CN117608863B - Cloud computing task tracking processing method and system based on intelligent resource allocation - Google Patents

Cloud computing task tracking processing method and system based on intelligent resource allocation Download PDF

Info

Publication number
CN117608863B
CN117608863B CN202410089339.7A CN202410089339A CN117608863B CN 117608863 B CN117608863 B CN 117608863B CN 202410089339 A CN202410089339 A CN 202410089339A CN 117608863 B CN117608863 B CN 117608863B
Authority
CN
China
Prior art keywords
user
time
trust level
value
user trust
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202410089339.7A
Other languages
Chinese (zh)
Other versions
CN117608863A (en
Inventor
李沙莎
金保国
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Liaocheng Luoxi Information Technology Co ltd
Original Assignee
Liaocheng Luoxi Information Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Liaocheng Luoxi Information Technology Co ltd filed Critical Liaocheng Luoxi Information Technology Co ltd
Priority to CN202410089339.7A priority Critical patent/CN117608863B/en
Publication of CN117608863A publication Critical patent/CN117608863A/en
Application granted granted Critical
Publication of CN117608863B publication Critical patent/CN117608863B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/3003—Monitoring arrangements specially adapted to the computing system or computing system component being monitored
    • G06F11/3017—Monitoring arrangements specially adapted to the computing system or computing system component being monitored where the computing system is implementing multitasking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/48—Program initiating; Program switching, e.g. by interrupt
    • G06F9/4806—Task transfer initiation or dispatching
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
    • G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
    • G06F9/5038—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the execution order of a plurality of tasks, e.g. taking priority or time dependency constraints into consideration
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

The invention relates to the field of resource allocation of electric digital processing, in particular to a cloud computing task tracking processing method and a cloud computing task tracking processing system based on intelligent resource allocation, which are used for acquiring required service time and user trust level of each user process in a cloud server; for the current process, obtaining the number of planning time slices according to the trust level of the user and the required service time; obtaining tolerance values according to the number of planning time slices, the percentile coefficient and the number of running time slices of the current process; optimizing the time slice rotation process by combining the tolerance value with the rotation mark in the time slice rotation algorithm; obtaining a new user trust level according to the difference between the required service time and the actual running time, the tolerance value and the user trust level; and optimizing the computing power resource of the scheduling server according to the trust level of the new user of each user process. The invention avoids the resource preemption caused by the false work of the user and completes the intelligent allocation and scheduling of the computing power resources of the cloud server.

Description

Cloud computing task tracking processing method and system based on intelligent resource allocation
Technical Field
The application relates to the field of resource allocation of electric digital processing, in particular to a cloud computing task tracking processing method and system based on intelligent resource allocation.
Background
In the age of industrial big data, the demands of industry on computing resources are increasing in a explosive manner, and cloud computing server technology is generated. The cloud computing server provides various cloud computing services such as computing, storage and network for users, and the server can automatically configure various resources in the server according to the process proposed by the users, so that the effect of sharing the computing resources and improving the utilization rate of the data equipment is achieved.
In the resource scheduling of cloud service, the situation that a user process needs to wait in a queue usually occurs in a peak period of the use of an operation process, and at this time, a relevant algorithm for computing resource allocation by a server is needed to complete the allocation of computing resources.
In the conventional process of calculating resource scheduling by a server, a commonly used algorithm is a multi-stage feedback queue scheduling algorithm, which is a resource scheduling algorithm in which a time slice rotation scheduling algorithm and a priority scheduling algorithm are mixed, the priorities of processes are designated by externally set priorities, the processes are ordered according to the priorities, and the time slice rotation algorithm is adopted to replace the process prediction running time given by a user, so that the abnormal resource scheduling caused by the wrong process time given by the user maliciously is prevented.
The algorithm sorts the processes according to the process priority, and if a process with a higher priority continuously enters, starvation phenomenon can be caused, so that a process with a low priority cannot be solved for a long time; when the time slice rotation algorithm is adopted, the process prediction running time given by a user is completely abandoned, and if the time slice classification of the process is unreasonable or the size of the time slice is not matched with the overall condition of the process, a high-frequency process switching phenomenon can occur, so that the calculation cost is increased.
Disclosure of Invention
In order to solve the technical problems, the invention provides a cloud computing task tracking processing method and a cloud computing task tracking processing system based on intelligent resource allocation, and the adopted technical scheme is as follows:
in a first aspect, an embodiment of the present invention provides a cloud computing task tracking processing method based on intelligent resource allocation, where the method includes the following steps:
acquiring required service time and user trust level of each user process in a cloud server;
Obtaining a percentile coefficient of the user process according to the user trust level of the user process; for the current process, obtaining a dynamic allocation time discrimination value according to the user trust level and the required service time; obtaining the time slice size according to the dynamic allocation time discrimination value and the required service time; correcting according to the dynamic allocation time discrimination value to obtain the number of planning time slices;
Obtaining tolerance values according to the number of planning time slices, the percentile coefficient and the number of running time slices of the current process; optimizing the time slice rotation process by combining the tolerance value with the rotation mark in the time slice rotation algorithm; obtaining a new user trust level according to the difference between the required service time and the actual running time, the tolerance value and the user trust level;
and optimizing the computing power resource of the scheduling server according to the trust level of the new user of each user process.
Preferably, the obtaining the percentile coefficient of the user process according to the user trust level of the user process includes:
When the user trust level of the user process accounts for the first 10% in all user ranks, the percentile coefficient of the user process is 0.55; when the user trust level of the user process accounts for the first 10% -20% in all user ranks, the percentile coefficient of the user process is 0.60; and so on, when the user trust level of the user process accounts for the first 90% -100% of all user ranks, the percentile coefficient of the user process is 1.
Preferably, the obtaining the dynamic allocation time discrimination value according to the user trust level and the required service time includes:
acquiring the number of unfinished computing processes in a current server;
Acquiring the sum of user trust levels of all unfinished computing processes of a current server, and computing the ratio of the sum of the user trust levels of the current process and the user trust levels to which the current process belongs as a first ratio;
Acquiring the sum of the required service time of all unfinished computing processes of the current server, and computing the ratio of the required service time of the current process to the sum of the required service time as a second ratio;
And taking the product of the difference value of the first ratio and the second ratio and the number of processes as a dynamic allocation time discrimination value.
Preferably, the obtaining the time slice size according to the dynamic allocation time discrimination value and the required service time includes:
when the dynamic allocation time discrimination value is less than or equal to 1, the service time is required to be used as the time slice size;
when the dynamic allocation time discrimination value is greater than 1, the ratio of the service time to the upward rounding value of the dynamic allocation time discrimination value is required as the time slice size.
Preferably, the correcting according to the dynamic allocation time discrimination value to obtain the number of planned time slices includes:
When the dynamic allocation time discrimination value is larger than 1, taking the upward rounding value of the dynamic allocation time discrimination value as the number of planning time slices;
When the dynamic allocation time discrimination value is less than or equal to 1, 1 is taken as the number of planning time slices.
Preferably, the obtaining the tolerance value according to the number of planned time slices, the percentile coefficient and the number of running time slices of the current process includes:
when the number of times of the running time slices of the current process is smaller than or equal to the number of times of the planning time offset, setting the tolerance value to 0;
When the number of times of the running time slices of the current process is larger than the number of times of the planning time deviations, calculating a difference value result between the number of times of the running time slices of the current process and the number of times of the planning time slices and taking a product of the number of times of the running time slices of the current process as a first product; calculating the product of the number of planning time slices and the number of time slices which do not finish running of the current server as a second product;
And taking the downward rounded value of the product of the ratio result of the first product and the second product and the percentile coefficient as the tolerance value.
Preferably, the step of obtaining the new user trust level according to the difference between the required service time and the actual running time, the tolerance value and the user trust level adjustment includes:
Obtaining a user trust level adjustment scaling factor according to the user trust level, the time slice size and the running time slice times when the current process is finished;
Obtaining a user trust level adjustment proportion according to the user trust level adjustment proportion coefficient;
Calculating the product of the difference value of subtracting the user trust level adjustment proportion from 1 and the user trust level as a third product;
calculating the average value of all user trust levels of the current server, calculating the difference value between the new user protection coefficient and the tolerance value, and calculating the product of the difference value, the average value and the user trust level adjustment proportion as a fourth product;
And taking the sum value of the third product and the fourth product as a new user trust level.
Preferably, the obtaining the user trust level adjustment scaling factor according to the user trust level, the time slice size, and the number of times of the time slices that have been run when the current process ends includes:
Calculating the product of the time slice size and the number of time slices operated by the current process as a fifth product, and calculating the product of the average value of the time slice sizes of all the unfinished processes and the average value of the time slice numbers of all the unfinished processes as a sixth product; calculating the ratio of the fifth product to the sixth product;
The ratio result of the user trust level to the average value of all the user trust levels and the absolute value of the difference value of the average value of all the user trust levels are obtained; and taking the product of the result of subtracting the absolute value of the difference value from 1 and the ratio as a user trust level adjustment scaling factor.
Preferably, the obtaining the user trust level adjustment ratio according to the user trust level adjustment ratio coefficient includes:
when the user trust level adjustment scaling factor is greater than 0.9, setting the user trust level adjustment scaling to 0.9;
When the user trust level adjustment scaling factor is less than or equal to 0.1, setting the user trust level adjustment scaling to 0.1;
Otherwise, the user trust level adjustment proportion is set as a user trust level adjustment proportion coefficient.
In a second aspect, an embodiment of the present invention further provides a cloud computing task tracking processing system based on intelligent resource allocation, where the cloud computing task tracking processing system includes a memory, a processor, and a computer program stored in the memory and running on the processor, where the processor implements the steps of any one of the methods described above when executing the computer program.
The invention has at least the following beneficial effects:
The invention gives the allocation of the computing resources to the users requiring the service time for the long-term faithful report by constructing the user trust level, reduces the allocation of the computing resources to the users requiring the service time for the false report, eliminates the risk of registration attack by setting the update of the user trust level, and avoids the resource preemption problem caused by the falsification of the users while fully utilizing the requiring service time;
Meanwhile, the service time information is added to the time slice division, so that a high-frequency process switching phenomenon is avoided, and the calculation cost is reduced; finally, compared with the traditional time slice rotation algorithm, the method and the device utilize the service time to reduce the calculation cost, prevent the resource preemption caused by the false work of the user, and complete the intelligent resource allocation of the cloud computing server.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a cloud computing task tracking processing method based on intelligent resource allocation provided by the invention;
FIG. 2 is a flowchart of an optimized scheduling algorithm.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description is given below of the cloud computing task tracking processing method and system based on intelligent resource allocation according to the invention, and the detailed description of the specific implementation, structure, characteristics and effects thereof is given below with reference to the accompanying drawings and the preferred embodiment. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the cloud computing task tracking processing method and system based on intelligent resource allocation provided by the invention with reference to the accompanying drawings.
The embodiment of the invention provides a cloud computing task tracking processing method and a cloud computing task tracking processing system based on intelligent resource allocation.
Specifically, the following cloud computing task tracking processing method based on intelligent resource allocation is provided, please refer to fig. 1, and the method includes the following steps:
Step S001, obtaining the required service time of each user process in the cloud server.
In the computing resource scheduling of the cloud server, in order to perform reasonable computing power allocation scheduling on each user process, so that the computing resources of the cloud server are utilized to the greatest extent, and the computing power resource waste is avoided, in this embodiment, the process prediction running time of each user process to be run is obtained, and the time is the time that the process given by the user in advance may run, and is generally named as the required service time。
So far, the service time required by each user process can be obtained according to the method。
Step S002, the time slice size of each user process is adjusted according to the service time required by each user process and the user trust level, and the user trust level is adjusted in time.
In order to prevent users from maliciously reporting the process prediction running time while fully utilizing the process prediction running time, the embodiment adopts the steps of setting trust level for each userScoring users according to trust level/>Performing resource allocation on a process submitted by a user; calculating trust level/>, in a manner that penalizes users of malicious reporting process prediction run time and rewards users of correct reporting process run time。
The trust level of the user can dynamically change along with the calculation condition of the progress submitted by the user; for the user using the cloud server for the first time, the trust level is an initial value。
Each process entered into the cloud computing server will obtain a user trust levelPercentile coefficients/>, ranking user trust levels among all usersAs its characteristic value. The present embodiment defines the percentile coefficients of all user ranks as: when the user trust level/>When the user rank is the top 10%, then/>The user ranks 10% to 20%,/>The user ranks 20% to 30%,/>By analogy, when the user ranks top 90% to 100% >。
In the method of the embodiment, a time slice round-robin scheduling algorithm is adopted to allocate resources to each process, so as to eliminate the problem of unsuitable time slice size easily occurring in the traditional time slice round-robin scheduling algorithm, and the time slice size dynamically allocated to the process of each newly added cloud computing server is combined with the trust level of the userThe formula is as follows:
in the method, in the process of the invention, Representing the time slice size of the current process allocation,/>Indicating the required service time of the current process,Is the dynamic allocation time discrimination value of the current process,/>Representing an upward rounding symbol;
is an adjustable super parameter, represents the size division scale of the time slice, and takes the checked value 2 in the embodiment; /(I) The sum of the required service time of all unfinished computing processes in the current server; /(I)Is the user trust level of the current process; /(I)Is the sum of the user trust levels of all outstanding computing processes in the current server, +1 is to prevent the denominator from being 0; /(I)Is the number of outstanding computing processes in the current server, where/>For the first ratio of the current process,/>A second ratio of the current process.
Used in the formulaDivided by/>The ratio of the required service time of the current process to the required service time of all processes is evaluated, and the ratio is multiplied by/>If/>Equal to the average of all program demand service times, then the resulting value is/>; If/>Less than the average of all program demand service times, the resulting value is less than/>; If it isGreater than the average of all program demand service times, the resulting value is greater than/>The time slices are dynamically divided according to the value, the larger the value is, the longer the required service time of the process is relative to the whole, and the required service time is divided into more time slices, so that the phenomenon of server congestion waiting caused by the longer processing process of the current program is avoided, and the resource waste is caused.
Further, the method comprises the steps of,Divided by/>The trust level ranking condition of the user trust level corresponding to the current process in all current binaries is obtained. The higher the value is, the higher the trust level of the user is, the better the cloud server usage record representing the user is, and certain rewards are correspondingly given to the cloud server, so that the cloud server obtains more server resources, and finally the larger the time slice size of the current process is.
To sum up, inWhen the service time is less than or equal to 1, the service time required by the current process is generally less than the size of a time slice in the system, or the larger the user trust level of the current process is compared with the proportion of the service time required by the user trust level of the current process to all unfinished computing processes in the current server, the size of the time slice allocated to the current process is equal to the service time required by the user trust level of the current process, so that the user trust level can run in one time slice; when/>Greater than 1, representing that the current process is longer, multiple time slices should be allocated to it, avoiding the program from taking up computing resources.
The final dynamic allocation effect of the time slices is: the users with high user trust level are distributed to larger time slices, the size of the time slices is dynamically adjusted by combining the required service time of all processes in the current server, the high-frequency switching process is avoided, meanwhile, rewards are given to the users with high user trust level, and server resources are reasonably and dynamically distributed.
In order to prevent users from maliciously reporting required service time, a tolerance value is set according to a time slice in which actual calculation of a process is performed, the process exceeding the required service time is classified according to the tolerance value, computing resources are dynamically allocated, users requiring the service time for maliciously reporting are punished in time, and meanwhile, the processes of other users are guaranteed to be normally solved, and the method comprises the following specific processes:
First, dynamically allocating time discrimination values Performing numerical correction to obtain the number of planning time slices of the current processThe method is characterized by comprising the following steps:
in the method, in the process of the invention, Representing the number of planning time slices of the current process,/>Is the dynamic allocation time discrimination value of the current process,/>Representing rounding up symbols.
When (when)When the time is less than or equal to 1, the service time required for representing the current process is shorter and the priority is possibly higher, and the process can be completed within 1 time slice according to the planning of the process, so the number of times of planning the time slice is 1; when/>Above 1, it represents that the current process is too large and may be of lower priority, as planned in/>The process can be completed in the secondary time slices, so that the number of times of planning the time slices is/>。
Obtaining tolerance value of the current process based on the planned time slice times, the operated time slice times and the time slice times of unfinished operation of the current process：
In the method, in the process of the invention,Tolerance value representing current progress,/>Is the number of run-time slices of the current process,/>Is the number of planning time slices of the current process,/>Is the number of time slices of incomplete runs of all programs in the current server,/>Is a percentile coefficient of the user trust level of the current process, where/>Is the first product of the current process,/>Is the second product of the current process.
In the formula, when the number of running time slices of the current process is smaller than or equal to the number of planning time slices, the tolerance value of the current process is set to 0, namely, when the number of running time slices is within a reasonable range, the current process is not likely to be a malicious program, so that the reasonable process is not tolerant in the running process; when the number of time slices that the current process has run is greater than the number of time slices planned,Is the number of time slices that the current process has run minus the number of planning time slices for the current process, divided by/>Representing how many times the current process is run more than the number of time slices planned; the more times a number of runs, the greater the value obtained, the more likely that a service time is required to preempt resources on behalf of a user for malicious reporting, and the more computing resources that the process has preempted maliciously, so that upon subsequent allocation of computing resources, they should be de-allocated, corresponding tolerance values/>The larger should be the need to reduce the time slice size of the current process.
The ratio of the number of time slices of the current program to the number of time slices of the overall planning of the server is represented by the ratio, the larger the value is, the more the current program has acquired the computing resources, the less the allocation should be performed to the current program to ensure the resource allocation fairness and prevent the starvation phenomenonThe larger should be.
According to tolerance valueIn the method described in this embodiment, the flow of each time slice calculation is as follows:
In the current round, all processes go through the round robin flag bit To determine whether to calculate the process.
Program initialization for joining a server process: when a process just joins the calculation server, the characteristic value is obtained through calculation: user trust levelPercentile coefficients/>Time slice size/>Time slice number of planning/>. Then the process is added into the area to be calculated of the calculation server, and the rotation flag bit/>, of the process is turnedSet to 1.
Loop calculation flow of server calculation element: traversing the rotation zone bit of all processes of the computing server to be computed when the computing element of the server is idleIf the cycle flag bit is traversed/>Are not 0, the rotation flag bit/>Subtracting 1; if the process with the rotation flag bit of 0 is traversed, the rotation flag bit/>, is executedAdd/>Simultaneously reading the time slice size/>, of the process corresponding to the rotation flag bit。
If inRemoving the process from the area to be calculated of the calculation server and outputting a calculation result after the process is calculated in the time; if at/>Interrupting the process when the process is not calculated in time, and calculating tolerance value/>, of the processRotate the flag bit/>Add/>. According to experience/>Is 3, and avoids starvation of the process due to too high tolerance values.
In summary, by adding tolerance values in the time-rotation algorithm processCompared with the traditional time rotation algorithm, the method adopts a time slice rotation hierarchical ranking mode, and prevents users who maliciously report the required service time from occupying computing resources while fully utilizing the required service time of each process.
The initial value of the user trust level is set to 0, and the trust level of the user needs to be adjusted according to the difference between the required service time of the process reported by the user and the actual running time in the running process. The specific process is as follows:
after one process is finished, the user trust level is updated for the user who proposes the process, and the method specifically comprises the following steps:
in the method, in the process of the invention, Is a new user trust level,/>Is the user trust level,/>Is the user trust level adjustment ratio,/>The new user protection coefficient is equal to 2 when the user trust level is adjusted for the first three times for a newly registered user, and then the new user protection coefficient is 0, so that the user trust level of the new user is quickly increased, and the new user is quickly influenced by the algorithm described in the embodiment; /(I)Is the tolerance value of the current process after the current process is finished; /(I)Is the average of the trust levels of all users when the users do not update the trust levels of the users;
is the user trust level adjustment scaling factor,/> Is the time slice size of the current process,/>Is the number of time slices that have been run at the end of the current process,/>Is the average value of the time slice sizes in all the running processes; /(I)Is the average of the number of time slices in all the unexecuted processes, where/(v >)As a result of the third product of the products,Is the fourth product,/>As a result of the fifth product of the products,Is the sixth product.
In the formulaIn/>The absolute value of the user trust level average value of all the running processes is subtracted from the user trust level, the smaller the value is, the closer the trust level of the current user is to the average value, and in order to improve the difference of the user trust levels among different users, the larger the user trust level adjustment proportion coefficient is opposite to the value, so that the trust level of the user can deviate from the average value rapidly, and the trust level of the user has rapid variability;
representing the ratio of the running time of the current process to the required service time of all the incompletely running processes, wherein the larger the value is, the longer the running time of the current process is, the larger the content of the current process is, and the ratio/>, of trust level adjustment of a user, after the corresponding current process is ended The larger the server resources are used because the content usage is large.
AndIs to prevent the current user trust level from deviating too high from the mean value or the running time of the current process from being too large or too small, thereby leading to/>Too large or too small a value of (1) makes the user trust level adjustment ratio unreasonable, thus limiting the user trust level adjustment ratio to/>Within a range of (2).
Finally in the formulaIn/>The larger the user trust level of the user corresponding to the current process is, the smaller the ratio of the user trust level is in the new user trust level, and the minimum ratio/>；/>An adjustment term called user trust level,/>Representing the difference between the required service time and the actual running time of the current process corresponding to the user report, and if the actual running time of the process is about twice as long as the required service time of the user report, the adjustment item of the user trust level of the user is a positive value; if about two times and more are reached, the adjustment item of the user trust level is 0 or a negative value; adjustment of the previous three user trust levels for newly registered users by/>Making the final adjustment result bigger; /(I)Is an adjustment scale for defining a user trust level, the scale is related to the overall user trust level, +1 is used for preventing the user trust level from being unable to be adjusted due to the initial condition that all the user trust levels are 0,/>Is the ratio of the adjustment item representing the user trust level in the new user trust level,/>The larger the duty cycle, the more aggressive the user level adjustment.
And step S003, optimizing the computing power resource of the scheduling server according to the trust level of the new user of each user.
Finally, the process of each user can calculate a new user trust level, when the process ends the current time slice, the new user trust level condition can be adjusted, and the larger the process given by the user or the closer the trust level of the user is to the average value of all the user trust levels, the larger the adjustment amplitude is; the required service time and the actual running time reported by the user are not greatly different, and the tolerance value is improvedEqual to 0, the user trust level generally rises; the reported required service time of the user is much shorter than the actual run time, then tolerance value/>Above 0, the user's trust level decreases, possibly reporting a service time requirement maliciously on behalf of the user.
When the trust level of the user drops to a negative number, the user is marked as a malicious user, and access to the cloud computing server is forbidden.
And for a new user, the user trust level rises faster, but the trust level of the old user is always higher than that of the new user, and the higher the user trust, the more resources the process allocates to than the new user. The method and the device can ensure that the new user is influenced by the algorithm quickly, and avoid that the user maliciously refreshes the trust level of the user to acquire additional computing resources in a new user registration mode.
In summary, the flowchart of the optimized scheduling algorithm of this embodiment is shown in fig. 2. In fig. 2, for a newly registered user, the user trust level is set to 0, and the newly registered user is added into the user trust level database; the user sends out the process, divides the time slices for the process by calculating the trust level of the user, and enters the time slice rotation; and after the process calculation is finished, recording the tolerance value and the time slice rotation times, calculating the updated user trust level, and updating the trust level of the user in a user trust level database.
According to the method, the operation program of the cloud computing server is optimized, dynamic task allocation and scheduling of the cloud computing server are completed, and a better operation effect of the cloud computing server is obtained.
Based on the same inventive concept as the method, the embodiment of the invention also provides a cloud computing task tracking processing system based on intelligent resource allocation, which comprises a memory, a processor and a computer program stored in the memory and running on the processor, wherein the processor realizes the steps of any one of the cloud computing task tracking processing methods based on intelligent resource allocation when executing the computer program.
According to the embodiment of the invention, the user trust level is constructed, the allocation of the computing resources is given to the users requiring the service time for long-term faithful reporting, the allocation of the computing resources is reduced for the users requiring the service time for false reporting, meanwhile, the risk of registration attack is eliminated through the setting of the user trust level update, and the resource preemption problem caused by the falsification of the users is avoided while the service time is fully utilized;
meanwhile, the service time information is added to the time slice division, so that a high-frequency process switching phenomenon is avoided, and the calculation cost is reduced; finally, compared with the traditional time slice rotation algorithm, the embodiment of the invention utilizes the service time to reduce the calculation cost, prevents the resource preemption caused by the false work of the user, and completes the intelligent allocation of the resources of the cloud computing server.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. And the foregoing description has been directed to specific embodiments of this specification. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and the same or similar parts of each embodiment are referred to each other, and each embodiment mainly describes differences from other embodiments.
The above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; the technical solutions described in the foregoing embodiments are modified or some of the technical features are replaced equivalently, so that the essence of the corresponding technical solutions does not deviate from the scope of the technical solutions of the embodiments of the present application, and all the technical solutions are included in the protection scope of the present application.

Claims (4)

1. The cloud computing task tracking processing method based on intelligent resource allocation is characterized by comprising the following steps of:
acquiring required service time and user trust level of each user process in a cloud server;
Obtaining a percentile coefficient of the user process according to the user trust level of the user process; for the current process, obtaining a dynamic allocation time discrimination value according to the user trust level and the required service time; obtaining the time slice size according to the dynamic allocation time discrimination value and the required service time; correcting according to the dynamic allocation time discrimination value to obtain the number of planning time slices;
Obtaining tolerance values according to the number of planning time slices, the percentile coefficient and the number of running time slices of the current process; optimizing the time slice rotation process by combining the tolerance value with the rotation mark in the time slice rotation algorithm; obtaining a new user trust level according to the difference between the required service time and the actual running time, the tolerance value and the user trust level;
Optimizing computing power resources of a scheduling server according to the trust level of the new user of each user process;
the obtaining the percentile coefficient of the user process according to the user trust level of the user process comprises the following steps:
When the user trust level of the user process accounts for the first 10% in all user ranks, the percentile coefficient of the user process is 0.55; when the user trust level of the user process accounts for the first 10% -20% in all user ranks, the percentile coefficient of the user process is 0.60; and so on, when the user trust level of the user process accounts for the first 90% -100% in all user ranks, the percentile coefficient of the user process is 1;
The obtaining the dynamic allocation time discrimination value according to the user trust level and the required service time comprises the following steps:
acquiring the number of unfinished computing processes in a current server;
Acquiring the sum of user trust levels of all unfinished computing processes of a current server, and computing the ratio of the sum of the user trust levels of the current process and the user trust levels to which the current process belongs as a first ratio;
Acquiring the sum of the required service time of all unfinished computing processes of the current server, and computing the ratio of the required service time of the current process to the sum of the required service time as a second ratio;
Taking the product of the difference value of the first ratio and the second ratio and the number of processes as a dynamic allocation time discrimination value;
the obtaining the time slice size according to the dynamic allocation time discrimination value and the required service time comprises the following steps:
when the dynamic allocation time discrimination value is less than or equal to 1, the service time is required to be used as the time slice size;
When the dynamic allocation time discrimination value is larger than 1, the ratio of the required service time to the upward rounding value of the dynamic allocation time discrimination value is used as the time slice size;
The step of correcting according to the dynamic allocation time discrimination value to obtain the number of planned time slices comprises the following steps:
When the dynamic allocation time discrimination value is larger than 1, taking the upward rounding value of the dynamic allocation time discrimination value as the number of planning time slices;
when the dynamic allocation time discrimination value is less than or equal to 1, taking 1 as the number of planning time slices;
obtaining tolerance values according to the planned time slice times, the percentile coefficients and the time slice times of the current process, wherein the tolerance values comprise:
when the number of times of the running time slices of the current process is smaller than or equal to the number of times of the planning time offset, setting the tolerance value to 0;
When the number of times of the running time slices of the current process is larger than the number of times of the planning time deviations, calculating a difference value result between the number of times of the running time slices of the current process and the number of times of the planning time slices and taking a product of the number of times of the running time slices of the current process as a first product; calculating the product of the number of planning time slices and the number of time slices which do not finish running of the current server as a second product;
taking the downward rounding value of the product of the ratio result of the first product and the second product and the percentile coefficient as a tolerance value;
the step of obtaining the new user trust level according to the difference between the required service time and the actual running time, the tolerance value and the user trust level adjustment comprises the following steps:
Obtaining a user trust level adjustment scaling factor according to the user trust level, the time slice size and the running time slice times when the current process is finished;
Obtaining a user trust level adjustment proportion according to the user trust level adjustment proportion coefficient;
Calculating the product of the difference value of subtracting the user trust level adjustment proportion from 1 and the user trust level as a third product;
calculating the average value of all user trust levels of the current server, calculating the difference value between the new user protection coefficient and the tolerance value, and calculating the product of the difference value, the average value and the user trust level adjustment proportion as a fourth product;
And taking the sum value of the third product and the fourth product as a new user trust level.
2. The cloud computing task tracking processing method based on intelligent resource allocation according to claim 1, wherein the obtaining the user trust level adjustment scaling factor according to the user trust level, the time slice size, and the number of time slices that have been run when the current process is finished comprises:
Calculating the product of the time slice size and the number of time slices operated by the current process as a fifth product, and calculating the product of the average value of the time slice sizes of all the unfinished processes and the average value of the time slice numbers of all the unfinished processes as a sixth product; calculating the ratio of the fifth product to the sixth product;
The ratio result of the user trust level to the average value of all the user trust levels and the absolute value of the difference value of the average value of all the user trust levels are obtained; and taking the product of the result of subtracting the absolute value of the difference value from 1 and the ratio as a user trust level adjustment scaling factor.
3. The cloud computing task tracking processing method based on intelligent resource allocation according to claim 1, wherein the obtaining the user trust level adjustment scale according to the user trust level adjustment scale coefficient comprises:
when the user trust level adjustment scaling factor is greater than 0.9, setting the user trust level adjustment scaling to 0.9;
When the user trust level adjustment scaling factor is less than or equal to 0.1, setting the user trust level adjustment scaling to 0.1;
Otherwise, the user trust level adjustment proportion is set as a user trust level adjustment proportion coefficient.
4. Cloud computing task tracking processing system based on intelligent resource allocation, comprising a memory, a processor and a computer program stored in the memory and running on the processor, characterized in that the processor implements the steps of the method according to any of claims 1-3 when executing the computer program.
CN202410089339.7A 2024-01-23 2024-01-23 Cloud computing task tracking processing method and system based on intelligent resource allocation Active CN117608863B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202410089339.7A CN117608863B (en) 2024-01-23 2024-01-23 Cloud computing task tracking processing method and system based on intelligent resource allocation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202410089339.7A CN117608863B (en) 2024-01-23 2024-01-23 Cloud computing task tracking processing method and system based on intelligent resource allocation

Publications (2)

Publication Number Publication Date
CN117608863A CN117608863A (en) 2024-02-27
CN117608863B true CN117608863B (en) 2024-04-23

Family

ID=89960160

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202410089339.7A Active CN117608863B (en) 2024-01-23 2024-01-23 Cloud computing task tracking processing method and system based on intelligent resource allocation

Country Status (1)

Country Link
CN (1) CN117608863B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101308468A (en) * 2008-06-13 2008-11-19 南京邮电大学 Grid calculation environment task cross-domain control method
US10275267B1 (en) * 2012-10-22 2019-04-30 Amazon Technologies, Inc. Trust-based resource allocation
CN111767134A (en) * 2020-05-18 2020-10-13 鹏城实验室 Multitask dynamic resource scheduling method
CN115408152A (en) * 2022-08-23 2022-11-29 吉兴信(广东)信息技术有限公司 Adaptive resource matching obtaining method and system
CN116302509A (en) * 2023-02-21 2023-06-23 中船（浙江）海洋科技有限公司 Cloud server dynamic load optimization method and device based on CNN-converter
CN116991585A (en) * 2023-08-11 2023-11-03 上海数珩信息科技股份有限公司 Automatic AI calculation power scheduling method, device and medium

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101308468A (en) * 2008-06-13 2008-11-19 南京邮电大学 Grid calculation environment task cross-domain control method
US10275267B1 (en) * 2012-10-22 2019-04-30 Amazon Technologies, Inc. Trust-based resource allocation
CN111767134A (en) * 2020-05-18 2020-10-13 鹏城实验室 Multitask dynamic resource scheduling method
WO2021233261A1 (en) * 2020-05-18 2021-11-25 鹏城实验室 Multi-task dynamic resource scheduling method
CN115408152A (en) * 2022-08-23 2022-11-29 吉兴信(广东)信息技术有限公司 Adaptive resource matching obtaining method and system
CN116302509A (en) * 2023-02-21 2023-06-23 中船（浙江）海洋科技有限公司 Cloud server dynamic load optimization method and device based on CNN-converter
CN116991585A (en) * 2023-08-11 2023-11-03 上海数珩信息科技股份有限公司 Automatic AI calculation power scheduling method, device and medium

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
曹洁 ; 曾国荪 ; 姜火文 ; 马海英 ; .云环境下服务信任感知的可信动态级调度方法.通信学报.2014,(11),全文. *
束柬 ; 梁昌勇 ; 徐健 ; .基于信任的云服务系统多目标任务分配模型.计算机研究与发展.2018,(06),全文. *

Also Published As

Publication number Publication date
CN117608863A (en) 2024-02-27

Similar Documents

Publication Publication Date Title
CN111045795A (en) Resource scheduling method and device
CN108270805B (en) Resource allocation method and device for data processing
CN106789118B (en) Cloud computing charging method based on service level agreement
CN112181613B (en) Heterogeneous resource distributed computing platform batch task scheduling method and storage medium
CN109976901B (en) Resource scheduling method, device, server and readable storage medium
CN114327811A (en) Task scheduling method, device and equipment and readable storage medium
CN105488134A (en) Big data processing method and big data processing device
CN111176840A (en) Distributed task allocation optimization method and device, storage medium and electronic device
CN111221631A (en) Task scheduling method, device and storage medium
CN117608863B (en) Cloud computing task tracking processing method and system based on intelligent resource allocation
CN115640113A (en) Multi-plane flexible scheduling method
CN117608840A (en) Task processing method and system for comprehensive management of resources of intelligent monitoring system
CN112862312B (en) Manufacturing service resource dynamic scheduling method and system based on random online algorithm
CN111488209B (en) Heuristic Storm node task scheduling optimization method
CN111813546B (en) Resource allocation method, system and related device for multi-network connection application
CN110427217B (en) Content-based publish-subscribe system matching algorithm lightweight parallel method and system
CN112598284A (en) Limited resource dynamic allocation method for endowment system
CN114064294A (en) Dynamic resource allocation method and system in mobile edge computing environment
Thai et al. Algorithms for optimising heterogeneous Cloud virtual machine clusters
CN106886451B (en) Multi-workflow task allocation method based on virtualization container technology
CN110941493A (en) Task scheduling method and device
CN112598112A (en) Resource scheduling method based on graph neural network
CN112306642A (en) Workflow scheduling method based on stable matching game theory
CN114816720B (en) Scheduling method and device of multi-task shared physical processor and terminal equipment
CN114443258B (en) Resource scheduling method, device, equipment and storage medium for virtual machine

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant