CN117851029A - Cloud rendering server resource access control method, device, equipment and medium - Google Patents

Cloud rendering server resource access control method, device, equipment and medium Download PDF

Info

Publication number
CN117851029A
CN117851029A CN202311619454.2A CN202311619454A CN117851029A CN 117851029 A CN117851029 A CN 117851029A CN 202311619454 A CN202311619454 A CN 202311619454A CN 117851029 A CN117851029 A CN 117851029A
Authority
CN
China
Prior art keywords
server
rendering
resource access
access request
quality prediction
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.)
Pending
Application number
CN202311619454.2A
Other languages
Chinese (zh)
Inventor
丁刚毅
崔程远
关正
吴羽琛
唐明湘
梁栋
闫大鹏
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.)
Beijing Institute of Technology BIT
Original Assignee
Beijing Institute of Technology BIT
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 Beijing Institute of Technology BIT filed Critical Beijing Institute of Technology BIT
Priority to CN202311619454.2A priority Critical patent/CN117851029A/en
Publication of CN117851029A publication Critical patent/CN117851029A/en
Pending legal-status Critical Current

Links

Classifications

    • 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

Landscapes

  • Image Generation (AREA)

Abstract

The present invention relates to the field of cloud rendering technologies, and in particular, to a method, an apparatus, a device, and a medium for controlling resource access of a cloud rendering server. According to the invention, the rendering quality prediction model is used, and the rendering quality prediction result output by the rendering quality prediction model is obtained according to the resource occupation parameter and the server parameter, so that the rendering resources can be distributed according to the rendering quality prediction result; compared with the traditional concurrent resource allocation method based on the fixed threshold, the method provided by the invention has the advantages that the rendering quality of the cloud rendering server in the critical state is obviously improved, and the system stability and the user experience are further improved.

Description

Cloud rendering server resource access control method, device, equipment and medium
Technical Field
The present invention relates to the field of cloud rendering technologies, and in particular, to a method, an apparatus, a device, and a medium for controlling resource access of a cloud rendering server.
Background
Models in metaspace are generally complex in design, large in number of faces and strong in interactivity, so that running and rendering costs are high, and traditional terminal equipment is difficult to provide such strong computational support. Therefore, the cloud rendering mode is currently the mainstream. In a cloud rendering mode, a plurality of users and a cloud rendering server are connected through a network, a user side transmits a control flow to the rendering server, the rendering server decodes the flow content after receiving the control flow of the users to obtain control information of the users, a system performs rendering according to the control information to obtain audio and video flow data corresponding to the users, and then the audio and video flow is transmitted to user side equipment through the network and output.
In the prior art, for a plurality of users, a commonly adopted method is time sequence-based queuing control, whether a rendering request is taken out of a queuing queue or not is judged based on a fixed threshold value, and the rendering request is added into a ready queue, and a cloud server redistributes corresponding resources to tasks of the ready queue and converts the tasks into running tasks.
However, due to the insufficient number of rendering resources, there may be a problem of allocating rendering resources for users that cannot meet the computational requirements. In a critical state that the rendering resources are close to the rendering requirements, a plurality of three-dimensional rendering instances of the cloud rendering server compete for limited computing resources, so that the picture rendering speed and quality cannot meet the expected requirements, and the system stability and the user experience are poor.
Disclosure of Invention
The invention provides a method, a device, equipment and a medium for controlling resource access of a cloud rendering server, which are used for solving the defect of poor picture rendering quality caused by rendering resource allocation in the prior art and realizing reasonable allocation of rendering resources, thereby improving the rendering quality and ensuring user experience.
The invention provides a cloud rendering server resource access control method, which comprises the following steps:
determining a current resource access request, and acquiring a resource occupation parameter according to the current resource access request, wherein the resource occupation parameter is used for representing the resource occupation amount of the current resource access request;
obtaining server parameters, wherein the server parameters are used for representing the running state of a cloud rendering server;
inputting the resource occupation parameters and the server parameters to a rendering quality prediction model to obtain a rendering quality prediction result output by the rendering quality prediction model;
and processing the current resource access request according to the rendering quality prediction result.
According to the cloud rendering server resource access control method provided by the invention, the server parameters comprise a first server parameter, a second server parameter and a third server parameter, wherein:
the first server parameter is used for representing the hardware running state of all rendering nodes in the server;
the second server parameter is used for representing the average rendering quality of the server;
the third server parameter is used to characterize the highest rendering quality of the server.
According to the cloud rendering server resource access control method provided by the invention, the server parameters further comprise fourth server parameters, and the fourth server parameters are used for representing the hardware running state of the rendering node with the lowest resource occupancy rate in the server.
According to the cloud rendering server resource access control method provided by the invention, the rendering quality prediction result comprises a first result and a second result; and processing the current resource access request according to the rendering quality prediction result, including:
if the rendering quality prediction result is a first result, suspending a current resource access request;
and if the rendering quality prediction result is a second result, adding the current resource access request into a ready queue.
According to the cloud rendering server resource access control method provided by the invention, before the step of inputting the resource occupation parameter and the server parameter into the rendering quality prediction model to obtain the rendering quality prediction result output by the rendering quality prediction model, the method further comprises a fixed threshold judgment step, specifically comprising:
acquiring the number of tasks in current operation and the number of tasks in a ready queue according to the server parameters;
acquiring the maximum task number which can be operated currently according to the resource occupation parameter;
and judging whether to continue to process the current resource access request according to the number of the tasks in the current operation, the number of the tasks in the ready queue and the maximum number of the tasks which can be operated at present.
According to the cloud rendering server resource access control method provided by the invention, the resource access request is determined specifically as follows:
obtaining a queuing queue, wherein the queuing queue is a user side resource access request queue based on time sequencing;
and determining the resource access request at the head of the queuing queue as the current resource access request.
According to the cloud rendering server resource access control method provided by the invention, the queue obtaining step comprises the following steps:
initializing a queuing queue;
establishing connection with a user terminal, acquiring a user terminal resource access request corresponding to the user terminal, and adding the user terminal resource access request into the tail of a queuing queue;
and continuously monitoring the connection, and if the connection is closed, removing the user side resource access request corresponding to the connection from the queuing queue.
The invention also provides a cloud rendering server resource access control device, which comprises:
the request determining module is used for determining a current resource access request, and acquiring a resource occupation parameter according to the current resource access request, wherein the resource occupation parameter is used for representing the resource occupation amount of the current resource access request;
the server parameter acquisition module is used for acquiring server parameters, wherein the server parameters are used for representing the running state of the cloud rendering server;
the rendering quality prediction module is used for inputting the resource occupation parameter and the server parameter into the rendering quality prediction model and obtaining a rendering quality prediction result output by the rendering quality prediction model;
and the request processing module is used for processing the current resource access request according to the rendering quality prediction result.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the cloud rendering server resource access control method according to any one of the above when executing the program.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a cloud rendering server resource access control method as described in any of the above.
The invention also provides a computer program product comprising a computer program which when executed by a processor implements a cloud rendering server resource access control method as described in any one of the above.
According to the cloud rendering server resource access control method, device, equipment and medium, the rendering quality prediction model is used, and the rendering quality prediction result output by the rendering quality prediction model is obtained according to the resource occupation parameter and the server parameter, so that the rendering resources can be distributed according to the rendering quality prediction result; compared with the traditional concurrent resource allocation method based on the fixed threshold, the method provided by the invention has the advantages that the rendering quality of the cloud rendering server in the critical state is obviously improved, and the system stability and the user experience are further improved.
Drawings
In order to more clearly illustrate the invention or the technical solutions 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 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 schematic flow chart of a cloud rendering server resource access control method provided by the invention;
fig. 2 is a schematic structural diagram of a resource access control device of a cloud rendering server provided by the invention;
fig. 3 is a schematic structural diagram of an electronic device provided by the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The cloud rendering server resource access control method, device, equipment and medium of the present invention are described below with reference to fig. 1 to 3.
Fig. 1 is a flow chart of a cloud rendering server resource access control method provided by the present invention, and as shown in fig. 1, the cloud rendering server resource access control method includes steps 110, 120, 140 and 150, specifically includes the following steps:
step 110, determining a current resource access request, and acquiring a resource occupation parameter according to the current resource access request, wherein the resource occupation parameter is used for representing the resource occupation amount of the current resource access request;
step 120, obtaining server parameters, wherein the server parameters are used for representing the running state of a cloud rendering server;
step 140, inputting the resource occupation parameter and the server parameter to a rendering quality prediction model, and obtaining a rendering quality prediction result output by the rendering quality prediction model;
and 150, processing the current resource access request according to the rendering quality prediction result.
In the embodiment of the invention, the resource occupation parameters can include computer performance parameters representing the resource occupation amount of the current resource access request, such as GPU occupation, CPU occupation, memory occupation, and the like, and the resource occupation parameters can be obtained directly according to the resource access request or can be obtained by calculating the power demand according to the rendering target of the resource access request, thereby obtaining the resource occupation parameters.
In the embodiment of the invention, the server parameters can include computer performance parameters used for representing the running state of hardware of the cloud rendering server, such as GPU occupancy rate, CPU occupancy rate, memory occupancy rate and the like of the server node, and also can include computer performance parameters used for representing the running state of tasks in the cloud rendering server, such as frame rate, delay and the like of the current running instance.
In the embodiment of the invention, the rendering quality prediction model adopts a supervised learning model based on an XGBoost algorithm. Other machine learning-based or statistical methods, such as LightGBM, may also be employed in embodiments of the present invention to achieve the goal of obtaining rendering quality predictions.
According to the cloud rendering server resource access control method provided by the embodiment of the invention, the rendering quality prediction result output by the rendering quality prediction model is obtained according to the resource occupation parameter and the server parameter by using the rendering quality prediction model, so that the rendering resources can be distributed according to the rendering quality prediction result; compared with the traditional concurrent resource allocation method based on the fixed threshold, the method provided by the invention has the advantages that the rendering quality is obviously improved, and the system stability and the user experience are further improved.
In the embodiment of the invention, the server parameters comprise a first server parameter, a second server parameter and a third server parameter, wherein:
the first server parameter is used for representing the hardware running state of all rendering nodes in the server;
the second server parameter is used for representing the average rendering quality of the server;
the third server parameter is used to characterize the highest rendering quality of the server.
Specifically, the first server parameter may include at least one of an average CPU occupancy rate of all rendering nodes in the server cluster, an average GPU occupancy rate of all rendering nodes in the server cluster, and an average memory occupancy rate of all rendering nodes in the server cluster;
the second server parameter may include at least one of an average render frame rate of all currently running instances within the server cluster, an average off-screen render delay of all currently running instances within the server cluster;
the third server parameter may include at least one of a highest render frame rate for all currently running instances within the server cluster, a lowest off-screen render delay for all currently running instances within the server cluster.
According to the embodiment of the invention, the total number of feature dimensions can be controlled and the model performance can be increased by setting the server parameters including the first server parameter, the second server parameter and the third server parameter; meanwhile, the first server parameter, the second server parameter and the third server parameter set by the embodiment of the invention can enable the rendering quality prediction model to obtain more correlations between computing resources and rendering quality, comprehensively consider the running condition and resource occupation of a macroscopic task, increase the generalization capability of the rendering quality prediction model in the rendering quality prediction task and improve the prediction accuracy of the model.
Further, the server parameters of the embodiment of the invention further comprise a fourth server parameter, wherein the fourth server parameter is used for representing the hardware running state of the rendering node with the lowest resource occupancy rate in the server;
specifically, the fourth server parameter may include one or more of GPU utilization of the rendering node with the lowest current resource occupancy rate, CPU utilization of the rendering node with the lowest current resource occupancy rate, and memory utilization of the rendering node with the lowest current resource occupancy rate.
The existing server resource allocation mode is mostly greedy rules, namely tasks are added to nodes with lower resource occupancy rate preferentially, so that the fourth server parameter is introduced into the rendering quality prediction model, the influence of the greedy rule-based server resource allocation mode on the rendering quality can be considered more, the rendering quality prediction result output according to the rendering quality prediction model can be matched with the result of processing the resource access request in the greedy rule-based server resource allocation mode better, and the picture rendering quality output after the resource access request is processed in the greedy rule-based server resource allocation mode is improved.
In the embodiment of the invention, the rendering quality prediction result comprises a first result and a second result; and processing the current resource access request according to the rendering quality prediction result, including:
if the rendering quality prediction result is a first result, suspending a current resource access request;
and if the rendering quality prediction result is a second result, adding the current resource access request into a ready queue.
The first result of the embodiment of the invention indicates that the predicted rendering quality is lower than the preset quality lower limit threshold, and the second result indicates that the predicted rendering quality is higher than the preset quality lower limit threshold, so that if the rendering quality predicted result is the first result, the current resource access request is suspended, and the rendering resource is temporarily not allowed to be allocated to the current resource access request; when rendering resources change, namely resources which are released exist, or when resource occupation parameters corresponding to a current resource access request change, or at intervals, the resource occupation parameters and server parameters can be input into a rendering quality prediction model again, a rendering quality prediction result output by the rendering quality prediction model is obtained, and whether the current resource access request is added into a ready queue is judged;
after the current resource access request is added into the ready queue, the server allocates rendering resources to perform rendering tasks according to the current resource access request.
The rendering quality prediction result of the embodiment of the invention is a classification result, the characteristics of the rendering quality prediction model can be fully exerted, the existing XGBoost model is more suitable for the requirements of real-time concurrent threshold reasoning and updating, and the method is suitable for the rendering quality prediction task, and better prediction speed and prediction effect are obtained.
In the embodiment of the present invention, before step 140, a fixed threshold determining step 130 is further included, specifically:
acquiring the number of tasks in current operation and the number of tasks in a ready queue according to the server parameters;
acquiring the maximum task number which can be operated currently according to the resource occupation parameter;
and judging whether to continue to process the current resource access request according to the number of the tasks in the current operation, the number of the tasks in the ready queue and the maximum number of the tasks which can be operated at present.
Judging whether to continue to process the current resource access request according to the number of tasks in current operation, the number of tasks in a ready queue and the maximum number of tasks which can be operated at present, wherein the expression is as follows:
Maximum_cnt–Running_cnt–Ready_cnt>0
wherein maximum_cnt is the Maximum number of tasks currently Running, running_cnt is the number of tasks currently Running, and ready_cnt is the number of tasks currently in a Ready queue; if the expression is satisfied, the current resource access request is continuously processed, and if the expression is not satisfied, the current resource access request is paused.
In the embodiment of the invention, through the fixed threshold judging step, whether the computing resource can meet the current resource access request can be quickly determined, and when the fixed threshold judging expression is not established, the computing resource cannot necessarily meet the current resource access request, so that rendering quality prediction is not needed, and the computing resource of a server is saved.
Further, the server parameters include a fifth server parameter including a maximum number of tasks currently operable, a number of tasks currently in operation, and a number of tasks currently in a ready queue. Under the condition of introducing a fixed threshold judgment step, the generalization capability of the model is improved according to the acquired information of the task quantity condition and the resource occupation quantity.
In the embodiment of the invention, the resource access request is determined by:
obtaining a queuing queue, wherein the queuing queue is a user side resource access request queue based on time sequencing;
and determining the resource access request at the head of the queuing queue as the current resource access request.
According to the cloud rendering server resource access control method, the queuing queues are introduced, the resource access requests of users are added into the queuing queues according to the time sequence, then the current resource access requests are processed according to the rendering quality prediction results, the conversion of the resource access requests in the queuing queues and the ready queues is controlled based on the rendering quality prediction results, the monopolization of rendering resources bound by tasks is guaranteed, the probability of resource allocation failure caused by task contention for the rendering resources is avoided, the resource allocation efficiency is improved, the rendering quality prediction results are acquired one by one when a plurality of users make requests, the calculation resources and the concurrency pressure of the rendering quality prediction model can be reduced, and the overall processing efficiency of multi-user requests is improved.
Further, in an embodiment of the present invention, the method for obtaining the queuing queue includes the following steps:
initializing a queuing queue;
establishing connection with a user terminal, acquiring a user terminal resource access request corresponding to the user terminal, and adding the user terminal resource access request into the tail of a queuing queue;
and continuously monitoring the connection, and if the connection is closed, removing the user side resource access request corresponding to the connection from the queuing queue.
Further, when the current resource access request is added into the ready queue, the connection with the user terminal is closed.
The queuing queue in the embodiment of the invention is a dynamic queuing queue, when a plurality of users request, the user side resource access request is added into the queuing queue according to the time sequence, and the queuing queue is dynamically updated through monitoring connection, so that the computing power resource can be ensured to be used for outputting a real rendering quality prediction result of the resource access request by a model, and the computing power waste is avoided.
In the embodiment of the invention, the rendering quality prediction model training is performed on real data accumulated in the real running process of the cloud rendering service, namely, whether to add a resource access request into a ready queue is judged only through steps 110-130 without introducing model judgment. The meta-universe application operated by the cloud rendering service is a meta-universe application developed based on Unreal Engine, rendering nodes are all NVIDIA TESLA A6000-model graphics cards, and the theoretical operation capacity of a single graphics card is 5 paths of concurrency of the single graphics card.
The training data intercepts real operation data within 4-7 days after the application is online, and records server parameters, average frame rate of a task after the user requests the task, and off-screen rendering delay. The supervision labels in the training process are whether the quality of service that results in the appearance of one or more application instances within the cloud rendering server when joining the task is below a lower rendering quality limit. Identifying as a first result when the quality of service appears to be below the lower rendering quality limit; when the quality of service of all instances is above the lower rendering quality limit, a second result is identified.
1023 pieces of data are accumulated in the running process of the system, and a training set and a testing set are divided according to the 8-2 principle, wherein 818 pieces of data are used for model training. Training was performed on NVIDIA TESLA A GPU 800 to obtain a render quality prediction model.
Table 1 shows the quality comparison result of the method of the embodiment of the present invention compared with the fixed threshold method of the prior art, and can find that compared with the conventional fixed threshold method, the server adopting the dynamic access control method of the resource has a significant improvement in service quality, reduces the average hardware overhead of the rendering node, and improves the running stability of the cloud rendering service; compared with the method of directly distributing tasks to the lowest current occupancy rate according to a single node operation threshold, the rendering resource dynamic distribution method can improve the maximum concurrency capacity of the system and maximize the utilization of server resources.
TABLE 1 method of embodiments of the invention and fixed threshold method render quality contrast results
In summary, according to the cloud rendering server resource access control method provided by the embodiment of the invention, by using the rendering quality prediction model, the rendering quality prediction result output by the rendering quality prediction model is obtained according to the resource occupation parameter and the server parameter, so that the rendering resources can be allocated according to the rendering quality prediction result; compared with the traditional concurrent resource allocation method based on the fixed threshold, the method provided by the invention has the advantages that the rendering quality of the cloud rendering server in the critical state is obviously improved, and the system stability and the user experience are further improved.
The cloud rendering server resource access control device provided by the invention is described below, and the cloud rendering server resource access control device described below and the cloud rendering server resource access control method described above can be referred to correspondingly.
Fig. 2 is a schematic structural diagram of a cloud rendering server resource access control device provided by the present invention, as shown in fig. 2, including:
a request determining module 210, configured to determine a current resource access request, and obtain a resource occupation parameter according to the current resource access request, where the resource occupation parameter is used to characterize a resource occupation amount of the current resource access request;
a server parameter obtaining module 220, configured to obtain a server parameter, where the server parameter is used to characterize an operating state of a cloud rendering server;
the rendering quality prediction module 240 is configured to input the resource occupation parameter and the server parameter to a rendering quality prediction model, and obtain a rendering quality prediction result output by the rendering quality prediction model;
and the request processing module 250 is used for processing the current resource access request according to the rendering quality prediction result.
According to the cloud rendering server resource access control device provided by the embodiment of the invention, the rendering quality prediction model can be used, and the rendering quality prediction result output by the rendering quality prediction model can be obtained according to the resource occupation parameter and the server parameter, so that the rendering resources can be distributed according to the rendering quality prediction result; compared with the traditional concurrent resource allocation method based on the fixed threshold, the method provided by the invention has the advantages that the rendering quality of the cloud rendering server in the critical state is obviously improved, and the system stability and the user experience are further improved.
Fig. 3 illustrates a physical schematic diagram of an electronic device, as shown in fig. 3, where the electronic device may include: processor 310, communication interface (Communications Interface) 320, memory 330 and communication bus 340, wherein processor 310, communication interface 320, memory 330 accomplish communication with each other through communication bus 340. Processor 310 may invoke logic instructions in memory 330 to perform a cloud rendering server resource access control method, comprising the steps of:
determining a current resource access request, and acquiring a resource occupation parameter according to the current resource access request, wherein the resource occupation parameter is used for representing the resource occupation amount of the current resource access request;
obtaining server parameters, wherein the server parameters are used for representing the running state of a cloud rendering server;
inputting the resource occupation parameters and the server parameters to a rendering quality prediction model to obtain a rendering quality prediction result output by the rendering quality prediction model;
and processing the current resource access request according to the rendering quality prediction result.
Further, the logic instructions in the memory 330 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product, where the computer program product includes a computer program, where the computer program may be stored on a non-transitory computer readable storage medium, and when the computer program is executed by a processor, the computer is capable of executing the cloud rendering server resource access control method provided by the above methods, and the method includes the following steps:
determining a current resource access request, and acquiring a resource occupation parameter according to the current resource access request, wherein the resource occupation parameter is used for representing the resource occupation amount of the current resource access request;
obtaining server parameters, wherein the server parameters are used for representing the running state of a cloud rendering server;
inputting the resource occupation parameters and the server parameters to a rendering quality prediction model to obtain a rendering quality prediction result output by the rendering quality prediction model;
and processing the current resource access request according to the rendering quality prediction result.
In still another aspect, the present invention further provides a non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor is implemented to perform the cloud rendering server resource access control method provided by the above methods, comprising the steps of:
determining a current resource access request, and acquiring a resource occupation parameter according to the current resource access request, wherein the resource occupation parameter is used for representing the resource occupation amount of the current resource access request;
obtaining server parameters, wherein the server parameters are used for representing the running state of a cloud rendering server;
inputting the resource occupation parameters and the server parameters to a rendering quality prediction model to obtain a rendering quality prediction result output by the rendering quality prediction model;
and processing the current resource access request according to the rendering quality prediction result.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. The cloud rendering server resource access control method is characterized by comprising the following steps of:
determining a current resource access request, and acquiring a resource occupation parameter according to the current resource access request, wherein the resource occupation parameter is used for representing the resource occupation amount of the current resource access request;
obtaining server parameters, wherein the server parameters are used for representing the running state of a cloud rendering server;
inputting the resource occupation parameters and the server parameters to a rendering quality prediction model to obtain a rendering quality prediction result output by the rendering quality prediction model;
and processing the current resource access request according to the rendering quality prediction result.
2. The cloud rendering server resource access control method of claim 1, wherein the server parameters comprise a first server parameter, a second server parameter, and a third server parameter, wherein:
the first server parameter is used for representing the hardware running state of all rendering nodes in the server;
the second server parameter is used for representing the average rendering quality of the server;
the third server parameter is used to characterize the highest rendering quality of the server.
3. The cloud rendering server resource access control method of claim 2, wherein the server parameters further comprise a fourth server parameter, the fourth server parameter being used to characterize a hardware running state of a rendering node with a lowest resource occupancy rate in a server.
4. The cloud rendering server resource access control method of any of claims 1-3, wherein the rendering quality prediction result comprises a first result and a second result; and processing the current resource access request according to the rendering quality prediction result, including:
if the rendering quality prediction result is a first result, suspending a current resource access request;
and if the rendering quality prediction result is a second result, adding the current resource access request into a ready queue.
5. The cloud rendering server resource access control method according to claim 1, wherein before the step of inputting the resource occupation parameter and the server parameter to the rendering quality prediction model to obtain the rendering quality prediction result output by the rendering quality prediction model, the method further comprises a fixed threshold judgment step, specifically:
acquiring the number of tasks in current operation and the number of tasks in a ready queue according to the server parameters;
acquiring the maximum task number which can be operated currently according to the resource occupation parameter;
and judging whether to continue to process the current resource access request according to the number of the tasks in the current operation, the number of the tasks in the ready queue and the maximum number of the tasks which can be operated at present.
6. The cloud rendering server resource access control method according to claim 1, wherein the determining the current resource access request specifically includes:
obtaining a queuing queue, wherein the queuing queue is a user side resource access request queue based on time sequencing;
and determining the resource access request at the head of the queuing queue as the current resource access request.
7. The cloud rendering server resource access control method of claim 6, wherein the acquiring the queuing queue comprises the steps of:
initializing a queuing queue;
establishing connection with a user terminal, acquiring a user terminal resource access request corresponding to the user terminal, and adding the user terminal resource access request into the tail of a queuing queue;
and continuously monitoring the connection, and if the connection is closed, removing the user side resource access request corresponding to the connection from the queuing queue.
8. A cloud rendering server resource access control apparatus, comprising:
the request determining module is used for determining a current resource access request, and acquiring a resource occupation parameter according to the current resource access request, wherein the resource occupation parameter is used for representing the resource occupation amount of the current resource access request;
the server parameter acquisition module is used for acquiring server parameters, wherein the server parameters are used for representing the running state of the cloud rendering server;
the rendering quality prediction module is used for inputting the resource occupation parameter and the server parameter into the rendering quality prediction model and obtaining a rendering quality prediction result output by the rendering quality prediction model;
and the request processing module is used for processing the current resource access request according to the rendering quality prediction result.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the cloud rendering server resource access control method of any of claims 1 to 7 when the program is executed by the processor.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the cloud rendering server resource access control method of any of claims 1 to 7.
CN202311619454.2A 2023-11-29 2023-11-29 Cloud rendering server resource access control method, device, equipment and medium Pending CN117851029A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311619454.2A CN117851029A (en) 2023-11-29 2023-11-29 Cloud rendering server resource access control method, device, equipment and medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311619454.2A CN117851029A (en) 2023-11-29 2023-11-29 Cloud rendering server resource access control method, device, equipment and medium

Publications (1)

Publication Number Publication Date
CN117851029A true CN117851029A (en) 2024-04-09

Family

ID=90527932

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311619454.2A Pending CN117851029A (en) 2023-11-29 2023-11-29 Cloud rendering server resource access control method, device, equipment and medium

Country Status (1)

Country Link
CN (1) CN117851029A (en)

Similar Documents

Publication Publication Date Title
US11436050B2 (en) Method, apparatus and computer program product for resource scheduling
CN113377540A (en) Cluster resource scheduling method and device, electronic equipment and storage medium
CN112181613B (en) Heterogeneous resource distributed computing platform batch task scheduling method and storage medium
CN112333234B (en) Distributed machine learning training method and device, electronic equipment and storage medium
CN110413776B (en) High-performance calculation method for LDA (text-based extension) of text topic model based on CPU-GPU (Central processing Unit-graphics processing Unit) collaborative parallel
CN111506434B (en) Task processing method and device and computer readable storage medium
CN112416585A (en) GPU resource management and intelligent scheduling method for deep learning
CN112035238A (en) Task scheduling processing method and device, cluster system and readable storage medium
CN112052950B (en) Neural network training method, model calculation server and storage medium
CN112422977A (en) Distribution method and distribution device for audio and video transcoding tasks
CN112559147B (en) Dynamic matching method, system and equipment based on GPU (graphics processing Unit) occupied resource characteristics
CN114118433A (en) Recommendation method and device for configuration parameters of equipment
CN115586961A (en) AI platform computing resource task scheduling method, device and medium
CN110780991B (en) Deep learning task scheduling method and device based on priority
CN116684420A (en) Cluster resource scheduling method, device, cluster system and readable storage medium
CN114995997A (en) Task processing method
CN111193802A (en) Dynamic resource allocation method, system, terminal and storage medium based on user group
CN109840597B (en) Model prediction method and device, electronic equipment and storage medium
CN112685162A (en) High-efficiency scheduling method, system and medium for heterogeneous computing resources of edge server
CN116996938A (en) Internet of vehicles task unloading method, terminal equipment and storage medium
CN117311973A (en) Computing device scheduling method and device, nonvolatile storage medium and electronic device
CN111930485A (en) Job scheduling method based on performance expression
CN117851029A (en) Cloud rendering server resource access control method, device, equipment and medium
CN116010051A (en) Federal learning multitasking scheduling method and device
CN114866430A (en) Calculation force prediction method for edge calculation, calculation force arrangement method and system

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