A GHOST TASK DECOMPOSITION AND SAFE WORKING WINDOW DEFINITION SYSTEM

Abstract

This invention relates to a system (1) that enables the pre-calculation of the risk of conflicts between periodic or triggered batch jobs running on different task scheduler resources in large-scale enterprise information technology infrastructures, the differentiation of actual running tasks from ghost tasks / stale lock situations, and the determination of the safe working window.

Description

1 TARIFF A GHOST TASK: DECONTAMINATION AND SAFE WORKING WINDOW DETERMINATION SYSTEM Technical Area This invention can be used in large-scale enterprise information technology infrastructures for different tasks. Common to periodic or triggered batch jobs running on timer resources Pre-calculation of conflict risk based on access to technical resources, Differentiating between the actual working task and the ghost task / stale lock states, and It relates to a system that enables the determination of the safe operating window. 10 Previous Technique In current technology, single-task scheduling solutions mostly rely on their own 15 the runtime of tasks on the platform, the final run result, and the work It shows its history. Monitoring and alarm solutions, mission failure or The system can generate an alarm when an error occurs, but it should identify common sources before the error happens. The conflict cannot be calculated. Deployment automations, deployment steps It runs; however, another scheduled task running simultaneously uses the same database 20 It often checks whether the table conflicts with the file path or service component. It does not calculate. Furthermore, with current technology, issues such as lock cleaning or "already running" can occur. The solution usually involves manual process control, log control, or database session 25 It is done under control. 2 In the known state of the art, jobs that run on different task scheduler systems Simultaneous access to the same resource because it is not represented in a common resource graph. The risk cannot be foreseen; the query, file path, and service mentioned in the task content. Data in the form of a tip or messaging channel is often provided manually. It depends on the documentation; if the documentation is not up-to-date, there is a 5% chance of a conflict. becoming invisible, a task appearing to be running, actually running This may not mean that the task is disabled due to a stale lock or old status record. unnecessarily preventing access to the same resource during live acquisition If there is a batch task, errors, deadlocks, or data inconsistencies may occur; this Technical problems often arise where the risk is only understood after the distribution. 10 It is receiving. In the Chinese patent document numbered CN122018897, which is included in the known state of the art... A visual artificial intelligence workflow editing method and system is described. 15 The method in the invention involves deriving a demand for a manufacturing industry application scenario. and combined with a hint word template library to create a task dependency flow. creating the graph and establishing a low-code editing standard, establishing a low-code standard for the AI ​​node type identification, determination of the connection strength between nodes and a dynamic 20 Creating a workflow topological graph, configuring multiple model interfaces. obtaining parameters, performing capacity matching, difference compensation factors through the creation of a planning opportunity atlas and multi-model collaborative Planning is carried out based on the location of the execution bottleneck. Determining a dynamic analysis trigger condition, a fault node 25 identification, road reconstruction and self-adaptation Creating an execution sequence, through manufacturing industry application validation Performing node execution efficiency analysis, graphical node library. creation and generation of drag-and-drop editing interface data It includes the steps. The topology generation module in the invention is a low-code 30 3 Defining AI node types using orchestration standards, artificial The strength of the connection between data flow and control flow according to intelligence node types. to identify, using connection strength, a dynamic workflow topology graph. to create and orchestrate a node based on a dynamic workflow topology graph. It is used to determine the priority table. 5 However, in the document numbered CN122018897, in the near future window... Comparing tasks and distribution events on a graph, from the same source. The conflict score for accessing tasks is 10 for tasks that appear to be “already running”. ghost mission score, distribution-batch interaction score for live deployment events and from calculating the safe window score for alternative times It is not mentioned. Brief Description of the Invention 15 The aim of this invention is to use multiple timer sources instead of a single timer tracking. combining them in a shared work-resource graph, derived from task contents Deletion of dependency, classification of access type, recent Calculating the source intersection in the window, collision and ghost task 20 generating scores, postponing or safeguarding tasks / assignments based on score thresholds I would like the window to be installed, not just for monitoring or generating alarms, but also for troubleshooting. platform behavior by calculating technical resource conflicts before they occur. a ghost task decomposition and secure that enables the generation of transformative decisions. The system implements the working window determination system. 25 Detailed Description of the Invention 4 The "Ghost Task Decomposition" was carried out to achieve the purpose of this invention. And the "Safe Working Window Determination System" is shown in the figure; Figure 1: The subject of the invention is a ghost task decomposition and safe operating window. This is a schematic view of the determination system. 5 The parts shown in the figure are individually numbered, and the corresponding numbers correspond to these numbers. It is given below. 1. System 10 2. Application 3. Analytics server In large-scale enterprise information technology infrastructures, different task schedulers Common technical aspects of periodic or trigger-based batch jobs performed in resources 15 Pre-calculation of conflict risk based on access to resources, actual Separating working tasks from ghost tasks / stale lock states and ensuring safety. a ghost task, the subject of the invention, that enables the determination of the working window Decoupling and safe working window determination system (1); - connection with at least an information technology communication infrastructure at the corporate level 20 capable of installing, run on an authorized user's electronic device, authorized user's information, tasks, and flows transmitted through the information technology communication network in question the most configured to enable viewing and / or management at least one application (2), - connection with the information technology communication infrastructure with which the application (2) connects 25 The Quran, job descriptions within the communication infrastructure in question, and timing. statements, script and query contents, task work history, central log records, relational database session / lock metadata records, file system lock metadata data in the form of logs, deployment events, service inventory and maintenance window accessing it through the data sources layer on top of it, with at least one data 30 summation and normalization module, resource dependency inference module, work resource graph module, scoring engine, policy decision module, action and to operate the orchestration module, feedback and control module, on it The application connects with the data collection and normalization module (2). Data from different sources in the communication infrastructure are combined into a common model. 5 the conversion of the time information in the incoming data into a single time to ensure translation into the language and standardization of task and resource names, matching the environmental information in the subsequent data, from different sources of the same source. combining written documents and sensitive information such as customer or user data. Enabling masking and filtering in areas where data is located, 10 the application communicates with the resource dependency inference module (2) on it. from task scripts, query commands, and logs collected from the communication infrastructure Table, file path, service endpoint from signatures and configuration contents, messaging channel, shared storage, lock file and distribution component, etc. To enable the extraction of resources, then set the access type to “READ”, “WRITE”, 15 “DELETE”, “LOCK”, “SCHEMA_CHANGE”, “RESTART”, “DEPLOY”, To ensure it is classified as “PUBLISH”, “CONSUME” or “MOUNT”, the work resource graph module and the resource dependency inference module on it The work node of each task in the access types that have outputs, each technical resource It should be kept as a source node, with access at the edges between the task and the source. type, confidence score, last seen time, environment and resource criticality to ensure the level is stored, with the scoring engine on the application (2) in the near time window within the communication infrastructure with which it communicates To enable comparison of tasks and distribution events on a graph, the same The conflict score for tasks accessing the resource is 25, which appears as “already running”. Ghost mission score of tasks, distribution-batch for live deployment events interaction score and safe window score for alternate times to enable calculation, with its policy and decision layer and scoring engine. Comparing the calculated scores with predetermined threshold values, at low risk. Permission, medium risk warning, high risk manual approval, critical risk postponement or safe 30 6 to enable the production of decisions in the working window format, application (2) automatic on critical resources within the communication infrastructure with which it communicates limiting the action; irreversible actions are only recommendations. to ensure that the application (2) is shown to the authorized user, 5 generated in the policy and decision module with the action and orchestration module on it Depending on the outcome of the decisions, the task may be postponed, the task may be suspended, or a distribution step may be taken. Stopping, creating approval records, and authorizing the user. Suggest a safe window or lock clearing suggestion via application (2) old (2) interface without application on communication network platforms, approved and low-risk actions are in question. to ensure that it can be implemented with (2) user interface automation in the application, Action and orchestration with the feedback and control module located on it. the action that the module applies to the communication infrastructure in question, the server measuring the success rate of the task, working time, error rate, key Data in the form of formation, distribution result and manual confirmation information 15 This ensures that the information is recorded, allowing for a reduction in past error coefficients in subsequent scoring. and its use as a coefficient of success for action, verifiable evidence for each decision. at least one analytics server configured to enable the creation of the record (3) includes. The application (1) in the system that is the subject of the invention (2) authorized user's mobile phone, to work on an electronic device in the form of a tablet or computer It is being structured. The invention concerns the system (1) analysis server (3) application (2) connection 25 Included in the communication infrastructure it established: “job” (task ID, name, schedule, command, owner team, environment), “Run” (task example, start / end time, Result, log signature), “Resource” (resource type, resource name, environment, criticality), “Edge” (task-resource relationship, access type, trust score), “Lock” (lock path / session, last update, owner task), “DeploymentEvent” (deployment time, affected 30 7 component, undoability), “Action” (permission / warning / approval / delay), “Evidence” (decision) It is structured to access fundamental objects in the form of (evidence). Analysis server (3) connects to the communication infrastructure with which the application (2) connects. providing and running at least one resource extraction algorithm on it, these 5 this allows you to view job descriptions, scripts, queries, configurations, logs, and work history. retrieved from the connected communication infrastructure, the retrieved content line by line. parsing, then adding comment lines, temporary values ​​and masking the necessary values ​​should be removed from the relevant content, from the query commands table / schema, file path from file commands, service 10 from network calls the endpoint, the channel name from messaging commands, the component from deployment commands the removal of the name, the classification of each resource relationship according to its access type, The dependency confidence score is determined by verifying the same source from different sources. increasing the number of relationships achieved and at least recording them in a business-resource graph. It is structured to provide this. 15 Analysis server (3) connects to the communication infrastructure with which the application (2) connects. to provide and run at least one resource extraction algorithm on it, the word The algorithm uses the following formula: “Confidence Score = (Number of Source Verifications x 0.35) + (Live (Visibility in Log x 0.25) + (Visibility in Task History x 0.20) + (Inventory 20) To enable the execution of the formula "Match x 0.20)" It is being structured. Analysis server (3) connects to the communication infrastructure with which the application (2) connects. to set up, thanks to the work resource graph module on it, the graph is in the form "G=(V,E)" 25 modeling, from the task nodes and source nodes of the V cluster its formation, the set E showing the technical impact of the task on the resource. providing edge features such as access type, last seen time, and source security. It should include the score, mission criticality level, and environmental information, all from the same source. WRITE-WRITE, WRITE-READ, DELETE-READ, LOCK-WRITE or 30 8 Risky relationship combinations such as DEPLOY-WRITE are potential overlap candidates. It is configured to allow for marking. Concurrency and time window calculation on the analysis server (3) 5 obtained from the communication infrastructure with which the application (2) communicates during its implementation The planned start date of work, expected duration, and past duration for each task assigned. Creating a working window using the distribution, running two tasks Calculating the probability of concurrency if the windows intersect, task duration. If unknown, the median duration and p95 duration of the last N successful studies. To ensure its use, "Intersecting 10" is also used in calculating the probability of simultaneity. The formulation "Duration / Shortest Task Window" should be used. It is structured to provide this. Obtaining the collision score in the scoring engine on the analysis server (3) to ensure, the criticality level of the common resource in calculating the collision score, 15 access type coefficient, probability of concurrency, historical error coefficient, task duration, return data such as the inability to retrieve and the proximity to live organism are taken into consideration To ensure this, the collision score should be "Collision Score = Common Resource Criticality". Score x Access Type Coefficient x Probability of Concurrency x Historical Error Coefficient x The formula is "Irreversibility Coefficient x Affinity Coefficient". It is configured to enable the formula to be executed. By running the scoring engine on the analysis server (3), the said partner If the source has an overlap score between 0 and 30, it indicates low risk and the word... the subject is to ensure that it is allowed to fulfill its role in the communications infrastructure, joint resource 25 If the overlap score is between 31-60, it indicates a moderate risk and at least one To trigger an alert, if the common resource has a conflict score between 61-80. identification of high risk and implementation of the common resource in question (2) Ensuring that manual work approval is requested, common source between 81-100 If there is a conflict score, the critical risk should be identified and the relevant 30 common resource in question should be considered. 9 obstruction of my duties and distribution in the communication infrastructure, and then the authorized person to ensure that the safe window is suggested to the user via the application (2) It is structured accordingly. Analysis server (3) connection with the communication infrastructure with which the application (2) communicates 5 If the task appears to be "running" after installation, then that task is directly involved in the actual work. rejection, transaction existence, lock file last update time, log last write time, database session, heartbeat, and resource consumption checking them together, and generating a ghost task score with those checks. It is structured to provide. 10 Analysis server (3) connection with the communication infrastructure with which the application (2) communicates. in establishing and determining the ghost score related to the task in the relevant communication infrastructure “Lock Age Coefficient + Process Absence Coefficient + Log Silence Coefficient + The formula is "Session Absence Coefficient + Heartbeat Absence Coefficient", which is 15. It is configured to enable operation. Analysis server (3) connection with the communication infrastructure with which the application (2) communicates. After establishing and determining the ghost score related to the task in the relevant communication infrastructure, If the ghost score is low, it means the task is actually working on that communication infrastructure. 20 (2) It is suggested that manual control be recommended when the ghost score is at a high level. The relevant task is classified as a "stalock" or ghost mission by the authorized person. unlock or restart task via application (2) for the user to ensure the action is recommended, automatic lock cleaning is only low effective 25 and is structured to ensure its implementation in recoverable resources. 10 The analysis server (3) is the communication infrastructure with which the application (2) communicates. at least before a mission's deployment or configuration change begins, the source graph of the affected component located in the communication infrastructure in question having connected resources and accessing those resources within the same time window. If there is a task, calculate the "distribution-batch" interaction score; if the score is high, give 5. the distribution is delayed, then the authorized user via the application (2) manual approval is requested or the secure time interval is given to the authorized user. It is structured to ensure that it is recommended. The analysis server (3) is at least 10 within the communication infrastructure with which the application (2) communicates. at least one task, component, involves scanning candidate timeframes, and analyzing each candidate. number of active tasks, common resource conflicts, historical error density, maintenance window suitability, resource intensity and deployment schedule to ensure that the lowest risk range is evaluated through application (2) To ensure that it is suggested as a secure working window to the authorized user, 15 The Safe Window Score is calculated as “100 - (Collision Risk + Source)”. Density + Maintenance Inadequacy + Past Error Density + Distribution To enable the execution of the formula "proximity)" It is being structured. To run the decision engine on the analysis server (3), the application (2) at least one event within the communication infrastructure with which it communicates “D=[Task Criticality, Resource Criticality Score, Access Type, Probability of Concurrence, History Error Rate, Lock Status, Session Status, Log Silence, Deployment The decision vector in the form of "Proximity, Reversibility, Approval Requirement" is 25 to create, calculate the action priority score of the event in question (Action Priority Score = Expected Technical Benefit / Risk Score) to provide policy to ensure that the highest priority action is selected in accordance with the rules It is being structured. 11 The analysis server (3) postpone the task, stop the task, restart the task, Change runtime, postpone deployment step, suggest maintenance window (2) the transmission of decisions in the form of; waiting for live acquisition, approval Flow creation, planning according to the safe window, rollback readiness control. Actions of this nature are sent to the distribution automation channel in the relevant communication infrastructure. It is configured to enable sending. The analysis server (3) communicates with the application (2) to perform post-action tasks. Result, runtime, error log, resource lock, task delay, deployment result and the parameters in the form of manual confirmation output are in the specified observation window 10 to ensure monitoring, and if the action is successful, the weight of the action under the relevant conditions increasing the historical error coefficient and risk coefficient if the relevant action fails. upgrading it so that it is more accurate in the same combinations of resources and tasks. It is structured to facilitate the production of decisions. In one application of the invention, the analysis server (3) uses a relational table instead of a graph data structure. or to enable the use of document-based models, task-resource relationships, access types, time windows, and scores are involved enabling storage in relational table or document models, artificial intelligence the ability to use a rule-based parser instead of a supported parser, user 20 to enable the use of application interface integration instead of interface (2) It is structured accordingly. In one application of the invention, the analysis server (3) duty time estimation model, Capacity density estimation, coefficient update learning from past events 25 model, automatic notification integration, decision simulation, advanced visual resources Two-way planning integration in the form of a graph screen and change schedule. It is configured to run within its components. 12 In one application of the invention, the analysis server (3) uses an AI-powered discriminator, graphing database, user interface automation, notification integration, automatic lock optional components such as cleaning and an advanced capacity estimation model. It is configured to run. Industrial application of the invention In the sample scenario (1) of the system in question, at 02:00, the communication infrastructure A live deployment is planned for an application component. A 10-inch phase is also planned within the same timeframe. Task-A, which runs on the application server scheduler, specifies a particular relational data. WRITE access is performed on the base table in the operating system scheduler. Task-B, which is working, reports the same table with READ access. System (1), The expected time for Task A is 22 minutes, for Task B 18 minutes, and for live deployment... It calculates as 12 minutes. Time windows overlap by 70%. Common 15 Source criticality score 8 / 10, access type coefficient 1.4 for WRITE-READ, history The error coefficient is determined as 1.2 and the irrecoverability coefficient as 1.1. Collision score. 82 is calculated. The system (1) finds the time 02.00 to be critically risky, authorizing the live animal to be taken in. authorized through the application (2) running on the user's electronic device It suggests a meeting to the user between 02:40 and 03:05 and creates a manual confirmation record. Action 20 If the subsequent tasks are successfully completed, the relevant decision analysis server (3) is successful. This is marked as such; under the same conditions, the confidence score of subsequent recommendations is increased. The system in question (1) and at least one communication infrastructure in different 25 The problem of conflicting access to the same technical resource by tasks in timers Calculating the conflict score using the task-resource graph and access type classification. Solving it using, although the task appears to be working, the actual work The absence of a technical problem with process, lock, log, session and heartbeat signals Solving this through ghost task score calculation, 30 during live acquisition. 13 The invisible batch effect problem is related to the distribution event and the task-resource graph. This is resolved by intersecting the problem and suggesting a safe working window. The system in question (1) is within the scope of the Personal Data Protection Law (KVKK). It operates. In this system (1) customer data or transaction content 5 No data is collected; only technical operational metadata information is used. The invention concerns the decomposition of a ghost task and the determination of a safe operating window. It is possible to develop a wide variety of applications of the system (1), and the invention is here It cannot be limited to the examples given; it is essentially as stated in the claims. 10

Claims

14 REQUESTS 1. In large-scale corporate information technology infrastructures, different tasks Common to periodic or triggered batch jobs running on timer resources 5. Pre-calculation of conflict risk based on access to technical resources. Differentiating between the actual working task and the ghost task / stale lock states, and enables the determination of the safe operating window; - connection with at least an information technology communication infrastructure at the corporate level capable of installing, run on an authorized user's electronic device, authorized user's information, tasks, and flows transmitted through the information technology communication network in question 10 the most configured to enable viewing and / or management at least one application (2) and - connection with the information technology communication infrastructure with which the application (2) connects The Quran, job descriptions within the communication infrastructure in question, and timing. statements, script and query contents, task work history, central log records, 15 relational database session / lock metadata records, file system lock metadata data in the form of logs, deployment events, service inventory and maintenance window accessing it through the data sources layer on top of it, at least one data summation and normalization module, resource dependency inference module, work resource graph module, scoring engine, policy decision module, action and 20 to operate the orchestration module, feedback and control module, on it The application connects with the data collection and normalization module (2). data from different sources in the communication infrastructure are combined into a common model. the conversion of the time information in the incoming data into a single time to ensure translation into the language, standardization of task and resource names, 25 matching the environmental information in the subsequent data, from different sources of the same source. combining written documents and sensitive information such as customer or user data. To enable masking and filtering in areas where data is located, the application communicates with the resource dependency inference module (2) on it. from task scripts and query commands collected from the communication infrastructure, log 30 Table, file path, service endpoint from signatures and configuration contents, messaging channel, shared storage, lock file and distribution component, etc. to enable the extraction of resources, then set the access type to “READ”, “WRITE”, “DELETE”, “LOCK”, “SCHEMA_CHANGE”, “RESTART”, “DEPLOY”, Ensuring it is classified as “PUBLISH”, “CONSUME” or “MOUNT”, 5 the work resource graph module and the resource dependency inference module on it The work node of each task in the access types that have outputs, each technical resource maintaining it as a resource node, providing access at the edges between the task and the resource. type, confidence score, last seen time, environment and resource criticality to ensure the level is stored, the application (2) 10 with the scoring engine on it in the near time window within the communication infrastructure with which it communicates To enable comparison of tasks and distribution events on a graph, the same The conflict score for tasks accessing the resource appears as "already running". Ghost mission score of tasks, distribution-batch for live deployment events Interaction score and safe window score for alternate times is 15 to enable calculation, with its policy and decision layer and scoring engine. Comparing the calculated scores with predetermined threshold values, at low risk. permission, medium risk warning, high risk manual approval, critical risk postponement or safe to enable the production of decisions in the working window format, application (2) Automatic 20 on critical resources located in the communication infrastructure with which it communicates limiting the action; irreversible actions are only recommendations. to ensure that the application (2) is shown to the authorized user, produced in the policy and decision module with the action and orchestration module on it Depending on the outcome of the decisions, the task may be postponed, the task may be suspended, or a distribution step may be taken. Stopping, creating approval records, and authorizing the user 25 Suggest a safe window or lock clearing suggestion via application (2) old (2) interface without application on communication network platforms, approved and low-risk actions are involved. to ensure that it can be implemented with (2) user interface automation in the application, Action and orchestration 30 with feedback and control module located on it 16 the action that the module applies to the communication infrastructure in question, the server measuring the success rate of the task, working time, error rate, key data in the form of formation, distribution result and manual confirmation information This ensures that the information is recorded, allowing for a reduction in past error coefficients in subsequent scoring. and its use as an action success coefficient, verifiable evidence for each decision. 5 at least one analytics server configured to enable the creation of the record a ghost task parsing and safe working window characterized by (3). determination system (1).

2. The communication infrastructure with which the application (2) connects: “job” (task 10 ID, name, timing, command, owner team, environment), “Run” (task work instance, start / end time, result, log signature), “Resource” (resource type, resource name, Environment, criticality), “Edge” (task-resource relationship, access type, trust score), “Lock” (lock path / session, last update, owner task), “DeploymentEvent” (deployment) (time, affected component, undoability), “Action” (permission / warning / approval / delay), 15 To access the fundamental objects in the form of "evidence" (evidence of the decision). like the one in Claim 1 characterized by the configured analysis server (3). Ghost task decomposition and safe working window determination system (1).

3. Connecting with the communication infrastructure with which the application (2) connects and 20 to run at least one resource extraction algorithm on it, thereby completing the task definition, script, query, configuration, log and operation history are linked. obtaining the content from the communication infrastructure, parsing the obtained content line by line, then the comment lines, temporary values, and values ​​that need to be masked the removal of the aforementioned content, the table / schema from the query commands, 25 file path from file commands, service endpoint from network calls, messaging extracting the channel name from the commands and the component name from the deployment commands, classifying each resource relationship according to its access type, different access types of the same resource Increasing the dependency confidence score by verifying it from reliable sources, the obtained 30 to ensure that relationships are at least plotted on a business-resource graph 17 as in Claim 1 or 2 characterized by the configured analysis server (3). a ghost task parsing and safe working window determination system (1).

4. Connect with the communication infrastructure with which the application (2) connects running at least one resource extraction algorithm on it, the 5 in question In the algorithm, "Confidence Score = (Number of Source Verifications x 0.35) + (Live Log (Seen x 0.25) + (Seen in Task History x 0.20) + (Inventory Match x The analysis is structured to enable the execution of the formula "0.20)". any of the above requests characterized by its server (3) Ghost task parsing and safe working window determination system (1). 10 5. To connect with the communication infrastructure with which the application (2) connects, Thanks to the work resource graph module on it, the graph can be in the form of "G=(V,E)". modeling, from the task nodes and source nodes of the V cluster its formation, the set E showing the technical impact of the task on the resource 15 providing edge features such as access type, last seen time, and source security. It should include the score, mission criticality level, and environmental information, all from the same source. WRITE-WRITE, WRITE-READ, DELETE-READ, LOCK-WRITE or Risky relationship combinations such as DEPLOY-WRITE are potential overlap candidates. 20 characterized by the analysis server (3) configured to enable marking. Parsing and securing a ghost task like any of the above requests made working window determination system (1).

6. During the calculation of simultaneity and time window on it For each task obtained from the communication infrastructure with which the application (2) communicates, 25 planned work start date, expected duration and past time distribution creating a working window using, working windows for two tasks If they intersect, calculate the probability of concurrency; if the term of duty is unknown, then the end. To ensure the use of the median duration and p95 duration of N successful studies, The probability of concurrency is also calculated using the formula: "Intersecting Time / Shortest Task 30". 18 configured to enable the use of the formulation "window" any of the above requests characterized by the analysis server (3) Ghost task decomposition and safe working window determination system (1).

7. Ensure the collision score is obtained in the scoring engine, collision 5 In calculating the score, the criticality level of the common resource and the access type coefficient are considered. probability of concurrency, historical error rate, task duration, irreversibility coefficient and ensuring that data such as proximity to live organism acquisition is taken into account, avoiding overlap. The score is calculated as "Conflict Score = Common Resource Criticality Score x Access Type Coefficient". x Simultaneity Probability x Historical Error Coefficient x Irrecoverability Coefficient x 10 Enabling the execution of the formula "Acquisition Proximity Coefficient". The above is characterized by the analysis server (3) which is configured for the above. Parsing any of the ghost tasks and the safe working window determination system (1).

8. By running the scoring engine on it, you can get the score of the shared resource between 0-30. If there is a conflict score, the low risk should be identified and the communication in question should be considered. to ensure that his role in the infrastructure is permitted, joint source between 31-60 If there is a conflict score, it should identify the medium risk and at least generate a warning. To ensure that the common source has an overlap score between 61-80, the high risk is 20. determination and manual implementation of the said common resource through application (2) To ensure that work approval is requested, the joint source overlaps between 81-100. If a score exists, the identification of the critical risk and the relevant communication of the common resource in question. blocking my task and distribution in the infrastructure, then to the authorized user. 25 to ensure that the safe window is suggested via application (2). from the above requests characterized by the configured analysis server (3) Parsing any kind of ghost task and setting a safe working window. system (1). 19 9. Connect with the communication infrastructure with which the application (2) communicates and perform the task. If it appears to be "working," then the task in question is considered direct and actual work. not being done, process existence, lock file last update time, log last write time, database session, heartbeat, and resource consumption 5 characterized by the analysis server (3) configured to provide Parsing and securing a ghost task like any of the above requests working window determination system (1).

10. Connect with the communication infrastructure with which the application (2) communicates and the relevant 10 "Key Age" in determining the ghost score related to the task in communication infrastructure. Coefficient + Process Absence Coefficient + Log Silence Coefficient + Session The formula is "Absence Coefficient + Heartbeat Absence Coefficient" characterized by the analysis server (3) configured to enable its operation. Parsing and securing a ghost task like any of the above requests is crucial. working window determination system (1).

11. Connect with the communication infrastructure with which the application (2) communicates and the relevant After determining the ghost score related to the task in the communication infrastructure, the ghost score If the score is low, the task is considered to be actually working on the communication infrastructure in question. 20 (2) It is suggested that manual control be recommended when the ghost score is at a high level. The relevant task is classified as a "stalock" or ghost mission by the authorized person. unlock or restart task via application (2) for the user to ensure the action is recommended, automatic lock cleaning is only low effective 25 and analysis structured to ensure its application in recoverable resources. any of the above requests characterized by its server (3) Ghost task decomposition and safe working window determination system (1). 20 12. At least one communication infrastructure with which the application (2) communicates. Before the mission goes live or any configuration change begins, the component in the communication infrastructure that will be affected and its connection to the resource graph. finding the resources, and simultaneously, the task of accessing those resources within the same time window. If available, calculate the "distribution-batch" interaction score; if the score is high, distribute 5 units. to be held in abeyance, then manual approval from the authorized user via application (2). requesting or suggesting a safe time interval to the authorized user characterized by the analysis server (3) configured to provide Parsing and securing a ghost task like any of the above requests Working window determination system (1). 10 13. At least one communication infrastructure with which the application (2) communicates. The task is to scan the candidate timeframes for the component and identify active tasks related to each candidate. number, common resource conflict, past error density, maintenance window Evaluation of suitability, resource intensity and deployment schedule 15 to provide the lowest risk range to the authorized user via the application (2) To ensure it is recommended as a safe working window, Safe Window The score is calculated as “100 - (Collision Risk + Resource Intensity + Maintenance)”. (Incompatibility + Past Error Density + Distribution Proximity)” 20 with analysis server (3) configured to enable the formula to run a ghost mission like any of the above characterized requests Decoupling and safe working window determination system (1).

14. To start the decision engine on it, thereby enabling the application's (2) communication Within the communication infrastructure it established, at least one event was “D=[Mission Criticality, 25 Resource Criticality Score, Access Type, Concurrency Probability, Error History Coefficient, Lock Status, Session Status, Log Silence, Live Access Proximity, Creating a decision vector in the form of "Revocability, Approval Requirement", The subject is calculating the action priority score of the event (Action Priority Score). = Expected Technical Benefit / Risk Score) to be delivered, in compliance with policy rules, at least 30 21 Analysis structured to ensure the selection of a high-priority action. any of the above requests characterized by its server (3) Ghost task decomposition and safe working window determination system (1).

15. Task postponement, task suspension, task restart, working time 5 change, postpone deployment step, suggest maintenance window (2) transmitting decisions to implementation; waiting for live launch, approval flow creation, planning according to the safe window, rollback readiness control actions to the distribution automation channel in the relevant communication infrastructure 10 characterized by the analysis server (3) configured to enable sending. Parsing and securing a ghost task like any of the above requests made working window determination system (1).

16. By communicating with the application (2), the post-action task result, working time, Error log, resource lock, task delay, deployment result, and manual confirmation output 15 To enable the monitoring of parameters such as these within the defined observation window, If the action is successful, increasing the weight of the action under the relevant conditions, If the action fails, the historical error coefficient and the risk coefficient should be increased. This allows for more accurate decision-making in the same combinations of resources and tasks. 20 characterized by the analysis server (3) configured to provide Parsing and securing a ghost task like any of the above requests working window determination system (1).

17. Relational table or document-based models instead of graph data structures to enable its use, task-resource relationships, access types, time 25 windows and scores in the relational table or document models in question to ensure storage, rule-based instead of AI-powered parsing. the parser can be used by the application interface (2) instead of the user interface analytics server configured to enable integration 22 a ghost mission like any of the above requests characterized by (3) Decoupling and safe working window determination system (1).

18. Mission duration estimation model, capacity density estimation, from past events Learning coefficient update model, automatic notification integration, decision 5 The simulation is bidirectional with an advanced visual resource graph display and change schedule. to operate within its components in the form of planning integration from the above requests characterized by the configured analysis server (3) Parsing any kind of ghost task and setting a safe working window. system (1). 10 19. AI-powered parser, graph database, user interface automation, Notification integration, automatic lock clearing, advanced capacity estimation model. analysis server configured to run optional components such as these A ghost mission like any of the above requests characterized by (3) 15 Decoupling and safe working window determination system (1). 25