A collaborative method and system for automated deployment and operation and maintenance services of project software
By building a unified metadata model and intelligent decision-making, the problems of automated collaboration and risk assessment in the project software delivery process were solved, realizing full-process automation and refined risk control, and improving project delivery efficiency and quality.
Patent Information
- Application Number
- CN202611415770.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-11
- Publication Date
- 2026-11-17
AI Technical Summary
In existing technologies, the project software delivery process lacks an integrated automated collaboration mechanism across the entire chain, and the deployment and release process lacks intelligent risk assessment and adaptive control, resulting in low efficiency and potential quality risks.
Build a unified metadata model, conduct feasibility assessments based on basic project attribute information, dynamically match CI pipelines, adapt to decision-making target environments, deploy and verify health in real time, combine multi-dimensional testing to calculate risk values, adaptively match release strategies, and achieve fully automated workflow and intelligent decision-making.
It has achieved integrated and automated workflow for the entire software delivery process, improving deployment success rate and stability, finely controlling the risks in the release process, shortening the delivery cycle and improving quality and reliability.
Smart Images

Figure CN123114493A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of project management technology, and in particular to a collaborative method and system for automated deployment and operation and maintenance services of project software. Background Technology
[0002] In traditional software delivery models, development, testing, deployment, and operations are typically handled by different teams, resulting in significant process fragmentation and information silos between these stages. At project initiation, teams rely on manual experience and documentation for planning, making it difficult to quickly reuse best practices from previous projects. After development, build, deployment, and release operations heavily depend on manual execution by operations personnel, leading to inefficiency and susceptibility to deployment failures due to environmental differences or configuration errors.
[0003] While some automated deployment tools and operation and maintenance monitoring platforms exist in the current technology, they still have the following shortcomings: 1. Lack of a fully integrated automated collaboration mechanism across the entire process. Existing solutions typically only cover a specific stage of software delivery. For example, continuous integration tools are only responsible for code building and testing, deployment scripts are only responsible for application release, and operation and maintenance monitoring platforms only focus on online operational status. The entire process of a project, from initialization, planning, pipeline construction, deployment environment selection, release strategy decision-making to operation and maintenance troubleshooting, lacks a unified metadata-driven and automatic triggering mechanism. The connection between each stage still requires manual intervention, making it difficult to achieve end-to-end automated flow and collaboration.
[0004] 2. The deployment and release process lacks intelligent risk assessment and adaptive control. Existing systems rely heavily on the experience and judgment of operations personnel in areas such as deployment environment selection, resource quota allocation, and health checks. They lack quantitative assessment models based on project phases, resource requirements, and real-time environmental load, making it difficult to achieve optimal decision-making for the target environment and elastic resource scheduling. Furthermore, the release phase lacks systematic quantitative calculation of deployment risks and cannot adaptively match differentiated release strategies according to risk levels, resulting in significant uncertainty and quality risks during the release process. Summary of the Invention
[0005] The purpose of this invention is to provide a collaborative method and system for automated deployment and operation and maintenance services of project software, so as to solve the problems mentioned in the background art.
[0006] To achieve the above objectives, this invention provides a collaborative method for automated deployment and operation and maintenance services of project software, comprising the following steps: S1. Obtain the basic attribute information of the project input by the user, conduct a feasibility assessment of the project based on the basic attribute information, and after the assessment is passed, build a unified metadata model and initialize it, and bind the automatic triggering rules corresponding to each project node. S2. Dynamically match and generate CI pipelines based on a unified metadata model, and perform full-process verification; S3. Automatically trigger deployment instructions based on verification results, make adaptive decisions on the target environment based on project stage, resource requirements and real-time environmental load, dynamically calculate resource quotas and orchestrate infrastructure, perform deployment and verify deployment health in real time. S4. Conduct multi-dimensional testing on the project based on the deployment results, calculate the final deployment risk value based on the test results, adaptively match the release strategy according to the risk value, and execute the release. S5. Dynamically configure inspection strategies and alarm rules according to project phases to achieve the correlation and location of alarms or service work orders with fault locations; S6. Based on the location of the associated fault, repair the fault.
[0007] Preferably, step S1 specifically includes: S11. Obtain the basic attribute information of the project input by the user, including the project's unique code, the field to which the project belongs, the project type, the project priority, the product line to which it belongs, the expected delivery cycle, the project scale level, the technology stack type, and human and resource constraints. Among them, the project type includes new projects, iterative upgrade projects, patch repair projects, and migration projects. S12. Convert the project's basic attribute information into a standardized project feature vector. Based on a pre-set historical project database, calculate the feature similarity between the current project and all historically completed projects using a similarity matching algorithm, and select the project with the highest similarity. One project serves as a reference template; S13. Extract the WBS task structure, phase division rules, standard task duration, and typical deliverable list from the reference template to generate the initial planning framework for the current project. Based on the initial planning framework, conduct a feasibility evaluation according to the technology stack adaptability, resource capacity satisfaction, and historical delivery success rate. S14. After the feasibility assessment is passed, based on the matching reference template and combined with the total project delivery cycle, the plan is automatically divided into phases, the deliverables and completion standards of each phase are clarified, a first-level project plan is generated, and each phase of the first-level plan is broken down into executable task items to generate a second-level project plan. S15. Based on the project's secondary plan, map all project basic attributes, phase nodes, task lists, deliverable standards, configuration parameters, and product BOM dependency lists into the unified metadata model, initialize the unified metadata model, bind the automated triggering rules corresponding to each project node, and clarify the actions that need to be triggered after the node is achieved.
[0008] Preferably, the formula for calculating the similarity of project features is: ; In the formula, Indicates the current project and the first The feature similarity of each historical project, with a value ranging from 0 to 1. This indicates the total number of feature dimensions for the project. Indicates the first The weight coefficients of each feature dimension, with the sum of the weights of all dimensions being 1; Indicates the current project is in the [number]th [year]. Quantization values on each feature dimension; Indicates the first The first historical project in Quantization values on each feature dimension.
[0009] Preferably, step S2 specifically includes: S21. Based on the unified metadata model, retrieve the basic project attribute information of the corresponding project, including technology stack type, code repository address, product BOM dependency list, and project stage; S22. Call the pipeline template library, calculate the compatibility of each template with the current project through the template matching algorithm, select the optimal pipeline template, fill in the project parameters, and generate a dedicated CI pipeline. S23. Based on a dedicated CI pipeline, code retrieval, dependency installation, compilation and building, unit testing, code quality scanning and security vulnerability detection are performed sequentially according to the process. For each sub-step, the overall quality score of the current build is judged through a dynamic quality gate model, and each step is verified. S24. After all stages of verification are passed, a standardized product is generated. A unique version identifier for the product is generated according to unified rules, including the version identifier associated with the project ID, stage number, build sequence number, and code submission hash value. S25. Calculate the performance indicators of this build. If the quality is qualified, issue the deployment command and store the standardized product in the historical project database. If the quality is unqualified, roll back to the development task node.
[0010] Preferably, the overall quality score determined by the dynamic quality access control model includes: The formula for constructing the comprehensive quality score is as follows: ; In the formula, This indicates the overall quality score of the current construction. This indicates the code quality score; Indicates the test coverage score; Indicates the safety compliance score; , , Represents the dimension weight coefficient. ; The formula for calculating the quality access control threshold is: ; In the formula, This indicates the quality acceptance threshold for the current project phase. Indicates the benchmark pass threshold. Indicates the stage correction factor; The verification was successful.
[0011] Preferably, step S3 specifically includes: S31. Based on the unified metadata model, obtain information on project stage, product version, resource requirements, configuration parameters, and deployment scope, and obtain real-time load, remaining resources, architecture type, and security level data of available target environments; S32. Based on the acquired data, calculate the matching score between each available environment and the deployment requirements, and obtain the optimal target environment. The environment adaptability evaluation formula is: ; In the formula, This represents the suitability score for the candidate environment, with a value ranging from 0 to 1. This represents the resource satisfaction level, which is the ratio of the remaining resources in the candidate environment to the resources required for deployment. It is set to 1 when it is greater than or equal to 1, and the value is taken as the actual ratio when it is less than 1. Indicates the degree of matching at each stage; Indicates the health status of the load. Indicates the degree of security level matching; , , , Indicates the weighting coefficient. ; S33. Based on the defined target environment, dynamically calculate dynamic resource quotas and orchestrate infrastructure according to project scale and historical resource utilization rates. Infrastructure includes computing resources, storage resources, network resources, and operating environment. The formula for calculating dynamic resource quotas is as follows: ; In the formula, This indicates the resource quota allocated for this deployment; This represents the baseline resource quota, which corresponds to the standard resource allocation for a standard-sized project. Indicates the project size coefficient; This represents the historical utilization rate correction factor, calculated using the following formula: ,in, To achieve the target resource utilization rate, This represents the historical average resource utilization rate for similar projects at the same stage. S34. A batch rolling release strategy is adopted for deployment. For each batch, a health check is performed. Based on the node's liveness status, interface response rate, and error rate data, the deployment health of a single batch is calculated. When the health is higher than the set threshold, the deployment of this batch is determined to be successful and the next batch continues; otherwise, the deployment is automatically paused and the abnormal rollback logic is triggered. S35. After the full batch deployment is completed, a full health check will be performed.
[0012] Preferably, step S4 specifically includes: S41. Conduct multi-dimensional testing on the project based on the deployment results, including functional automation testing, performance stress testing, interface integration testing, and business scenario compatibility testing, and obtain multi-dimensional scores based on the test results; S42. Calculate the final deployment risk value based on the multi-dimensional scoring results; S43. Based on the final deployment risk value, the deployment is divided into low risk, medium risk and high risk. The preset release strategy is matched according to the risk results and the deployment is carried out according to the preset release strategy.
[0013] Preferably, the final deployment risk value calculation formula is as follows: ; in, The basic risk value is represented by the following formula: ; In the formula, The final deployment overall risk value ranges from 0 to 1; P represents the project phase weighting coefficient; S represents the change scope percentage, ranging from 0 to 1, indicating the proportion of modules involved in this change to the total number of modules in the project; F represents the historical failure rate of similar deployments, ranging from 0 to 1. , , Represents the weighting coefficients, satisfying ; This represents the correction coefficient for the verification result.
[0014] Preferably, step S5 specifically includes: S51. Obtain multi-dimensional information on project stage, business level, and deployment architecture based on a unified metadata model; S52. Set inspection strategies, inspection frequencies, alarm thresholds and emergency plans based on multi-dimensional information of the project. S53. When a monitoring alarm or a user service ticket is received, the version information, deployment records, configuration parameters, and change history of the corresponding project are obtained from the unified metadata model. The matching degree between the current alarm / ticket and historical fault cases is calculated through a similarity matching algorithm, and the investigation clues and possible root cause directions are output. S54. Based on the matching results, associate the alarm or service ticket with the fault location.
[0015] This invention also provides a collaborative system for automated deployment and operation and maintenance services of project software, which executes a collaborative method for automated deployment and operation and maintenance services of project software, including: The central engine connects to each module and serves as the system's data hub and scheduling core. It is used to uniformly store end-to-end metadata, monitor project nodes, issue execution instructions, and transfer data. The project initialization module is used to obtain the basic attribute information of the project input by the user, conduct a feasibility assessment of the project based on the basic attribute information, and after the assessment is passed, build a unified metadata model and initialize it, and bind the automatic triggering rules corresponding to each project node. The project generation module is used to dynamically match and generate CI pipelines based on a unified metadata model and perform full-process verification. The deployment module is used to automatically trigger deployment instructions based on the verification results. It adaptively decides the target environment based on the project stage, resource requirements and real-time environmental load, dynamically calculates resource quotas and orchestrates infrastructure, performs deployment and verifies deployment health in real time. The deployment module is used to perform multi-dimensional testing on the project based on the deployment results, calculate the final deployment risk value based on the test results, adaptively match the deployment strategy according to the risk value, and execute the deployment. The operation and maintenance service module is used to dynamically configure inspection strategies and alarm rules according to the project stage, realize the association and location of alarms or service work orders with fault locations, and repair faults.
[0016] Therefore, the present invention, employing the above-described collaborative method and system for automated deployment and operation and maintenance of project software, has the following beneficial effects: (1) This invention constructs a unified metadata model to include all information such as basic project attributes, stage nodes, task lists, deliverable standards, configuration parameters, and product BOM dependency lists into unified management, and binds the corresponding automated triggering rules for each project node. From project initialization assessment and plan generation to CI pipeline matching and construction verification, and then to deployment, release, and operation and maintenance, the entire chain is driven by unified metadata, realizing the integrated and automated flow of the entire project software delivery process, eliminating information transmission barriers and manual connection costs between each link.
[0017] (2) In the deployment phase, the present invention comprehensively evaluates the resource satisfaction, phase matching, load health and security level matching of candidate environments and adaptively decides the optimal target environment; at the same time, it dynamically calculates resource quotas based on project scale and historical resource utilization to achieve elastic resource scheduling, adopts a batch rolling release strategy, verifies deployment health in real time for each batch, and automatically pauses and triggers rollback when abnormal, which significantly improves the success rate and stability of deployment. (3) This invention obtains verification scores through multi-dimensional testing, combines factors such as project stage weights and historical similar deployment failure rates, quantifies the final deployment risk value, and automatically matches differentiated release strategies such as full release, gray release or manual review according to the risk level, thereby achieving refined risk control in the release process.
[0018] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description
[0019] Figure 1 This is a flowchart illustrating a collaborative method for automated deployment and operation and maintenance services of project software according to an embodiment of the present invention. Figure 2 This is a system architecture diagram of a collaborative system for automated deployment and operation and maintenance services of project software, according to an embodiment of the present invention. Detailed Implementation
[0020] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.
[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.
[0022] Example like Figure 1 As shown, this invention provides a collaborative method for automated deployment and operation and maintenance services of project software, including the following steps: S1. Obtain the basic project attribute information input by the user, conduct a feasibility assessment of the project based on the basic project attribute information, and after the assessment is passed, build and initialize a unified metadata model, and bind the corresponding automatic triggering rules for each project node. Specifically, this includes: S11. Obtain the basic attribute information of the project input by the user, including the project's unique code, the field to which the project belongs, the project type, the project priority, the product line to which it belongs, the expected delivery cycle, the project scale level, the technology stack type, and human and resource constraints. Among them, the project type includes new projects, iterative upgrade projects, patch repair projects, and migration projects. S12. Convert the project's basic attribute information into a standardized project feature vector. Based on a pre-set historical project database, calculate the feature similarity between the current project and all historical completed projects using a similarity matching algorithm. Select the three projects with the highest similarity as reference templates. The formula for calculating project feature similarity is as follows: ; In the formula, Indicates the current project and the first The feature similarity of each historical project, with a value ranging from 0 to 1. This indicates the total number of feature dimensions for the project. Indicates the first The weight coefficients of each feature dimension, with the sum of the weights of all dimensions being 1; Indicates the current project is in the [number]th [year]. Quantization values on each feature dimension; Indicates the first The first historical project in Quantization values on each feature dimension.
[0023] S13. Extract the WBS task structure, phase division rules, standard task duration, and typical deliverable list from the reference template to generate the initial planning framework for the current project. Based on the initial planning framework, conduct a feasibility evaluation according to the technology stack adaptability, resource capacity satisfaction, and historical delivery success rate.
[0024] Technology stack compatibility calculation involves analyzing the number of personnel required to cover the technology stack needed for the current project, the number of tool licenses, and the delivery experience of similar technology projects in the past to calculate the technology stack compatibility. : ; In the formula, This indicates the number of on-the-job personnel with the required skills in the target technology stack. This indicates the total number of personnel planned for the project. This indicates the number of target technology stack tool types that the company has covered; This represents the total number of different types of technology stack tools required for the project. , This represents the weighting coefficient; the sum of the two equals 1.
[0025] Resource capacity satisfaction calculation involves obtaining the total available computing power, manpower quota, and environmental resources within the project cycle, and then calculating the resource capacity satisfaction level. : ; In the formula, This represents the total amount of computing resources available during the project lifecycle; This indicates the total amount of computing resources required by the project plan; This represents the total available man-hours during the project period. This indicates the total man-hours required for the project plan; Indicates the number of deployment environments available during the project lifecycle; This indicates the number of deployment environments required by the project plan.
[0026] Historical delivery success rate calculation involves filtering similar projects with a similarity ≥ 0.7 to the current project, calculating their on-time delivery rate and first-time acceptance pass rate, and then calculating the historical delivery success rate. : ; In the formula, The number of similar historical projects delivered on schedule; This indicates the total number of similar historical items included in the statistics; This indicates the number of projects that passed acceptance testing in one go within the same historical project category; , Represents the weighting coefficients, satisfying The values are 0.5 and 0.5 respectively.
[0027] The overall feasibility score of the project is obtained by weighting and summing the indicators from the three dimensions. : ; In the formula, , , These represent weighting coefficients, and the sum of the three should equal 1. Preferred values are 0.3, 0.4, and 0.3. , , The values range from 0 to 1.
[0028] Preset three-level judgment rules: If deemed feasible, the process will automatically proceed to the plan refinement stage. If a risk is identified, the system will automatically output specific risk points and optimization suggestions, and submit them for manual review. If deemed infeasible, the project initiation process will be automatically terminated and an infeasibility analysis report will be generated.
[0029] S14. After the feasibility assessment is passed, based on the matching reference template and the total project delivery cycle, the system automatically divides the project into phases, clarifies the deliverables and completion standards for each phase, generates a primary project plan, and breaks down each phase of the primary plan into actionable tasks, generating a secondary project plan. The standard duration for each phase is dynamically adjusted using a schedule adjustment formula. ; In the formula, Indicates the current project number The planned duration for each phase; Indicates the first in the history template Standard project duration for each stage; This represents the project scale coefficient, which is determined based on the project's required scale and code volume, and ranges from 0.8 to 1.5, with higher values for larger scale projects. This represents the resource adjustment coefficient, derived from the resource capacity satisfaction level B. When resources are insufficient, the project period is automatically extended to ensure the plan can be implemented.
[0030] Once the primary and secondary plans are generated, they are automatically sent to the project manager for confirmation; after confirmation, the project plan is officially issued.
[0031] S15. Based on the project's secondary plan, map all project basic attributes, phase nodes, task lists, deliverable standards, configuration parameters, and product BOM dependency lists into the unified metadata model, initialize the unified metadata model, bind the automated triggering rules corresponding to each project node, and clarify the actions that need to be triggered after the node is achieved.
[0032] This invention selects the optimal reference template from a historical project database using a similarity matching algorithm, automatically gathers multiple pieces of information, generates an initial planning framework, and performs a quantitative feasibility evaluation based on three dimensions: technology stack adaptability, resource capacity satisfaction, and historical delivery success rate. Upon successful evaluation, the project is automatically divided into phases, and primary and secondary plans are generated, significantly improving the intelligence and planning efficiency of the project initiation phase and reducing repetitive manual labor.
[0033] S2. Dynamically generate CI pipelines based on a unified metadata model and perform full-process validation. Specifically, this includes: S21. Based on the unified metadata model, retrieve the basic project attribute information of the corresponding project, including technology stack type, code repository address, product BOM dependency list, and project stage; S22. Call the pipeline template library, calculate the fit between each template and the current project using a template matching algorithm, select the optimal pipeline template, fill in the project parameters, and generate a dedicated CI pipeline. The fit between each template and the current project is calculated based on technology stack matching, project stage matching, and project size matching. Technology stack matching: 1 if the template's supported technology stack is completely identical to the project's technology stack, 0.5 for partial compatibility, and 0 for no support. Project stage matching: 1 if the template's corresponding stage matches the project's current stage, 0.6 for adjacent stages, and 0.2 for no match. Project size matching: 1 if the template's size matches the project's size level, 0.7 for a one-level deviation, and 0.3 for two or more levels of deviation.
[0034] S23. Based on a dedicated CI pipeline, the code retrieval, dependency installation, compilation and building, unit testing, code quality scanning, and security vulnerability detection are performed sequentially according to the process. For each sub-step, a dynamic quality gate model is used to determine the overall quality score of the current build, and each step is verified. The dynamic quality gate model is used to determine the overall quality score of the current build, including: The formula for constructing the comprehensive quality score is as follows: ; In the formula, This represents the overall quality score of the current construction, ranging from 0 to 1. A higher score indicates better construction quality. This represents the code quality score, calculated based on code duplication rate, cyclomatic complexity, and number of code quality issues, and ranges from 0 to 1. This represents the test coverage score, which is calculated by weighting unit test coverage rate and test case pass rate, and has a value of 0 to 1. The security compliance score is calculated based on the number of high-risk vulnerabilities, the number of medium-risk vulnerabilities, and the dependency risk level, and the value ranges from 0 to 1. , , Represents the dimension weight coefficient. The values are 0.3, 0.4, and 0.3.
[0035] The formula for calculating the quality access control threshold is: ; In the formula, This indicates the quality acceptance threshold for the current project phase. Indicates the benchmark pass threshold. Indicates the stage correction factor; The verification was successful.
[0036] S24. After all stages of verification are passed, a standardized product is generated. A unique version identifier for the product is generated according to unified rules, including the version identifier associated with the project ID, stage number, build sequence number, and code submission hash value.
[0037] S25. Calculate the performance indicators for this build. If the quality is satisfactory, issue deployment instructions and store the standardized product in the historical project database. If it fails, roll back to the development task node. The performance indicator calculation formula is: ; In the formula, This represents the performance index; a higher value indicates better construction efficiency. This represents the standard build time for similar projects at the same stage, taken from historical project statistics. This indicates the actual time taken for this construction. This indicates a successful construction status, with 1 for success and 0 for failure.
[0038] S3. Automatically triggers deployment commands based on verification results, adaptively decides the target environment based on project stage, resource requirements, and real-time environmental load, dynamically calculates resource quotas and orchestrates infrastructure, performs deployment, and verifies deployment health in real time. Specifically, this includes: S31. Based on the unified metadata model, obtain information on project stage, product version, resource requirements, configuration parameters, and deployment scope, and obtain real-time load, remaining resources, architecture type, and security level data of available target environments; S32. Based on the acquired data, calculate the matching score between each available environment and the deployment requirements, and obtain the optimal target environment. The environment adaptability evaluation formula is: ; In the formula, This represents the suitability score for the candidate environment, with a value ranging from 0 to 1. This represents the resource satisfaction level, which is the ratio of the remaining resources in the candidate environment to the resources required for deployment. It is set to 1 when it is greater than or equal to 1, and the value is taken as the actual ratio when it is less than 1. Indicates the degree of matching at each stage; Indicates the health status of the load. Indicates the degree of security level matching; , , , Indicates the weighting coefficient. The values are 0.4, 0.2, 0.2, and 0.2.
[0039] S33. Based on the defined target environment, dynamically calculate dynamic resource quotas and orchestrate infrastructure according to project scale and historical resource utilization rates. Infrastructure includes computing resources, storage resources, network resources, and operating environment. The formula for calculating dynamic resource quotas is as follows: ; In the formula, This indicates the resource quota allocated for this deployment; This represents the baseline resource quota, which corresponds to the standard resource allocation for a standard-sized project. Indicates the project size coefficient; This represents the historical utilization rate correction factor, calculated using the following formula: ,in, To achieve the target resource utilization rate, This represents the historical average resource utilization rate for similar projects at the same stage.
[0040] After the infrastructure is orchestrated, an environmental consistency check is performed. The current environmental configuration is compared with the standard baseline, and an environmental consistency score is calculated. When the consistency score is ≥0.95, the environment is deemed qualified and proceeds to the deployment phase; otherwise, the discrepancies are automatically repaired until the consistency requirements are met.
[0041] S34. A batch-based rolling deployment strategy is adopted. For each batch, a health check is performed. Based on the node's liveness status, interface response rate, and error rate, the deployment health of a single batch is calculated. If the health score is higher than a set threshold, the deployment of this batch is considered successful, and the next batch continues; otherwise, deployment is automatically paused, triggering abnormal rollback logic. The formula for the single-batch health score is: ; In the formula, This indicates the health status of a single batch of deployments, with a value ranging from 0 to 1. This represents the node survival rate, which is the number of surviving nodes divided by the total number of nodes in the batch. This indicates the success rate of the API request, with a value ranging from 0 to 1. This represents the normalized error rate value; the lower the error rate, the higher the score. , , This represents the weighting coefficients, which must add up to 1. The default values are 0.4, 0.4, and 0.2.
[0042] when If the deployment is successful, the next batch will proceed; otherwise, deployment will be automatically paused, triggering an exception rollback.
[0043] S35. After the full batch deployment is completed, a full health check will be performed.
[0044] S4. Conduct multi-dimensional testing on the project based on the deployment results, calculate the final deployment risk value based on the test results, adaptively match the release strategy according to the risk value, and execute the release. Specifically, this includes: S41. Conduct multi-dimensional testing on the project based on the deployment results, including functional automation testing, performance stress testing, interface integration testing, and business scenario compatibility testing, and obtain multi-dimensional scores based on the test results. ,when If the verification is successful, proceed to the risk assessment stage; otherwise, if the verification fails, automatically trigger a deployment rollback to the previous stable version.
[0045] S42. Calculate the final deployment risk value based on the multi-dimensional scoring results. The formula for calculating the final deployment risk value is: ; in, The basic risk value is represented by the following formula: ; In the formula, The final deployment comprehensive risk value ranges from 0 to 1; P represents the project stage weighting coefficient, with 0.2 for the solution stage, 0.5 for the trial stage, 0.7 for the finalization stage, and 1.0 for the trial production stage; S represents the change scope ratio, ranging from 0 to 1, representing the proportion of modules involved in this change to the total number of modules in the project; F represents the historical failure rate of similar deployments, ranging from 0 to 1. , , Represents the weighting coefficients, satisfying The values are 0.3, 0.4, and 0.3 respectively. This represents the correction factor for the verification result, with a value of 0.3. It is used to reflect the amplifying effect of verification quality on risk; the lower the verification score, the higher the final risk value.
[0046] S43. Based on the final deployment risk value, the deployment is divided into low risk, medium risk and high risk. The preset release strategy is matched according to the risk results and the deployment is carried out according to the preset release strategy.
[0047] Low risk, automatic full release.
[0048] Medium risk, automatically implements gray-scale release, gradually increasing the volume proportionally.
[0049] High risk: Automatically upgrade the approval level, suspend the release process, and send it to manual review.
[0050] S5. Dynamically configure inspection strategies and alarm rules according to project phases to achieve the correlation and location of alarms or service orders with fault locations. Specifically, this includes: S51. Obtain multi-dimensional information on project stage, business level, and deployment architecture based on a unified metadata model; S52. Set inspection strategies, inspection frequencies, alarm thresholds and emergency plans based on multi-dimensional information of the project. S53. When a monitoring alarm or a user service ticket is received, the system retrieves the version information, deployment records, configuration parameters, and change history of the corresponding project from the unified metadata model. Using a similarity matching algorithm, it calculates the matching degree between the current alarm / ticket and historical fault cases, outputting investigation clues and possible root cause directions. The alarm / fault matching degree formula is as follows: ; In the formula, This indicates the degree of matching between the current alarm and historical fault cases, with a value ranging from 0 to 1; This indicates version similarity. A score of 1 is given when the versions are the same, 0.6 is given when the major versions are the same but the minor versions are different, and 0.2 is given when the versions are different. The similarity of modules is represented by a value of 1 when the modules involved in the fault are completely identical, 0.5 when there is partial overlap, and 0 when they are unrelated. This represents the similarity of fault phenomena, calculated by keyword matching, and ranges from 0 to 1. , , Represents the weighting coefficients, satisfying The values are 0.3, 0.4, and 0.3.
[0051] Select the three historical cases with the highest matching degree and automatically associate them with the corresponding solutions to help maintenance personnel quickly troubleshoot problems.
[0052] S54. Based on the matching results, associate the alarm or service ticket with the fault location.
[0053] This step significantly shortens the fault location time and improves the efficiency of operation and maintenance troubleshooting.
[0054] S6. Based on the location of the associated fault, repair the fault.
[0055] like Figure 2 As shown, the present invention also provides a collaborative system for automated deployment and operation and maintenance services of project software, used to execute a collaborative method for automated deployment and operation and maintenance services of project software, including: The central engine connects to each module and serves as the system's data hub and scheduling core. It is used to uniformly store end-to-end metadata, monitor project nodes, issue execution instructions, and transfer data. The project initialization module is used to obtain the basic attribute information of the project input by the user, conduct a feasibility assessment of the project based on the basic attribute information, and after the assessment is passed, build a unified metadata model and initialize it, and bind the automatic triggering rules corresponding to each project node. The project generation module is used to dynamically match and generate CI pipelines based on a unified metadata model and perform full-process verification. The deployment module is used to automatically trigger deployment instructions based on the verification results. It adaptively decides the target environment based on the project stage, resource requirements and real-time environmental load, dynamically calculates resource quotas and orchestrates infrastructure, performs deployment and verifies deployment health in real time. The deployment module is used to perform multi-dimensional testing on the project based on the deployment results, calculate the final deployment risk value based on the test results, adaptively match the deployment strategy according to the risk value, and execute the deployment. The operation and maintenance service module is used to dynamically configure inspection strategies and alarm rules according to the project stage, realize the association and location of alarms or service work orders with fault locations, and repair faults.
[0056] Therefore, the present invention adopts the above-mentioned collaborative method and system for automated deployment and operation and maintenance services of project software. Through the automated operation and intelligent decision-making of the whole process, it reduces the manual intervention links and human errors, shortens the project delivery cycle from development to online operation, and at the same time, by constructing multiple quality assurance mechanisms such as quality access control, deployment health verification, risk quantification assessment, and intelligent troubleshooting of operation and maintenance, a full-link quality closed loop from submission to online operation is formed, which effectively improves the quality and reliability of software delivery.
[0057] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.
Claims
1. A collaborative method for automated deployment and operation and maintenance services of project software, characterized in that, Includes the following steps: S1. Obtain the basic attribute information of the project input by the user, conduct a feasibility assessment of the project based on the basic attribute information, and after the assessment is passed, build a unified metadata model and initialize it, and bind the automatic triggering rules corresponding to each project node. S2. Dynamically match and generate CI pipelines based on a unified metadata model, and perform full-process verification; S3. Automatically trigger deployment instructions based on verification results, make adaptive decisions on the target environment based on project stage, resource requirements and real-time environmental load, dynamically calculate resource quotas and orchestrate infrastructure, perform deployment and verify deployment health in real time. S4. Conduct multi-dimensional testing on the project based on the deployment results, calculate the final deployment risk value based on the test results, adaptively match the release strategy according to the risk value, and execute the release. S5. Dynamically configure inspection strategies and alarm rules according to project phases to achieve the correlation and location of alarms or service work orders with fault locations; S6. Based on the location of the associated fault, repair the fault.
2. The collaborative method for automated deployment and operation and maintenance services of project software according to claim 1, characterized in that, Step S1 specifically includes: S11. Obtain the basic attribute information of the project input by the user, including the project's unique code, the field to which the project belongs, the project type, the project priority, the product line to which it belongs, the expected delivery cycle, the project scale level, the technology stack type, and human and resource constraints. Among them, the project type includes new projects, iterative upgrade projects, patch repair projects, and migration projects. S12. Convert the project's basic attribute information into a standardized project feature vector. Based on a pre-set historical project database, calculate the feature similarity between the current project and all historically completed projects using a similarity matching algorithm, and select the project with the highest similarity. One project serves as a reference template; S13. Extract the WBS task structure, phase division rules, standard task duration, and typical deliverable list from the reference template to generate the initial planning framework for the current project. Based on the initial planning framework, conduct a feasibility evaluation according to the technology stack adaptability, resource capacity satisfaction, and historical delivery success rate. S14. After the feasibility assessment is passed, based on the matching reference template and combined with the total project delivery cycle, the plan is automatically divided into phases, the deliverables and completion standards of each phase are clarified, a first-level project plan is generated, and each phase of the first-level plan is broken down into executable task items to generate a second-level project plan. S15. Based on the project's secondary plan, map all project basic attributes, phase nodes, task lists, deliverable standards, configuration parameters, and product BOM dependency lists into the unified metadata model, initialize the unified metadata model, bind the automated triggering rules corresponding to each project node, and clarify the actions that need to be triggered after the node is achieved.
3. The collaborative method for automated deployment and operation and maintenance services of project software according to claim 2, characterized in that, The formula for calculating the similarity of project features is: ; In the formula, Indicates the current project and the first The feature similarity of each historical project, with a value ranging from 0 to 1. This indicates the total number of feature dimensions for the project. Indicates the first The weight coefficients of each feature dimension, with the sum of the weights of all dimensions being 1; Indicates the current project is in the [number]th [year]. Quantization values on each feature dimension; Indicates the first The first historical project in Quantization values on each feature dimension.
4. The collaborative method for automated deployment and operation and maintenance services of project software according to claim 2, characterized in that, Step S2 specifically includes: S21. Based on the unified metadata model, retrieve the basic project attribute information of the corresponding project, including technology stack type, code repository address, product BOM dependency list, and project stage; S22. Call the pipeline template library, calculate the compatibility of each template with the current project through the template matching algorithm, select the optimal pipeline template, fill in the project parameters, and generate a dedicated CI pipeline. S23. Based on a dedicated CI pipeline, code retrieval, dependency installation, compilation and building, unit testing, code quality scanning and security vulnerability detection are performed sequentially according to the process. For each sub-step, the overall quality score of the current build is judged through a dynamic quality gate model, and each step is verified. S24. After all stages of verification are passed, a standardized product is generated. A unique version identifier for the product is generated according to unified rules, including the version identifier associated with the project ID, stage number, build sequence number, and code submission hash value. S25. Calculate the performance indicators of this build. If the quality is qualified, issue the deployment command and store the standardized product in the historical project database. If the quality is unqualified, roll back to the development task node.
5. The collaborative method for automated deployment and operation and maintenance services of project software according to claim 4, characterized in that: The dynamic quality access control model determines the overall quality score of the current system, including: The formula for constructing the comprehensive quality score is as follows: ; In the formula, This indicates the overall quality score of the current construction. This indicates the code quality score; Indicates the test coverage score; Indicates the safety compliance score; , , Represents the dimension weight coefficient. ; The formula for calculating the quality access control threshold is: ; In the formula, This indicates the quality acceptance threshold for the current project phase. Indicates the benchmark pass threshold. Indicates the stage correction factor; The verification was successful.
6. The collaborative method for automated deployment and operation and maintenance services of project software according to claim 5, characterized in that, Step S3 specifically includes: S31. Based on the unified metadata model, obtain information on project stage, product version, resource requirements, configuration parameters, and deployment scope, and obtain real-time load, remaining resources, architecture type, and security level data of available target environments; S32. Based on the acquired data, calculate the matching score between each available environment and the deployment requirements, and obtain the optimal target environment. The environment adaptability evaluation formula is: ; In the formula, This represents the suitability score for the candidate environment, with a value ranging from 0 to 1. This represents the resource satisfaction level, which is the ratio of the remaining resources in the candidate environment to the resources required for deployment. It is set to 1 when it is greater than or equal to 1, and the value is taken as the actual ratio when it is less than 1. Indicates the degree of matching at each stage; Indicates the health status of the load. Indicates the degree of security level matching; , , , Indicates the weighting coefficient. ; S33. Based on the defined target environment, dynamically calculate dynamic resource quotas and orchestrate infrastructure according to project scale and historical resource utilization rates. Infrastructure includes computing resources, storage resources, network resources, and operating environment. The formula for calculating dynamic resource quotas is as follows: ; In the formula, This indicates the resource quota allocated for this deployment; This represents the baseline resource quota, which corresponds to the standard resource allocation for a standard-sized project. Indicates the project size coefficient; This represents the historical utilization rate correction factor, calculated using the following formula: ,in, To achieve the target resource utilization rate, This represents the historical average resource utilization rate for similar projects at the same stage. S34. A batch rolling release strategy is adopted for deployment. For each batch, a health check is performed. Based on the node's liveness status, interface response rate, and error rate data, the deployment health of a single batch is calculated. When the health is higher than the set threshold, the deployment of this batch is determined to be successful and the next batch continues; otherwise, the deployment is automatically paused and the abnormal rollback logic is triggered. S35. After the full batch deployment is completed, a full health check will be performed.
7. The collaborative method for automated deployment and operation and maintenance services of project software according to claim 1, characterized in that: Step S4 specifically includes: S41. Conduct multi-dimensional testing on the project based on the deployment results, including functional automation testing, performance stress testing, interface integration testing, and business scenario compatibility testing, and obtain multi-dimensional scores based on the test results; S42. Calculate the final deployment risk value based on the multi-dimensional scoring results; S43. Based on the final deployment risk value, the deployment is divided into low risk, medium risk and high risk. The preset release strategy is matched according to the risk results and the deployment is carried out according to the preset release strategy.
8. The collaborative method for automated deployment and operation and maintenance services of project software according to claim 7, characterized in that, The final deployment risk value calculation formula is as follows: ; in, The basic risk value is represented by the following formula: ; In the formula, The final deployment overall risk value ranges from 0 to 1; P represents the project phase weighting coefficient; S represents the change scope percentage, ranging from 0 to 1, indicating the proportion of modules involved in this change to the total number of modules in the project; F represents the historical failure rate of similar deployments, ranging from 0 to 1. , , Represents the weighting coefficients, satisfying ; This represents the correction coefficient for the verification result.
9. The collaborative method and system for automated deployment and operation and maintenance services of project software according to claim 8, characterized in that: Step S5 specifically includes: S51. Obtain multi-dimensional information on project stage, business level, and deployment architecture based on a unified metadata model; S52. Set inspection strategies, inspection frequencies, alarm thresholds and emergency plans based on multi-dimensional information of the project. S53. When a monitoring alarm or a user service ticket is received, the version information, deployment records, configuration parameters, and change history of the corresponding project are obtained from the unified metadata model. The matching degree between the current alarm / ticket and historical fault cases is calculated through a similarity matching algorithm, and the investigation clues and possible root cause directions are output. S54. Based on the matching results, associate the alarm or service ticket with the fault location.
10. A collaborative system for automated deployment and operation and maintenance services of project software, employing the collaborative method for automated deployment and operation and maintenance services of project software as described in any one of claims 1-9, characterized in that, include: The central engine connects to each module and serves as the system's data hub and scheduling core. It is used to uniformly store end-to-end metadata, monitor project nodes, issue execution instructions, and transfer data. The project initialization module is used to obtain the basic attribute information of the project input by the user, conduct a feasibility assessment of the project based on the basic attribute information, and after the assessment is passed, build a unified metadata model and initialize it, and bind the automatic triggering rules corresponding to each project node. The project generation module is used to dynamically match and generate CI pipelines based on a unified metadata model and perform full-process verification. The deployment module is used to automatically trigger deployment instructions based on the verification results. It adaptively decides the target environment based on the project stage, resource requirements and real-time environmental load, dynamically calculates resource quotas and orchestrates infrastructure, performs deployment and verifies deployment health in real time. The deployment module is used to perform multi-dimensional testing on the project based on the deployment results, calculate the final deployment risk value based on the test results, adaptively match the deployment strategy according to the risk value, and execute the deployment. The operation and maintenance service module is used to dynamically configure inspection strategies and alarm rules according to the project stage, realize the association and location of alarms or service work orders with fault locations, and repair faults.
