the enterprise cluster by the cluster construction unit;
acquiring the average maintenance time of the faults within the real-time construction duration of the enterprise cluster, and marking the average maintenance time of the faults within the real-time construction duration of the enterprise cluster as WSC; acquiring the times of completing maintenance in a short time according to the average maintenance time of the faults, and marking the times of completing maintenance in the short time of the faults as CWS; in the application, the fault is completed in a short time, namely the time for completing the maintenance of the fault is less than the average maintenance time; by the formula
Acquiring a maintainability coefficient KW of the enterprise cluster, wherein d3 and d4 are preset proportionality coefficients, and d3 is greater than d4 and is greater than 0;
comparing the maintainability coefficient KW of the enterprise cluster with a maintainability coefficient threshold: if the maintainability coefficient KW of the enterprise cluster is not less than the maintainability coefficient threshold, judging that the maintainability of the enterprise cluster is qualified, generating a maintainability signal and sending the maintainability signal to the cluster construction unit; if the maintainability coefficient KW of the enterprise cluster is smaller than the maintainability coefficient threshold, judging that the maintainability of the enterprise cluster is unqualified, generating an unremainable signal and sending the unremainable signal to a cluster construction unit, and performing equipment replacement on a node corresponding to the unremainable signal by the cluster construction unit;
calculating sum values of fault average maintenance time and fault average time in the real-time construction duration of the enterprise cluster, calculating ratio values of the fault average time and the corresponding sum values, marking the corresponding ratio values as availability coefficients, and comparing the availability coefficients of the enterprise cluster with availability coefficient threshold values: if the availability coefficient of the enterprise cluster is larger than or equal to the availability coefficient threshold, judging that the availability of the enterprise cluster is qualified, generating an available signal and sending the available signal to a cluster construction unit; if the availability coefficient of the enterprise cluster is smaller than the availability coefficient threshold value, judging that the availability of the enterprise cluster is unqualified, generating an unavailable signal and sending the unavailable signal to a cluster construction unit, and after receiving the unavailable signal, the cluster construction unit performs local area network replacement and node equipment replacement on the enterprise cluster;
after the cluster building unit receives the cluster reliable signal, the maintainable signal and the available signal, the cluster building unit generates a mirror fault-tolerant signal and sends the mirror fault-tolerant signal to the mirror fault-tolerant unit, the mirror fault-tolerant unit is used for carrying out data backup on an enterprise cluster, and the situation that data loss is caused by network intrusion or instruction misoperation and irreparable loss is caused to an enterprise is prevented, and the specific backup process is as follows:
monitoring business data and instruction responses of an enterprise cluster, arranging a sub-server in the enterprise cluster, storing and counting the business data and the instruction responses of the enterprise cluster by the sub-server, deleting the sub-server in a delayed manner when the server in the enterprise cluster deletes the data, generating a judgment misoperation instruction by a sending terminal of the misoperation instruction if the data is deleted to be instruction misoperation, sending the judgment misoperation instruction to the server, and copying corresponding data in the sub-server after the data is received by the server; simultaneously generating a mirror image completion signal and sending the mirror image completion signal to a cluster construction unit;
the method comprises the steps that a cluster building unit generates a mirror image completion signal to a server, the server generates a cluster operation signal and sends the cluster operation signal to a cluster operation platform, the cluster operation platform is used for detecting cluster operation, and if abnormality exists in detection, a building analysis instruction is generated and sent to the cluster building platform;
as shown in fig. 2:
a cluster building method based on AI consultation database comprises the following steps:
step one, enterprise analysis, namely analyzing the enterprises in real time, acquiring an enterprise real-time cluster building coefficient according to the business access amount, the business order number and the corresponding data flow, judging whether the enterprises carry out cluster building or not according to the enterprise real-time cluster building coefficient, and entering a step two if the enterprises carry out cluster building;
step two, cluster construction, namely setting each department in an enterprise as each node, carrying out communication connection on a server and each node, and carrying out network coverage on the server and each node by a local area network in the enterprise after the server is in communication connection with each node to construct an enterprise cluster;
thirdly, load balancing detection, namely dividing each node in the enterprise cluster, dividing the node into an exclusive node, a standby node and a no-response node according to a real-time demand instruction, and performing instruction distribution on each node according to a division type;
analyzing, maintaining and detecting the cluster, performing analysis, maintenance and detection after the enterprise cluster is constructed, and judging the real-time running state of the enterprise cluster by analyzing the reliability, maintainability and availability of the enterprise cluster;
and fifthly, setting cluster mirror image fault tolerance, namely setting a sub-server according to a server in the enterprise cluster, transmitting data in the server to a sub-server for storage, and setting a delayed deletion instruction for the sub-server, if the data in the server is deleted and judged to be in misoperation, generating a misoperation judgment instruction through a misoperation sending terminal, sending the misoperation judgment instruction to the server, and restoring the data through the sub-server by the server.
The working principle of the invention is as follows: a database cluster building method and system based on AI consultation database, when in work, a cluster building platform is used for building database clusters for each enterprise, a server generates enterprise analysis signals and sends the enterprise analysis signals to an enterprise analysis unit, and the enterprise analysis unit analyzes the operation of the enterprise after receiving the enterprise analysis signals; generating a cluster building signal according to the analysis, sending the cluster building signal to a server, and building an enterprise cluster through a cluster building unit after the server receives the cluster building signal; after the cluster construction unit completes the cluster construction of the corresponding enterprise, the cluster construction unit detects and maintains the enterprise cluster, generates a load balancing signal and sends the load balancing signal to the load balancing unit; analyzing the nodes in the enterprise cluster through a load balancing unit; analyzing the operation of the enterprise cluster through an analysis maintenance unit; and carrying out data backup on the enterprise cluster through a mirror fault-tolerant unit.
The above formulas are all calculated by taking the numerical value of the dimension, the formula is a formula which obtains the latest real situation by acquiring a large amount of data and performing software simulation, and the preset parameters in the formula are set by the technical personnel in the field according to the actual situation.
The foregoing is merely illustrative and explanatory of the present invention, and various modifications, additions or substitutions as would be apparent to one skilled in the art to the specific embodiments described are possible without departing from the invention as claimed herein or beyond the scope thereof.
Claims (2)
1. A cluster building system based on an AI consultation database is characterized by comprising a cluster building platform; the cluster building platform is provided with a server, the server is in communication connection with an enterprise analysis unit, a cluster building unit and a plurality of nodes, and the cluster building unit is in communication connection with a load balancing unit, an analysis maintenance unit and a mirror fault tolerance unit;
the method comprises the steps that a database cluster is built for each enterprise through a cluster building platform, a server generates enterprise analysis signals and sends the enterprise analysis signals to an enterprise analysis unit, and the enterprise analysis unit analyzes the operation of the enterprise after receiving the enterprise analysis signals; generating a cluster building signal according to the analysis, sending the cluster building signal to a server, and building an enterprise cluster through a cluster building unit after the server receives the cluster building signal; after the cluster construction unit completes the cluster construction of the corresponding enterprise, the cluster construction unit detects and maintains the enterprise cluster, generates a load balancing signal and sends the load balancing signal to the load balancing unit; analyzing the nodes in the enterprise cluster through a load balancing unit; analyzing the operation of the enterprise cluster through an analysis maintenance unit; carrying out data backup on the enterprise cluster through a mirror fault-tolerant unit;
the analysis process of the enterprise analysis unit is as follows:
acquiring the quantity of business orders of an enterprise in real time, marking the quantity of the business orders of the enterprise acquired in real time as SL, acquiring the business access quantity corresponding to the enterprise according to the quantity of the business orders of the enterprise, and marking the business access quantity corresponding to the enterprise as FW; acquiring data traffic generated by enterprise business through the quantity of business orders and the business access quantity, and marking the data traffic generated by the enterprise business as LL;
by the formula
Acquiring an enterprise real-time cluster building analysis coefficient X, wherein a1, a2 and a3 are preset proportionality coefficients, a1 is larger than a2 and larger than a3 is larger than 0, and comparing the enterprise real-time cluster building analysis coefficient X with a cluster building analysis coefficient threshold value:
if the real-time cluster building analysis coefficient X of the enterprise is larger than or equal to the cluster building analysis coefficient threshold value, judging that the corresponding enterprise needs to carry out cluster building, generating a cluster building signal and sending the cluster building signal to the server; if the real-time cluster building analysis coefficient X of the enterprise is smaller than the cluster building analysis coefficient threshold, judging that the corresponding enterprise does not need to carry out cluster building, generating a cluster non-building signal and sending the cluster non-building signal to the server;
the balancing process of the load balancing unit is as follows:
collecting each node in the enterprise cluster, and setting the node with the labels i, i =1,2, …, n, n as a natural number more than 1; acquiring a demand instruction received by each node, analyzing according to keywords in the demand instruction, and judging whether the demand instruction is of the same type;
marking the nodes receiving the same type of demand instructions as the same order nodes, and marking the same order nodes as u, wherein u belongs to i; acquiring the response time of the demand instruction processed by the same-order node, and marking the response time of the demand instruction processed by the same-order node as XYu; acquiring the receiving frequency of the repeated instruction after the demand instruction of the same-order node responds, and marking the receiving frequency of the repeated instruction after the demand instruction of the same-order node responds as PLu; the repeated instruction in the application is represented as the same demand instruction received by the same order node; by the formula
Acquiring a response matching coefficient Zu of the same order of nodes, wherein c1 and c2 are preset proportionality coefficients, c1 is larger than c2 and larger than 0, and beta is an error correction factor and takes a value of 2.01;
comparing the response matching coefficient of the concordance node with a response matching coefficient threshold:
if the response matching coefficient of the corresponding order node is not less than the response matching coefficient threshold, judging that the processing efficiency of the corresponding order node is high, and marking the corresponding order node as a high-efficiency order node; if the response matching coefficient of the corresponding syntype node is less than the response matching coefficient threshold value, judging that the processing efficiency of the corresponding syntype node is low, and marking the corresponding syntype node as a low-efficiency syntype node;
sequencing the high-efficiency syntactical nodes according to the sequence from large to small of the corresponding response matching coefficient values, matching the first sequenced high-efficiency syntactical node with the corresponding demand instruction, marking the first sequenced high-efficiency syntactical node as an exclusive node of the corresponding demand instruction, and simultaneously marking the second sequenced high-efficiency syntactical node as a standby node of the corresponding demand instruction; marking the low-efficiency same-order node as a no-response node of the corresponding demand instruction, and transmitting the corresponding demand instruction and a demand instruction sending terminal to the exclusive node together if the no-response node receives the corresponding demand instruction; sending the exclusive node, the standby node, the no-response node and the corresponding demand instruction to a cluster construction unit for storage;
the analysis maintenance process of the analysis maintenance unit is as follows:
acquiring the time of establishing the enterprise cluster, and acquiring the real-time establishment duration of the enterprise cluster according to the time of establishing the enterprise cluster and the current time; acquiring the average fault time length and the minimum interval time length of adjacent faults in the real-time construction time length of the enterprise cluster, and respectively marking the average fault time length and the minimum interval time length of adjacent faults in the real-time construction time length of the enterprise cluster as PSC and DSC; by the formula
Obtaining reliability coefficient of enterprise clusterSS, wherein d1 and d2 are both preset proportionality coefficients, and d1 is greater than d2 and is greater than 0, and the reliability coefficient SS of the enterprise cluster is compared with a reliability coefficient threshold value:
if the reliability coefficient SS of the enterprise cluster is larger than or equal to the reliability coefficient threshold, judging that the reliability of the enterprise cluster is qualified, generating a cluster reliable signal and sending the cluster reliable signal to a cluster construction unit; if the reliability coefficient SS of the enterprise cluster is smaller than the reliability coefficient threshold value, judging that the reliability of the enterprise cluster is unqualified, generating a cluster unreliable signal and sending the cluster unreliable signal to a cluster construction unit;
acquiring the average maintenance time of the faults within the real-time construction duration of the enterprise cluster, and marking the average maintenance time of the faults within the real-time construction duration of the enterprise cluster as WSC; acquiring the times of completing maintenance in a short time according to the average maintenance time of the faults, and marking the times of completing maintenance in the short time of the faults as CWS; by the formula
Acquiring a maintainability coefficient KW of the enterprise cluster, wherein d3 and d4 are preset proportionality coefficients, and d3 is greater than d4 and is greater than 0;
comparing the maintainability coefficient KW of the enterprise cluster with a maintainability coefficient threshold: if the maintainability coefficient KW of the enterprise cluster is not less than the maintainability coefficient threshold, judging that the maintainability of the enterprise cluster is qualified, generating a maintainability signal and sending the maintainability signal to the cluster construction unit; if the maintainability coefficient KW of the enterprise cluster is smaller than the maintainability coefficient threshold, judging that the maintainability of the enterprise cluster is unqualified, generating an unremainable signal and sending the unremainable signal to a cluster construction unit;
calculating the sum of the average maintenance time of faults and the average time of faults in the real-time construction time of the enterprise cluster, calculating the ratio of the average time of faults to the corresponding sum, marking the corresponding ratio as an availability coefficient, and comparing the availability coefficient of the enterprise cluster with an availability coefficient threshold value: if the availability coefficient of the enterprise cluster is larger than or equal to the availability coefficient threshold, judging that the availability of the enterprise cluster is qualified, generating an available signal and sending the available signal to a cluster construction unit; if the availability coefficient of the enterprise cluster is less than the availability coefficient threshold value, judging that the availability of the enterprise cluster is unqualified, generating an unavailable signal and sending the unavailable signal to a cluster construction unit;
the backup process of the image fault-tolerant signal is as follows:
monitoring business data and instruction responses of an enterprise cluster, arranging a sub-server in the enterprise cluster, storing and counting the business data and the instruction responses of the enterprise cluster by the sub-server, deleting the sub-server in a delayed manner when the server in the enterprise cluster deletes the data, generating a judgment misoperation instruction by a sending terminal of the misoperation instruction if the data is deleted to be instruction misoperation, sending the judgment misoperation instruction to the server, and copying corresponding data in the sub-server after the data is received by the server; and simultaneously generating a mirror image completion signal and sending the mirror image completion signal to the cluster building unit.
2. The cluster building method of the cluster building system based on the AI consultation database according to claim 1, wherein the specific cluster building method comprises the following steps:
step one, enterprise analysis, namely analyzing the enterprises in real time, acquiring an enterprise real-time cluster building coefficient according to the business access amount, the business order number and the corresponding data flow, judging whether the enterprises carry out cluster building or not according to the enterprise real-time cluster building coefficient, and entering a step two if the enterprises carry out cluster building;
step two, cluster construction, namely setting each department in an enterprise as each node, carrying out communication connection on a server and each node, and carrying out network coverage on the server and each node by a local area network in the enterprise after the server is in communication connection with each node to construct an enterprise cluster;
thirdly, load balancing detection, namely dividing each node in the enterprise cluster, dividing the node into an exclusive node, a standby node and a no-response node according to a real-time demand instruction, and performing instruction distribution on each node according to a division type;
analyzing, maintaining and detecting the cluster, performing analysis, maintenance and detection after the enterprise cluster is constructed, and judging the real-time running state of the enterprise cluster by analyzing the reliability, maintainability and availability of the enterprise cluster;
and fifthly, setting cluster mirror image fault tolerance, namely setting a sub-server according to a server in the enterprise cluster, transmitting data in the server to a sub-server for storage, and setting a delayed deletion instruction for the sub-server, if the data in the server is deleted and judged to be in misoperation, generating a misoperation judgment instruction through a misoperation sending terminal, sending the misoperation judgment instruction to the server, and restoring the data through the sub-server by the server.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111470286.6A CN114138750B (en) | 2021-12-03 | 2021-12-03 | AI consultation database based cluster building method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111470286.6A CN114138750B (en) | 2021-12-03 | 2021-12-03 | AI consultation database based cluster building method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114138750A CN114138750A (en) | 2022-03-04 |
| CN114138750B true CN114138750B (en) | 2022-10-18 |
Family
ID=80387644
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111470286.6A Active CN114138750B (en) | 2021-12-03 | 2021-12-03 | AI consultation database based cluster building method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114138750B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116095031B (en) * | 2023-04-10 | 2023-06-20 | 科讯嘉联信息技术有限公司 | RPA automatic micro-adding system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104680356A (en) * | 2013-11-27 | 2015-06-03 | 上海墨芋电子科技有限公司 | Middle and small-sized enterprise cooperative service platform of novel cloud architecture of novel technology |
| CN108717661A (en) * | 2018-04-20 | 2018-10-30 | 江苏大学 | A kind of cluster-based storage and analysis method of financial circles Risk-warning |
| CN112685055A (en) * | 2021-01-05 | 2021-04-20 | 北京金山云网络技术有限公司 | Cluster construction method and device |
| CN113157253A (en) * | 2021-05-06 | 2021-07-23 | 湖南中科星图信息技术股份有限公司 | Method for establishing enterprise research and development private cloud |
Family Cites Families (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7987152B1 (en) * | 2008-10-03 | 2011-07-26 | Gadir Omar M A | Federation of clusters for enterprise data management |
| CN102724292A (en) * | 2012-05-22 | 2012-10-10 | 中国科学院上海应用物理研究所 | Real-time database cluster architecture of control system of accelerator and establishment method thereof |
| CN103593420A (en) * | 2013-11-01 | 2014-02-19 | 浪潮电子信息产业股份有限公司 | Method for constructing heterogeneous database clusters on same platform by sharing online logs |
| CN104378239A (en) * | 2014-11-26 | 2015-02-25 | 国家电网公司 | Rapid reliability index statistical system and method based on cluster frame |
| CN104572895B (en) * | 2014-12-24 | 2018-02-23 | 天津南大通用数据技术股份有限公司 | MPP databases and Hadoop company-datas interoperability methods, instrument and implementation method |
| US10498748B1 (en) * | 2015-12-17 | 2019-12-03 | Skyhigh Networks, Llc | Cloud based data loss prevention system |
| CN107315803A (en) * | 2017-06-23 | 2017-11-03 | 郑州云海信息技术有限公司 | A kind of automatic building method of data-base cluster and system |
| CN110213358A (en) * | 2019-05-23 | 2019-09-06 | 深圳壹账通智能科技有限公司 | Method, node, equipment and the storage medium of cluster resource scheduling |
| CN110659180A (en) * | 2019-09-05 | 2020-01-07 | 国家计算机网络与信息安全管理中心 | Data center infrastructure management system based on cluster technology |
| CN112181940A (en) * | 2020-08-25 | 2021-01-05 | 天津农学院 | Method for constructing national industrial and commercial big data processing system |
| CN112988704A (en) * | 2021-03-05 | 2021-06-18 | 无锡星凝互动科技有限公司 | AI consultation database cluster building method and system |
| CN113205808A (en) * | 2021-04-27 | 2021-08-03 | 无锡星凝互动科技有限公司 | AI (Artificial intelligence) consultation based cloud computing building method and system |
| CN113222560A (en) * | 2021-06-01 | 2021-08-06 | 无锡星凝互动科技有限公司 | Enterprise service system based on AI consultation and construction method thereof |
| CN113269534A (en) * | 2021-06-09 | 2021-08-17 | 无锡星凝互动科技有限公司 | Network service platform based on enterprise consultation and construction method |
-
2021
- 2021-12-03 CN CN202111470286.6A patent/CN114138750B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104680356A (en) * | 2013-11-27 | 2015-06-03 | 上海墨芋电子科技有限公司 | Middle and small-sized enterprise cooperative service platform of novel cloud architecture of novel technology |
| CN108717661A (en) * | 2018-04-20 | 2018-10-30 | 江苏大学 | A kind of cluster-based storage and analysis method of financial circles Risk-warning |
| CN112685055A (en) * | 2021-01-05 | 2021-04-20 | 北京金山云网络技术有限公司 | Cluster construction method and device |
| CN113157253A (en) * | 2021-05-06 | 2021-07-23 | 湖南中科星图信息技术股份有限公司 | Method for establishing enterprise research and development private cloud |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114138750A (en) | 2022-03-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| RU2525859C2 (en) | Controlling power outage and faulty state of electrical power system | |
| RU2518178C2 (en) | System and method for control of electric power system | |
| RU2554540C2 (en) | Utility grid command filter system | |
| CN111176879A (en) | Fault repairing method and device for equipment | |
| JPH08307524A (en) | Method and equipment for discriminating risk in abnormal conditions of constitutional element of communication network | |
| CN103761309A (en) | Operation data processing method and system | |
| CN114138750B (en) | AI consultation database based cluster building method and system | |
| CN113159598A (en) | Method and device for rush repair and scheduling of transformer area faults and terminal equipment | |
| CN115550139B (en) | Fault root cause positioning method, device, system, electronic equipment and storage medium | |
| CN112712606B (en) | Automatic inspection management method and system based on operation and maintenance service | |
| Jensen et al. | Availability of a SCADA/OMS/DMS system—A case study | |
| CN112965990A (en) | Low-voltage contact cabinet fault solution generation method and device | |
| CN113760634A (en) | Data processing method and device | |
| CN111146868A (en) | Power communication machine room dynamic loop monitoring standardized access method and system | |
| CN110957724A (en) | Method for evaluating operation mode and grid structure of power distribution network | |
| CN114548769B (en) | Intelligent power grid IT asset big data monitoring system and method | |
| CN102045186A (en) | Event analysis method and system | |
| CN114493926A (en) | Line loss lean management and electricity larceny prevention method and system based on terminal edge calculation | |
| CN113391611B (en) | Early warning method, device and system for power environment monitoring system | |
| CN113468472A (en) | Power distribution network fault positioning method and system | |
| Hawkins | Characteristics of automated power system monitoring and management platforms | |
| CN113923096A (en) | Network element fault early warning method and device, electronic equipment and storage medium | |
| CN112215727A (en) | System for intelligently crawling and analyzing data of power system based on RPA technology | |
| CN117792903A (en) | Enterprise center service evaluation and dynamic treatment method based on deep reinforcement learning | |
| WO2024040976A1 (en) | Energy consumption control method and apparatus for network device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |