A data center scheduling method and system adapted to multiple business scenarios

Abstract

The present invention is applicable to the field of business data scheduling, and provides a data middle-end scheduling method and system that is adapted to multiple business scenarios. The method collects business feature data in real time, abstracts the business departments involved in scheduling into game roles, and allocates resource demand attributes and game chip attributes to them. Based on the task urgency coefficient, resource utilization efficiency and scenario adaptation weight, a bidding calculation model is constructed to generate automatic bids. Exclusive resources are allocated to high-priority roles in combination with the resource allocation heat map, and low-priority roles are scheduled on demand through an elastic resource pool. The system dynamically monitors resource utilization and task indicators, evaluates the rationality of bids and adjusts chip attributes. The present invention motivates roles to compete efficiently through a game mechanism, and optimizes resource allocation in combination with dynamic feedback, thereby solving the problems of poor adaptability and unfair allocation of traditional methods, and significantly improving resource utilization efficiency and business continuity in multiple business scenarios.

Description

Technical Field

[0001] The present invention belongs to the field of business data scheduling, and in particular relates to a data middle-station scheduling method and system adapted to multiple business scenarios. Background Art

[0002] As enterprises deepen their digital transformation, the resource scheduling efficiency of the data center, as the core infrastructure supporting multiple business scenarios, directly impacts business continuity and responsiveness. Current mainstream data center scheduling technologies often use static rules or fixed priority mechanisms, such as a first-in-first-out (FIFO) strategy based on task queues or allocating resources based on preset weights. While these approaches can handle some simple scenarios, they have significant shortcomings in complex and changing business environments:

[0003] Traditional approaches lack a scientific, multi-dimensional evaluation mechanism, making resource allocation susceptible to subjective judgment. Inefficient tasks can preempt critical resources, while urgent, high-value tasks go unmet, resulting in wasted resources and business delays. Furthermore, without introducing mechanisms for competition and collaboration between roles, departments lack self-discipline in resource use and are unable to optimize resource allocation strategies based on historical performance feedback, leading to long-term inefficient resource utilization. Summary of the Invention

[0004] The purpose of the present invention is to provide a data middle-station scheduling method and system that is adaptable to multiple business scenarios, aiming to solve the technical problems existing in the existing technology identified in the background technology.

[0005] This invention is implemented as follows: a data center scheduling method adapted to multiple business scenarios, which achieves efficient and flexible resource scheduling by collecting and analyzing various business feature data in real time. The method first abstracts multiple business departments into gaming roles, dynamically identifies the current business scenario type based on real-time data, and accurately allocates resources based on the resource sensitivity of different scenarios.

[0006] The method collects heterogeneous data from multiple sources in real time, including business behavior data, system resource data, and business metadata, and normalizes it to generate structured feature vectors. Next, a lightweight classification model is used to identify the current business scenario and dynamically calibrate the resource sensitivity level of each scenario. Based on this, the solution defines resource demand attributes and gaming chip attributes for each gaming role, ensuring that each role's actual needs and influence are reflected in the scheduling process.

[0007] To effectively manage resource allocation, the solution introduces virtual resource quotas as departmental "budgets," ensuring that, given limited resources, each department can rationally apply for and utilize resources based on immediate business needs. Based on the current business scenario type and task deadline, the urgency coefficient of the task is calculated. Combined with the adaptation weights extracted from the matching matrix, an automatic bid is calculated to ensure rational resource scheduling.

[0008] During the resource allocation process, the solution generates a resource allocation heat map, prioritizes exclusive resource allocation to roles with high resource utilization efficiency and high scenario matching, and triggers elastic resource pool scheduling to meet the needs of low-priority roles. At the same time, the solution dynamically adjusts the budget and credit points of roles by evaluating the matching degree between automatic bids and actual performance, ensuring that roles with efficient bids can obtain more resource support, while roles with inefficient bids are limited by resource application quotas. In addition, when the resource demands of multiple roles exceed the system capacity, the solution can automatically downgrade the priority of inefficient tasks based on scenario adaptation weights and resource utilization efficiency, and generate alternative scheduling path recommendations for tasks that fail due to insufficient resources.

[0009] The beneficial effects of the present invention are:

[0010] This data middle-office scheduling method, which is adaptable to multiple business scenarios, brings significant advantages and beneficial effects by abstracting multiple business departments into gaming roles and assigning resource demand attributes and gaming chip attributes to each role. First, the abstract definition of gaming roles enables the system to utilize the mechanisms of game theory to promote the rational allocation and utilization of resources. Within this framework, departments form a dynamic relationship of competition and cooperation in the competition for resources, prompting them to pay more attention to efficiency and value contribution when applying for resources. Compared with traditional resource scheduling methods, this strategy not only enhances the self-management awareness of each department, but also encourages collaboration and cooperation between departments, thereby improving overall resource utilization efficiency and reducing unnecessary resource waste.

[0011] Secondly, each player is assigned a virtual resource quota as a ceiling for resource usage, clarifying the boundaries of resource utilization for each department. This design effectively reduces uncertainty and conflict in the resource application and utilization process, ensuring that each department can make rational use of available resources even when resources are scarce. Compared to existing technologies, many traditional methods often lack clear resource usage standards, which can easily lead to unfair or inefficient resource allocation. The establishment of virtual resource quotas, on the other hand, provides a clear benchmark, making the resource allocation process more transparent and fair, and helping to enhance trust in resource allocation among departments.

[0012] By collecting and analyzing business feature data in real time, the system dynamically identifies current business scenarios and calibrates resource sensitivity, enabling it to adapt promptly to diverse business environments. This dynamic adaptability is unmatched by traditional static scheduling methods. By incorporating scheduling decisions based on real-time data, the system can rapidly adjust resource scheduling strategies in the face of emergencies or changes in business needs, ensuring the timely completion of critical tasks and business continuity. This flexibility not only improves business response speed but also optimizes resource allocation, further enhancing overall business efficiency.

[0013] This system also enhances the system's feedback mechanism by combining performance evaluations of each player with dynamic adjustments to budgets and reputation points. This feedback mechanism not only makes resource scheduling decisions more scientific, but also fosters healthy competition among departments, ultimately driving improvements in overall corporate performance. Compared to traditional methods, this mechanism can quickly identify and reward high-performance and penalize low-performance, effectively motivating departments to proactively utilize resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 A flowchart of a data center scheduling method adapted to multiple business scenarios provided by an embodiment of the present invention;

[0015] Figure 2 A flowchart of an embodiment of the present invention that abstractly defines multiple business departments involved in data scheduling as gaming roles and assigns resource requirement attributes and gaming chip attributes to each role;

[0016] Figure 3 A flowchart for generating automatic bids based on task urgency coefficient, resource utilization efficiency, and scenario adaptation weights provided by an embodiment of the present invention;

[0017] Figure 4 A flowchart of an embodiment of the present invention for allocating exclusive resources to gaming characters ranked above the bottom line, automatically triggering elastic resource pool scheduling for the remaining gaming characters, and allocating the remaining resources on demand;

[0018] Figure 5 A flowchart of calculating the bid rationality index and input-output ratio and dynamically adjusting the chip attributes for the next scheduling cycle provided by an embodiment of the present invention;

[0019] Figure 6 A structural block diagram of a data middle-station scheduling system adapted to multiple business scenarios provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0021] A data center scheduling method adapted for multiple business scenarios achieves efficient resource scheduling across multiple business scenarios through real-time collection and intelligent analysis of business feature data. The key idea is to abstract multiple business departments into gaming roles, dynamically identify the current business scenario type based on real-time data, and implement precise resource allocation based on the scenario's resource sensitivity. This method first standardizes multi-source heterogeneous data (such as business behavior data, system resource data, and business metadata) to generate structured feature vectors. It then utilizes a lightweight classification model to identify business scenarios and dynamically calibrate the rigidity or elasticity of each scenario's resource requirements, enabling rapid adaptation to diverse business environments.

[0022] In the specific implementation of resource allocation, the method assigns a virtual resource quota to each gaming role, serving as a "budget" or upper limit for each department's resource usage. This virtual resource quota is designed to provide clear boundaries for resource usage, ensuring that departments can rationally allocate and utilize available resources even when resources are scarce. By calculating the task urgency coefficient for each gaming role and combining it with the adaptive weights extracted from the matching matrix, an automated bid is generated, ensuring rational and efficient resource allocation.

[0023] In the resource scheduling mechanism, the method generates a resource allocation heat map and prioritizes exclusive resource allocation to roles with high resource utilization efficiency and high scenario matching. At the same time, for low-priority roles, the system automatically triggers elastic resource pool scheduling to allocate remaining resources on demand. By evaluating the matching degree between automatic bids and actual performance, the method dynamically adjusts the budget and credit points of the roles. For gaming roles with efficient bids, the budget will be increased, while roles with inefficient bids face the risk of budget cuts. In addition, when resource demand exceeds system capacity, the priority of inefficient tasks will be automatically downgraded based on scenario adaptation weights and resource utilization efficiency. The system also generates alternative scheduling path recommendations for tasks that fail due to insufficient resources, thereby ensuring the continuity and stability of various businesses.

[0024] Specifically:

[0025] Figure 1 A flowchart of a data center scheduling method adapted to multiple business scenarios provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the method includes:

[0026] S100: Based on the real-time collected business feature data, the current business scenario type is identified, and multiple business departments involved in data scheduling are abstractly defined as game roles, and resource requirement attributes and game chip attributes are assigned to each role;

[0027] By treating business departments as gaming roles, this step introduces the concept of game theory, encouraging each department to form a relationship of competition and cooperation in resource allocation. This role setting not only clarifies the division of labor and responsibilities of each department in resource utilization, but also encourages mutual coordination and optimization between departments. Each role's resource requirement attributes, such as computing resource baselines and data flow dependencies, can help the system accurately identify the specific resource needs of each department, avoiding resource waste and ineffective competition. At the same time, gaming chip attributes, such as dynamic budgets, reputation points, and business weights, give each role a certain competitive advantage and flexibility, allowing each department to reasonably compete for resources based on its own historical performance and current needs when resources are scarce.

[0028] Secondly, this abstraction and classification approach helps the system more clearly establish resource allocation priorities. By dynamically calculating the adaptation weights of each role, the system can reflect in real time the relative importance and urgency of each department in the current business scenario. This mechanism not only improves the scientific nature and transparency of decision-making, but also effectively reduces unfair resource allocation caused by subjective judgment, ensuring that resources are preferentially allocated to departments that contribute more to business value.

[0029] The framework established in this step significantly improves the flexibility and responsiveness of resource scheduling. Faced with complex and volatile business environments, the system can capture changes in business characteristics in real time and rapidly adjust resource allocation strategies to ensure the smooth progress of various tasks. This dynamic scheduling capability not only improves resource utilization efficiency but also enhances business continuity and stability, enabling enterprises to maintain flexibility and adaptability in the face of competition.

[0030] like Figure 2 As shown, the current business scenario type is identified, and multiple business departments involved in data scheduling are abstractly defined as game roles, and resource demand attributes and game chip attributes are assigned to each role, specifically including:

[0031] S110 collects multi-source heterogeneous data in real time, including business behavior data, system resource data, and business metadata, and performs standardization processing to generate structured feature vectors;

[0032] S120, based on the structured feature vector, builds a lightweight classification model to identify the current business scenario type and dynamically calibrate the resource sensitivity level of the business scenario, including:

[0033] Highly sensitive scenarios: resource demands are rigid, and over-quota preemption is allowed;

[0034] Low-sensitivity scenarios, flexible resource demand, and support for delayed scheduling;

[0035] S130, mapping the business department to a gaming role, and defining resource demand attributes and gaming chip attributes, wherein the gaming chip attributes include: dynamic budget, reputation points, and business weight;

[0036] S140, generating a matching matrix between business scenarios and game roles, and identifying the adaptation weight of each game role in the current business scenario ,in:

[0037] ;

[0038] in, Score the fitness, indicating the role of the game In business scenarios The specific adaptability under For business scenarios The resource demand sensitivity coefficient is used to reflect the business scenario The demand for resources is rigid. For business scenarios total resource demand.

[0039] S200 collects historical scheduling data for each role, including task completion timeliness, resource utilization, and business value contribution, establishes a bid calculation model, and generates automatic bids based on task urgency coefficient, resource utilization efficiency, and scenario adaptation weight;

[0040] This step collects historical scheduling data for each game role. The purpose of collecting historical scheduling data is to establish a performance profile of each game role in past tasks. The integration of this historical data enables the system to evaluate the efficiency and contribution of each role in resource utilization, thereby providing a scientific basis for subsequent decision-making. This analysis of historical performance means that resource allocation is no longer based on subjective judgment or static rules, but is based on a comprehensive understanding of the real experience and operations of each department. In this way, the system can identify which roles have performed well in past resource utilization and which roles need improvement, and make dynamic adjustments based on this.

[0041] The introduction of a task urgency coefficient further emphasizes the time sensitivity of tasks. This coefficient not only takes into account the remaining time but also incorporates the scenario's time tolerance factor, ensuring that urgent and important tasks are prioritized in resource allocation. This dynamic assessment mechanism ensures timely response to changing business needs, prioritizing resource allocation where it is most needed, thereby maximizing business execution efficiency.

[0042] By setting efficiency and stability indicators, the system can incentivize resource utilization performance among game players. Players with high efficiency and stability are given additional bonus factors to encourage them to maintain high performance in future scheduling. This incentive mechanism not only boosts the enthusiasm of each department but also enhances overall resource utilization efficiency, reducing resource waste and unnecessary competition. This performance-based resource allocation strategy helps cultivate healthy competition among departments, leading to more efficient and coordinated development of the entire system.

[0043] Furthermore, the system uses an efficiency stability metric to evaluate the resource utilization efficiency of each player in historical scheduling. This metric is expressed as the ratio of the standard deviation of historical resource utilization to the mean. This metric not only reflects the stability of a player's resource use but also enables the system to award additional rewards to players with outstanding performance, encouraging their enthusiasm and efficiency in future scheduling. This mechanism has effectively heightened the competitive spirit among players, making departments more cautious when applying for resources, and thus improving overall resource utilization efficiency.

[0044] When calculating automatic bids, the system combines computing resource baselines, task urgency coefficients, and adaptive weights to generate bids. This bidding model is highly flexible and can promptly reflect the true needs and value of each player in the current scenario. Furthermore, to ensure fair resource allocation, the system sets a reputation score threshold. Players with reputation scores below this threshold are penalized by a weighted reduction in their bids. This encourages self-discipline and improvement in resource utilization among each player.

[0045] like Figure 2 As shown, the bid calculation model is established to generate automatic bids based on the task urgency coefficient, resource utilization efficiency and scenario adaptation weight, specifically including:

[0046] S210, based on the current business scenario type and task deadline , calculate the task urgency coefficient :

[0047] ;

[0048] in, Indicates the current time;

[0049] S220, calculating the ratio of the standard deviation to the mean of the historical resource utilization rate of each gaming role, and defining it as an efficiency stability index. Simultaneously, a threshold for the index is set, and an efficiency bonus factor is added to gaming roles whose efficiency stability index is higher than the threshold.

[0050] S230, extract the adaptation weight of each game role in the current business scenario from the matching matrix and calculate the automatic bid :

[0051] ;

[0052] in, Represents resource demand attributes, represents the adaptation weight, represents a dynamic budget, Represents the weight coefficient.

[0053] In this step, the automatic bid is calculated:

[0054] For game players in highly sensitive scenarios, an emergency scenario bonus coefficient is added to the calculated automatic bid;

[0055] Set a score threshold. For game players whose credit score is lower than the score threshold, add a penalty factor to the calculated automatic bid.

[0056] The updated automatic bidding formula is:

[0057] ;

[0058] in, Indicates the emergency scenario bonus coefficient, represents the penalty reduction coefficient.

[0059] S300: Generate a resource allocation heat map based on the automatic bidding results, set a ranking bottom line, allocate exclusive resources to game characters ranked above the ranking bottom line, and automatically trigger elastic resource pool scheduling for the remaining game characters, allocating the remaining resources on demand;

[0060] The resource priority matrix generated in this step is constructed by comprehensively considering bid scenario adaptation weights, efficiency reward factors, and gaming role reputation points. This multi-dimensional evaluation criteria comprehensively reflects the relative importance of each role within the current business scenario, ensuring that resource allocation is based not only on bid levels but also on historical performance and suitability. This scientific prioritization mechanism allows for prioritization of roles that contribute significantly to business value even when resources are limited, effectively improving overall business efficiency.

[0061] By allocating exclusive resources to gaming roles with high resource utilization efficiency and a strong match with the scenario, we implement hard resource isolation. This strategy ensures that important tasks receive the necessary resources at critical moments, preventing them from being preempted by other lower-priority tasks. This hard resource isolation measure, particularly in highly sensitive scenarios, effectively reduces conflicts caused by resource competition and improves system stability and reliability. Furthermore, for low-priority roles, the system automatically triggers elastic resource pool scheduling. This flexible resource allocation mechanism ensures the system maintains resilience in the face of uncertainty and allows it to quickly respond to changing business needs.

[0062] Furthermore, this step can bring transparency to the resource allocation process. The generation of a resource allocation heat map not only visually displays the resource usage of each player, but also helps management quickly identify bottlenecks and potential issues in resource usage. The introduction of this visualization tool not only improves decision-making effectiveness but also provides important data support for subsequent resource scheduling.

[0063] Through scientific priority setting, resource hard isolation, and flexible scheduling mechanisms, efficient resource allocation and utilization are achieved in complex and ever-changing business environments. This not only ensures resource availability for critical tasks and improves overall business execution efficiency, but also provides management with a transparent and intuitive view of resource utilization. This highly flexible and dynamic resource scheduling capability enables enterprises to respond quickly to uncertainty and maintain a competitive advantage.

[0064] like Figure 4 As shown, the resource allocation heat map is generated based on the automatic bidding results, a ranking bottom line is set, exclusive resources are allocated to gaming roles ranked higher than the ranking bottom line, and elastic resource pool scheduling is automatically triggered for the remaining gaming roles to allocate the remaining resources on demand, specifically including:

[0065] S310: Generate a resource priority matrix based on the automatic bidding results. The generation rules are:

[0066] First priority: bid scenario adaptation weight;

[0067] Second priority: efficiency bonus factor;

[0068] The third priority: gaming role reputation points;

[0069] S320: Set a ranking bottom line. For gaming roles ranked higher than the ranking bottom line, implement resource hard isolation:

[0070] Allocate dedicated computing nodes and prohibit other tasks from preempting them;

[0071] Set resource utilization thresholds and automatically expand capacity when the threshold is exceeded;

[0072] S330, after the gaming roles ranked higher than the bottom line have been allocated resources, the remaining resources are divided into a flexible resource pool and a flexible allocation rule is set for allocation.

[0073] S400: Collect the actual resource utilization and task completion indicators of each role, calculate the bid rationality index and input-output ratio, and dynamically adjust the chip attributes of the next scheduling cycle;

[0074] This step, by grading bids based on their efficiency, rationality, and inefficiency, allows the system to quickly identify which players excel in resource utilization and which require improvement. This alignment assessment not only enables the system to effectively adjust itself but also provides a basis for subsequent resource scheduling. Players with efficient bids receive budget increases, creating an incentive mechanism that encourages departments to continue optimizing resource utilization, creating a virtuous cycle. For players with inefficient bids, budget reductions act as a constraint, prompting them to be more cautious in future resource requests and enhancing their awareness of resource self-management.

[0075] Roles with resource utilization rates ≥80% and on-time task completion receive increased reputation points, increasing their priority in subsequent auctions. This not only rewards high-performing roles but also sets an example for others, fostering a strong competitive atmosphere. This mechanism not only fosters fair competition among roles but also enhances overall business execution efficiency. Conversely, roles with resource utilization rates ≤50% and task failures receive credit points deductions and resource application quotas capped. This further strengthens the rigor of resource utilization and enhances the importance of resource utilization across departments.

[0076] Furthermore, the dynamic adjustment capabilities of this step ensure the system's flexibility in resource scheduling. By providing real-time feedback and adjusting chip attributes and auction rule parameters, the system maintains high sensitivity to changes in business scenarios and rapidly responds to changes in the external environment. For example, if demand for a particular business scenario increases dramatically, the system can leverage historical data and real-time feedback to adjust resource allocation for relevant roles, ensuring timely support for critical tasks.

[0077] By building feedback mechanisms and dynamically adjusting strategies, the system not only optimizes resource allocation but also promotes the initiative and rationality of each player in business execution, improving resource utilization efficiency across the entire organization. This strategy ensures that companies can respond flexibly and efficiently in the face of rapidly changing and complex market environments, maintaining their competitive advantage.

[0078] This step dynamically evaluates and adjusts the performance of game characters, optimizing resource scheduling and improving the system's responsiveness to changing business needs. This mechanism makes resource utilization more efficient and rational, ensuring the continuous and stable operation of the company's various businesses, thereby achieving long-term business goals.

[0079] like Figure 5 As shown, the calculation of the bid rationality index and input-output ratio, and the dynamic adjustment of the chip attributes of the next scheduling cycle are as follows:

[0080] S410, evaluating the matching degree between the automatic bidding and the actual performance:

[0081] ;

[0082] in, Represents the game role The matching degree between automatic bidding and actual performance, Represents the game role Automatic bidding for Represents the role of the game the actual effectiveness of

[0083] S420, perform grading based on the matching degree:

[0084] Efficient bidding: >1.2, indicating that resource input-output is higher than the industry average;

[0085] Reasonable bid: 0.8< <1.2, indicating reasonable resource allocation;

[0086] Inefficient bidding: <0.8, indicating resource waste;

[0087] S430: For the gaming role with efficient bidding, the budget for the next scheduling cycle is increased; for the gaming role with inefficient bidding, the budget for the next scheduling cycle is reduced, and the minimum amount is not less than 50% of the original budget;

[0088] S440: For roles with resource utilization ≥ 80% and tasks completed on time, reputation points are increased, raising their priority in subsequent auctions;

[0089] S450: For characters whose resource utilization rate is ≤50% or whose tasks fail, their credit points will be deducted and their resource application quota will be limited.

[0090] S500, identifying tasks that failed due to insufficient resources and generating alternative scheduling path suggestions.

[0091] Figure 6A structural block diagram of a data middle-station scheduling system adapted to multiple business scenarios provided by an embodiment of the present invention, such as Figure 6 As shown, the system includes:

[0092] The game role definition module 100 is used to identify the current business scenario type based on the business feature data collected in real time, abstractly define multiple business departments involved in data scheduling as game roles, and assign resource demand attributes and game chip attributes to each role;

[0093] The role bid calculation module 200 is used to collect historical scheduling data of each role, including task completion timeliness, resource utilization, and business value contribution, establish a bid calculation model, and generate automatic bids based on task urgency coefficient, resource utilization efficiency, and scenario adaptation weight;

[0094] The resource allocation module 300 is used to generate a resource allocation heat map based on the automatic bidding results, set a ranking bottom line, allocate exclusive resources to game characters ranked above the ranking bottom line, and automatically trigger elastic resource pool scheduling for the remaining game characters to allocate the remaining resources on demand;

[0095] Resource utilization monitoring module 400, used to collect actual resource utilization and task completion indicators of each role, calculate bid rationality indicators and input-output ratio, and dynamically adjust chip attributes for the next scheduling cycle;

[0096] The task secondary allocation module 500 is used to identify tasks that fail due to insufficient resources and generate alternative scheduling path suggestions.

[0097] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When executed, the program can include the processes of the above-described method embodiments. Any reference to memory, storage, database, or other media used in the various embodiments provided herein may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0098] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0099] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

[0100] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A data center scheduling method adapted to multiple business scenarios, characterized in that: The method comprises: Based on the real-time collected business feature data, the current business scenario type is identified, and multiple business departments involved in data scheduling are abstractly defined as game roles, and resource demand attributes and game chip attributes are assigned to each role; Collect historical scheduling data for each role, including task completion timeliness, resource utilization, and business value contribution, establish a bid calculation model, and generate automatic bids based on task urgency coefficient, resource utilization efficiency, and scenario adaptation weights; Generate a resource allocation heat map based on the automatic bidding results, set a ranking bottom line, allocate exclusive resources to game characters ranked above the bottom line, and automatically trigger elastic resource pool scheduling for the remaining game characters, allocating remaining resources on demand; Collect the actual resource utilization rate and task completion indicators of each role, calculate the bid rationality index and input-output ratio, and dynamically adjust the chip attributes of the next scheduling cycle; Identify tasks that fail due to insufficient resources and generate alternative scheduling path suggestions; The identification of the current business scenario type and the abstract definition of multiple business departments involved in data scheduling as game roles, and the allocation of resource demand attributes and game chip attributes to each role, specifically include: Real-time collection of multi-source heterogeneous data, including business behavior data, system resource data, and business metadata, and standardization processing to generate structured feature vectors; Based on the structured feature vector, a lightweight classification model is established to identify the current business scenario type and dynamically calibrate the resource sensitivity level of the business scenario, including: Highly sensitive scenarios: resource demands are rigid, and over-quota preemption is allowed; Low-sensitivity scenarios, flexible resource demand, and support for delayed scheduling; Map business departments into gaming roles and define resource demand attributes and gaming chip attributes. Gaming chip attributes include dynamic budget, reputation points, and business weight. Generate a matching matrix between business scenarios and game roles, and identify the adaptation weight of each game role in the current business scenario ,in: ； in, Score the fitness, indicating the role of the game In business scenarios The specific adaptability under For business scenarios The resource demand sensitivity coefficient is used to reflect the business scenario The demand for resources is rigid. For business scenarios Total resource requirements; The establishment of a bid calculation model to generate automatic bids based on task urgency coefficient, resource utilization efficiency, and scenario adaptation weight specifically includes: Based on the current business scenario type and task deadline , calculate the task urgency coefficient : ； in, Indicates the current time; Calculate the ratio of the standard deviation to the mean of each game character's historical resource utilization rate and define it as the efficiency stability index. At the same time, set the index threshold and add an efficiency bonus factor to the game characters whose efficiency stability index is higher than the index threshold. Extract the adaptation weight of each gaming role in the current business scenario from the matching matrix and calculate the automatic bid : ； in, Represents resource demand attributes, represents the adaptation weight, represents a dynamic budget, represents the weight coefficient; The calculation of the bid rationality index and input-output ratio and the dynamic adjustment of the chip attributes of the next scheduling cycle are specifically as follows: Evaluate how well automated bidding matches actual performance: ； in, Represents the role of the game The matching degree between automatic bidding and actual performance, Represents the role of the game Automatic bidding for Represents the role of the game the actual effectiveness of Grading based on matching degree: Efficient bidding: >1.2, indicating that resource input-output is higher than the industry average; Reasonable bid: 0.8< <1.2, indicating reasonable resource allocation; Inefficient bidding: <0.8, indicating resource waste; For game players with efficient bidding, the budget for the next scheduling cycle will be increased; for game players with inefficient bidding, the budget for the next scheduling cycle will be reduced, and the minimum amount will not be less than 50% of the original budget; For roles with resource utilization ≥ 80% and tasks completed on time, reputation points will be increased, raising their priority in subsequent auctions; For characters whose resource utilization rate is ≤50% or whose tasks fail, their credit points will be deducted and their resource application quota will be limited.

2. The method according to claim 1, characterized in that After the automatic bid calculation: For game players in highly sensitive scenarios, an emergency scenario bonus coefficient is added to the calculated automatic bid; Set a score threshold. For game players whose credit score is lower than the score threshold, add a penalty factor to the calculated automatic bid. The updated automatic bidding formula is: ； in, Indicates the emergency scenario bonus coefficient, represents the penalty reduction coefficient.

3. The method according to claim 2, characterized in that The resource allocation heat map is generated based on the automatic bidding results, a ranking bottom line is set, exclusive resources are allocated to gaming roles ranked above the ranking bottom line, and elastic resource pool scheduling is automatically triggered for the remaining gaming roles to allocate the remaining resources on demand, specifically including: Based on the automatic bidding results, a resource priority matrix is ​​generated, and the generation rules are as follows: First priority: bid scenario adaptation weight; Second priority: efficiency bonus factor; The third priority: gaming role reputation points; Set a ranking bottom line, and implement hard resource isolation for gaming roles that rank higher than the bottom line: Allocate dedicated computing nodes and prohibit other tasks from preempting them; Set resource utilization thresholds and automatically expand capacity when the threshold is exceeded; After the gaming roles ranked higher than the bottom line have completed resource allocation, the remaining resources will be divided into elastic resource pools and elastic allocation rules will be set for allocation.

4. The method according to claim 1, wherein The system for implementing the data middle-station scheduling method adapted to multiple business scenarios includes: The game role definition module is used to identify the current business scenario type based on the real-time collected business feature data, abstractly define multiple business departments involved in data scheduling as game roles, and assign resource demand attributes and game chip attributes to each role; The role bid calculation module is used to collect historical scheduling data for each role, including task completion timeliness, resource utilization, and business value contribution, establish a bid calculation model, and generate automatic bids based on task urgency coefficient, resource utilization efficiency, and scenario adaptation weight; The resource allocation module is used to generate a resource allocation heat map based on the automatic bidding results, set a ranking bottom line, allocate exclusive resources to game characters ranked above the ranking bottom line, and automatically trigger elastic resource pool scheduling for the remaining game characters, allocating the remaining resources on demand; The resource utilization monitoring module is used to collect the actual resource utilization rate and task completion indicators of each role, calculate the bid rationality index and input-output ratio, and dynamically adjust the chip attributes of the next scheduling cycle; The task reassignment module is used to identify tasks that fail due to insufficient resources and generate alternative scheduling path suggestions.

Images