CN116974744B - Interactive control system and interactive control method for shared space - Google Patents
Interactive control system and interactive control method for shared space Download PDFInfo
- Publication number
- CN116974744B CN116974744B CN202310241663.1A CN202310241663A CN116974744B CN 116974744 B CN116974744 B CN 116974744B CN 202310241663 A CN202310241663 A CN 202310241663A CN 116974744 B CN116974744 B CN 116974744B
- Authority
- CN
- China
- Prior art keywords
- node
- application program
- mem
- performance mode
- nodes
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 51
- 230000002452 interceptive effect Effects 0.000 title claims abstract description 27
- 230000008569 process Effects 0.000 claims abstract description 26
- 101150053844 APP1 gene Proteins 0.000 claims abstract description 19
- 101100055496 Arabidopsis thaliana APP2 gene Proteins 0.000 claims abstract description 19
- 101100189105 Homo sapiens PABPC4 gene Proteins 0.000 claims abstract description 19
- 102100039424 Polyadenylate-binding protein 4 Human genes 0.000 claims abstract description 19
- 101100016250 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) GYL1 gene Proteins 0.000 claims abstract description 19
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000012800 visualization Methods 0.000 claims description 7
- 101100084404 Mus musculus Prodh gene Proteins 0.000 claims description 6
- 101150004094 PRO2 gene Proteins 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 6
- 101100194363 Schizosaccharomyces pombe (strain 972 / ATCC 24843) res2 gene Proteins 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- 238000011161 development Methods 0.000 abstract description 8
- 238000012360 testing method Methods 0.000 description 20
- 230000009467 reduction Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 102100038359 Xaa-Pro aminopeptidase 3 Human genes 0.000 description 4
- 101710081949 Xaa-Pro aminopeptidase 3 Proteins 0.000 description 4
- 238000000605 extraction Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013467 fragmentation Methods 0.000 description 1
- 238000006062 fragmentation reaction Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035807 sensation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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]
- G06F9/5061—Partitioning or combining of resources
-
- 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
- G06F9/44—Arrangements for executing specific programs
- G06F9/445—Program loading or initiating
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Stored Programmes (AREA)
Abstract
The invention provides an interactive control system and an interactive control method for a shared space, and belongs to the technical field of data identification and control. The method comprises the following steps of S100: acquiring K application program starting processes to be operated; s200: acquiring the current available resource space size of N nodes in a server cluster; s300: dividing K application programs { APP1, APP2, … … APPK } into M groups; s400: and selecting M nodes from the server cluster based on the current available resource space sizes of N nodes in the server cluster, distributing M groups of application programs to the M nodes for operation, and operating a group of application programs on each node. According to the technical scheme, the existing cluster node resources can be identified to be matched based on the starting data of the APP, so that more APP loading and running can be completed by utilizing the existing resources as much as possible, and the APP development efficiency is improved.
Description
Technical Field
The invention belongs to the technical field of data identification and control, and particularly relates to an interactive control system and an interactive control method for a shared space.
Background
A cluster refers to a group (several) of computers that are independent of each other, and utilizes a larger computer service system that is formed by a high-speed communication network, where each cluster node (i.e., each computer in the cluster) is an independent server running a respective service. These servers may communicate with each other, cooperatively provide applications, system resources, and data to users, and be managed in a single system mode. When a user requests a clustered system, the cluster gives the user the sensation of a single independent server, whereas in reality the user requests a group of clustered servers, i.e. a shared resource space.
Because of the sharing convenience of cluster resources, the method is widely applied in the development stage of application programs. A target application typically contains different load and run sub-phases that require the scheduling of cooperative or independent completion tests of different compute nodes in the cluster. In the prior art, the loading mode of the application program is fixed, namely loading and testing of all sub-phases are required to be completed at one time to obtain a test result.
When the cluster receives a larger number of application programs to be tested (to be run) at the same time, due to limited cluster resources, certain application programs may need to wait or delay testing, so that the development efficiency of the application programs is slowed down; meanwhile, an unreasonable cluster scheduling strategy and a fixed application loading mode can also cause that cluster resources cannot be fully utilized.
Disclosure of Invention
In order to solve the technical problems, the invention provides an interactive control system and an interactive control method for a shared space, which can fully identify the existing cluster node resources for matching based on the starting data of the APP and adaptively adjust the loading and testing modes of the APP, so that more APP sub-stage loading and running can be completed by utilizing the existing resources as much as possible, and the APP development efficiency is improved.
In a first aspect of the invention, an interactive control method for a shared space is proposed, the method being applied to a server cluster, the server cluster being in communication with a plurality of applications, each of the applications having access to a shared space corresponding to at least one node in the server cluster,
The method comprises the following steps:
S100: acquiring starting processes { Pro1, pro2, …, proK } of K application programs { APP1, APP2, … … APPK } to be operated;
at least K 1 of the K applications of the first type have a plurality of performance modes including a low performance mode, a standard performance mode, and a high performance mode, K 1 being an integer greater than 1.
The step S100 further includes: the multiple threads loaded by the start-up process Proi of the ith application APPi are pre-executed in standard performance mode to determine the running resource requirement size, resi, of the ith application.
S200: acquiring the current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster;
In the step S200, each packet includes at least one application of the first type.
S300: dividing K application programs { APP1, APP2, … … APPK } into M groups, wherein each group at least comprises two application programs;
S400: selecting M nodes from the server cluster based on the current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster, distributing the M groups of application programs to the M nodes for operation, and operating a group of application programs on each node; m < N;
the step S100 further includes: determining an operating resource requirement size Resi of the ith application program based on a start-up process Proi of the ith application program APPi; i=1, 2, … …, K;
The current available resource space size { Mem chose1,Memchose2,…,MemchoseM } of the M nodes { node chose1,nodechose2,…,nodechoseM } selected in step S400 satisfies the following condition:
Wherein, min { … } and max { … } respectively represent the minimum and maximum values of a plurality of values, 1< chose < N. The step S400 specifically includes:
The performance mode of each application is determined based on the current available resource space size Mem chose of the selected node chose and the operating resource requirement size of a set of applications running on node chose, at least one application being in a low performance mode.
In a second aspect of the invention, an interactive control system for a shared space is presented, the system comprising a server cluster and a plurality of terminal devices in communication with the server cluster.
Specifically, the system further comprises:
The application program starting detection unit is used for acquiring starting processes { Pro1, pro2, …, proK } of K application programs { APP1, APP2, … … APPK } to be operated; the executable mirror image packages of the K application programs are positioned in K terminal devices;
An application program pre-executing unit, configured to pre-execute a plurality of threads loaded by the start process Proi of the ith application program APPi, so as to determine a running resource requirement size Resi of the ith application program;
A node resource obtaining unit, configured to obtain a current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster;
A terminal grouping unit, configured to divide K terminal devices corresponding to K application programs { APP1, APP2, … … APPK } into M groups based on the running resource requirement sizes { Res1, res2, …, resK } of K application programs { APP1, APP2, … … APPK }, where each group includes at least two terminal devices;
A terminal access unit, configured to select M nodes from the server cluster based on a current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster, and access the M groups of terminals to the M nodes for operation, where each node operates a group of terminal devices;
Wherein 1< M < K; m, N, K are integers greater than 1.
And each group of terminal equipment accesses a shared space corresponding to at least one node in the server cluster.
At least K 1 of the K applications of the first type have a plurality of performance modes including a low performance mode, a standard performance mode, and a high performance mode; k 1 is an integer greater than 1.
The executable image package of each application program comprises a plurality of independent image layers, and the image layers are instantiated by providing container space through nodes in the server cluster so as to run each application program.
In a third aspect of the present invention, an electronic device is provided, the electronic device comprising a memory and a processor, the memory having stored thereon computer program instructions, the program instructions being executable by the processor for implementing all the steps of an interactive control method for a shared space according to the first aspect.
According to the technical scheme, the K application programs are divided into M groups by acquiring the starting processes of the K application programs to be operated and the current available resource space sizes of N nodes in the server cluster; selecting M nodes from the server cluster based on the current available resource space of N nodes in the server cluster, distributing the M groups of application programs to the M nodes for operation, wherein each node is operated with a group of application programs, and at least K 1 first-class application programs in the K application programs have multiple performance modes, wherein the multiple performance modes comprise a low-performance mode, a standard performance mode and a high-performance mode; the executable mirror package of each application program comprises a plurality of mirror layers which are independent of each other, the mirror layers are instantiated by providing a container space through nodes in the server cluster so as to run each application program, existing cluster node resources can be fully identified for matching based on starting data of the APP, loading and testing modes of the APP are adaptively adjusted, and therefore more APP sub-stage loading and running are completed by utilizing the existing resources as much as possible, and development efficiency of the APP is improved.
Further advantages of the invention will be further elaborated in the description section of the embodiments in connection with the drawings.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a main flow chart of an interactive control method for shared space according to one embodiment of the invention
FIG. 2 is a schematic diagram of an interactive control method for shared space to implement packet matching of applications and server nodes in accordance with the present invention
FIG. 3 is a schematic view of a scenario of an interactive control system for shared space according to one embodiment of the present invention
FIG. 4 is a schematic diagram of the functional unit composition of an interactive control system for shared space embodying one embodiment of the present invention
Detailed Description
The invention will be further described with reference to the drawings and detailed description.
Before describing various embodiments of the present application, application scenarios of the present application are first described, thereby leading to technical problems to be solved by the present application, so as to better understand improvements of technical solutions of the present application.
A target application typically contains different load and run sub-phases that require the scheduling of cooperative or independent completion tests of different compute nodes in the cluster. In the prior art, the loading mode of the application program is fixed, namely loading and testing of all sub-stages are required to be completed at one time to obtain a test result; furthermore, a plurality of applications to be tested are each located on a different terminal device. Because the resources provided by the terminal equipment are limited, an application developer loads a plurality of application programs to be tested on one or more nodes of the server cluster in a mode of leasing the server cluster resources, and each application program can access a shared space corresponding to at least one node in the server cluster, so that the whole testing process is completed.
However, when the cluster receives a larger number of applications to be tested (to be run) at the same time, due to limited cluster resources, it may cause some applications to need to wait or delay testing, thereby slowing down the development efficiency of the applications; meanwhile, an unreasonable cluster scheduling strategy and a fixed application loading mode can also cause that cluster resources cannot be fully utilized.
Accordingly, the present invention proposes one or more solutions to the above technical problems as follows.
Referring to fig. 1, fig. 1 is a main flow chart of an interactive control method for a shared space according to an embodiment of the present invention.
One interaction control method for a shared space shown in fig. 1 is applied to a server cluster, where the server cluster communicates with a plurality of application programs, and each of the application programs can access the shared space corresponding to at least one node in the server cluster.
As a specific embodiment, the shared space corresponding to the node may supply computing resources required for loading and running the application program, including memory resources, storage resources, graphics card resources, audio resources, and the like.
As a first advantage of the present invention, when developing the application program, the application program to be tested is assembled in the form of an executable image package, where each executable image package of the application program includes a plurality of independent image layers, and a container space is provided by a node in the server cluster to instantiate the plurality of independent image layers so as to load and run each application program.
It will be appreciated that the loading and running of the mirror layer is two distinct phases; when the available resources are sufficient, the loading and running phases can be continuously completed; when the available resources are insufficient, the mirror layer may only be loaded first and then run while waiting for the resources to be sufficient.
In various embodiments of the present invention, an executable image package includes a plurality of image layers that are independent of each other, the resource requirements of the different image layers being different; and, when testing the application program, all the mirror layers can be loaded, or all the mirror layers can be loaded, corresponding to different performance modes of the application program.
When all the mirror layers are loaded, the application program is in a standard performance mode; when all the mirror layers are loaded and run, the application program is in a high-performance mode; when the partial mirror layer is loaded, the application is placed in a low performance mode.
As a specific, but non-limiting example, assume that an application to be tested and developed is an image processing program, and the application includes several independent sub-stages requiring testing, such as image feature extraction, image feature preprocessing, image feature dimension reduction, image feature matrixing, and image feature visualization.
It will be appreciated that image feature extraction is a sub-stage of the necessary test, and that image feature preprocessing, image feature dimension reduction, image feature matrixing, image feature visualization are optional and independent of each other. For example, after the image features are extracted, the preprocessing may be performed and then the dimension reduction processing may be performed, or the dimension reduction processing may be directly performed; when the resources are sufficient, the image feature preprocessing and the image feature dimension reduction processing can be directly tested; when the resources are insufficient, one or more of image feature preprocessing, image feature dimension reduction, image feature matrixing and image feature visualization can be selectively executed.
The application program corresponding to the development to be tested is an image processing program, and comprises a plurality of mutually independent mirror image layers, namely an image feature extraction mirror image layer, an image feature preprocessing mirror image layer, an image feature dimension reduction mirror image layer, an image feature matrixing mirror image layer and an image feature visualization mirror image layer.
The resource requirements of the different mirror layers are different. For example, the image feature visualization mirror layer requires graphics card resources, while the image feature preprocessing mirror layer does not require graphics card resources; the image feature matrixing mirror layer may require GPU resources, while the image feature preprocessing mirror layer requires CPU resources, etc.
The plurality of mutually independent mirror layers are instantiated through the current available type container resource space provided by the nodes in the server cluster so as to load and run different sub-phases of each application program.
Specifically, steps S100 to S400 included in the method illustrated in fig. 1 are performed as follows:
S100: acquiring starting processes { Pro1, pro2, …, proK } of K application programs { APP1, APP2, … … APPK } to be operated;
S200: acquiring the current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster;
s300: dividing K application programs { APP1, APP2, … … APPK } into M groups, wherein each group at least comprises two application programs;
S400: selecting M nodes from the server cluster based on the current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster, distributing the M groups of application programs to the M nodes for operation, and operating a group of application programs on each node; m < N;
Wherein 1< M < K; m, N, K are integers greater than 1.
Specifically, the step S100 further includes: determining an operating resource requirement size Resi of the ith application program based on a start-up process Proi of the ith application program APPi; i=1, 2, … …, K;
The starting process Proi of the ith application program APPi at least comprises a starting thread, and the running resource requirement size Resi of the ith application program is determined by pre-executing the starting thread;
Specifically, the pre-executing includes analyzing the multiple threads loaded by the start process Proi of the ith application APPi, and determining a resource pointer of each thread, so as to determine the running resource requirement size Resi of the ith application.
The current available resource space size { Mem chose1,Memchose2,…,MemchoseM } of the M nodes { node chose1,nodechose2,…,nodechoseM } selected in step S400 satisfies the following condition:
wherein, min { … } and max { … } respectively represent the minimum and maximum values of a plurality of values, 1< chose < N.
It can be seen that the above limitation conditions ensure that the maximum available resources of the selected M nodes are fully utilized from the global aspect, and also ensure that the minimum resource requirements of each application program are matched from the individual local aspect, and simultaneously ensure that the concentration of the remaining resources of the cluster is higher, so as to avoid fragmentation of the cluster resources.
Preferably, at least K 1 of the K applications of the first class have a plurality of performance modes, including a low performance mode, a standard performance mode, and a high performance mode, K 1 being an integer greater than 1.
As described above, the executable mirror package includes a plurality of mirror layers that are independent of each other, and the resource requirements of different mirror layers are different; and, when testing the application program, all the mirror layers can be loaded, or all the mirror layers can be loaded, corresponding to different performance modes of the application program.
When all the mirror layers are loaded, the application program is in a standard performance mode; when all the mirror layers are loaded and run, the application program is in a high-performance mode; when the partial mirror layer is loaded, the application is placed in a low performance mode.
In the step S200, each packet includes at least one application of the first type.
Preferably, the step S100 further includes: the multiple threads loaded by the start-up process Proi of the ith application APPi are pre-executed in standard performance mode to determine the running resource requirement size, resi, of the ith application.
The step S400 specifically includes:
The performance mode of each application is determined based on the current available resource space size Mem chose of the selected node chose and the operating resource requirement size of a set of applications running on node chose, at least one application being in a low performance mode.
Fig. 2 is a schematic diagram of an interactive control method for shared space to implement packet matching between an application and a server node according to the present invention.
FIG. 2 shows five applications APP1-APP5 to be tested; assume that a server cluster contains 5 available computing nodes.
By executing the method shown in FIG. 1, APP1 and APP3 are distributed to node2 as a first group, APP2, APP4 and APP5 are distributed to node4 as a second group;
And APP3 is currently in low performance mode in the first group and APP4 is in low performance mode in the second group; simultaneously, APP1 in the first group is ensured to be in a high-performance mode, APP2 in the second group is ensured to be in a standard-performance mode, and APP5 is ensured to be in a high-performance mode.
As a more specific example, at this time, it is assumed that the graphics card resources on node2 are sufficient, but the memory resources are insufficient (most of the memory resources preferably satisfy the APP1 to perform the high performance mode), so APP3 may load only a portion of the visualized image layer (no other image layer is loaded), and perform the visualized process loading and displaying with limited memory resources to implement the visualized test in the low performance mode without waiting.
Therefore, at the moment, the APP1-APP5 can be simultaneously ensured to be simultaneously allocated to the resource to execute corresponding loading and testing; in the prior art, APP3, APP2 and APP4 can only wait or delay testing due to insufficient resources.
FIG. 3 is a schematic diagram of a scenario of an interactive control system for shared space according to one embodiment of the present invention.
In fig. 3, an interactive control system for a shared space is shown, the system comprising a server cluster and a plurality of terminal devices in communication with the server cluster, the plurality of terminal devices assembling an application to be tested in the form of an executable mirror package, each group of terminal devices accessing the shared space corresponding to at least one node in the server cluster for subsequent testing,
FIG. 4 is a schematic diagram of the functional unit components of an interactive control system for shared space implementing one embodiment of the present invention.
In fig. 4, the system comprises:
The application program starting detection unit is used for acquiring starting processes { Pro1, pro2, …, proK } of K application programs { APP1, APP2, … … APPK } to be operated; the executable mirror image packages of the K application programs are positioned in K terminal devices;
An application program pre-executing unit, configured to pre-execute a plurality of threads loaded by the start process Proi of the ith application program APPi, so as to determine a running resource requirement size Resi of the ith application program;
A node resource obtaining unit, configured to obtain a current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster;
A terminal grouping unit, configured to divide K terminal devices corresponding to K application programs { APP1, APP2, … … APPK } into M groups based on the running resource requirement sizes { Res1, res2, …, resK } of K application programs { APP1, APP2, … … APPK }, where each group includes at least two terminal devices;
A terminal access unit, configured to select M nodes from the server cluster based on a current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster, and access the M groups of terminals to the M nodes for operation, where each node operates a group of terminal devices; m < N;
Wherein 1< M < K; m, N, K are integers greater than 1.
At least K 1 of the K applications of the first type have a plurality of performance modes including a low performance mode, a standard performance mode, and a high performance mode; k 1 is an integer greater than 1.
The executable image package of each application program comprises a plurality of independent image layers, and the image layers are instantiated by providing container space through nodes in the server cluster so as to run each application program.
Specifically, the terminal access unit selects M nodes from the server cluster based on the current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster, accesses the M groups of terminals to the M nodes for operation, and operates a group of terminal devices on each node, which is specifically implemented as follows:
the current available resource space size { Mem chose1,Memchose2,…,MemchoseM } of the M nodes { node chose1,nodechose2,…,nodechoseM } selected by the terminal access unit satisfies the following condition:
wherein, min { … } and max { … } respectively represent the minimum and maximum values in a plurality of values, 1< chose < N; m < N;
The performance mode of each application is determined based on the current available resource space size Mem chose of the selected node chose and the operating resource requirement size of a set of applications running on node chose, at least one application being in a low performance mode.
It will be appreciated that each functional unit described in fig. 4 may correspondingly perform each step of the method described in fig. 1, and a detailed description thereof will not be repeated here.
For the prior art, the method divides K application programs into M groups by acquiring the starting processes of the K application programs to be operated and the current available resource space sizes of N nodes in the server cluster; selecting M nodes from the server cluster based on the current available resource space of N nodes in the server cluster, distributing the M groups of application programs to the M nodes for operation, wherein each node is operated with a group of application programs, and at least K 1 first-class application programs in the K application programs have multiple performance modes, wherein the multiple performance modes comprise a low-performance mode, a standard performance mode and a high-performance mode; the executable mirror package of each application program comprises a plurality of mirror layers which are independent of each other, the mirror layers are instantiated by providing a container space through nodes in the server cluster so as to run each application program, existing cluster node resources can be fully identified for matching based on starting data of the APP, loading and testing modes of the APP are adaptively adjusted, and therefore more APP sub-stage loading and running are completed by utilizing the existing resources as much as possible, and development efficiency of the APP is improved.
In the various embodiments of the present invention, the embodiments of the present invention have been shown and described, but it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principle and spirit of the present invention, the scope of which is defined by the appended claims and their equivalents.
Claims (5)
1. An interactive control method for a shared space, the method being applied to a server cluster, the server cluster being in communication with a plurality of applications, each of the applications having access to a shared space corresponding to at least one node in the server cluster, the method comprising the steps of:
S100: acquiring starting processes { Pro1, pro2, …, proK } of K application programs { APP1, APP2, … … APPK } to be operated;
the starting process Proi of the ith application program APPi at least comprises a starting thread, and the running resource requirement size Resi of the ith application program is determined by pre-executing the starting thread; the pre-execution comprises the steps of analyzing a plurality of threads loaded by a starting process Proi of an ith application program APPi, and determining a resource pointer of each thread so as to determine the running resource demand size Resi of the ith application program;
S200: acquiring the current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster;
s300: dividing K application programs { APP1, APP2, … … APPK } into M groups, wherein each group at least comprises two application programs;
S400: selecting M nodes from the server cluster based on the current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster, distributing the M groups of application programs to the M nodes for operation, and operating a group of application programs on each node;
wherein M is more than 1 and less than K; m, N, K are integers greater than 1;
the step S100 further includes: determining an operating resource requirement size Resi of the ith application program based on a start-up process Proi of the ith application program APPi; i=1, 2, … …, K;
The current available resource space size { Mem chose1,Memchose2,…,MemchoseM } of the M nodes { node chose1,nodechose2,…,nodechoseM } selected in step S400 satisfies the following condition:
Wherein, min { … } and max { … } respectively represent the minimum and maximum values in a plurality of values, and M is less than N;
the step S400 specifically includes: determining a performance mode for each application based on the current available resource space size Mem chose of the selected node chose and the operating resource requirement size of a set of applications running on node chose, at least one application being in a low performance mode; loading a corresponding mirror layer of the application program in a corresponding performance mode;
the loading and running of the mirror layer are two different phases; when the available resources are sufficient, the loading and running phases are continuously completed; when the available resources are insufficient, loading the mirror image layer, and executing an operation stage when waiting for the sufficient resources;
At least K 1 of the K applications of the first class have a plurality of performance modes including a low performance mode, a standard performance mode, and a high performance mode, K 1 being an integer greater than 1;
The executable mirror image packages of the K application programs are positioned in K terminal devices; the terminal equipment assembles the application program to be tested in the form of an executable mirror image package;
the executable mirror image comprises a plurality of mirror image layers which are mutually independent, and the resource requirements of different mirror image layers are different; the image feature visualization mirror layer needs display card resources; the image feature matrixing mirror layer requires GPU resources; CPU resources are needed for the image feature preprocessing mirror layer; when all the mirror layers are loaded, the application program is in a standard performance mode; when all the mirror layers are loaded and run, the application program is in a high-performance mode; when the partial mirror layer is loaded, the application is placed in a low performance mode.
2. The interactive control method according to claim 1, wherein in step S200, each group contains at least one application of the first type.
3. The interactive control method for a shared space according to claim 1, wherein said step S100 further comprises: the multiple threads loaded by the start-up process Proi of the ith application APPi are pre-executed in standard performance mode to determine the running resource requirement size, resi, of the ith application.
4. An interactive control system for a shared space, the system comprising a server cluster and a plurality of terminal devices in communication with the server cluster, the system further comprising:
The application program starting detection unit is used for acquiring starting processes { Pro1, pro2, …, proK } of K application programs { APP1, APP2, … … APPK } to be operated; the executable mirror image packages of the K application programs are positioned in K terminal devices; the starting process Proi of the ith application program APPi at least comprises a starting thread, and the running resource requirement size Resi of the ith application program is determined by pre-executing the starting thread; the pre-execution comprises the steps of analyzing a plurality of threads loaded by a starting process Proi of an ith application program APPi, and determining a resource pointer of each thread so as to determine the running resource demand size Resi of the ith application program;
An application program pre-executing unit, configured to pre-execute a plurality of threads loaded by the start process Proi of the ith application program APPi, so as to determine a running resource requirement size Resi of the ith application program; i=1, 2, … …, K;
A node resource obtaining unit, configured to obtain a current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster;
A terminal grouping unit, configured to divide K terminal devices corresponding to K application programs { APP1, APP2, … … APPK } into M groups based on the running resource requirement sizes { Res1, res2, …, resK } of K application programs { APP1, APP2, … … APPK }, where each group includes at least two terminal devices;
Wherein at least K 1 of the K applications of the first class have a plurality of performance modes including a low performance mode, a standard performance mode, and a high performance mode; k 1 is an integer greater than 1; the executable mirror package of each application program comprises a plurality of mirror layers which are independent from each other, and the mirror layers are instantiated by providing container space through nodes in the server cluster so as to run each application program; the terminal equipment assembles the application program to be tested in the form of an executable mirror image package;
a terminal access unit, configured to select M nodes from the server cluster based on a current available resource space size { Mem 1,Mem2,…,MemN } of N nodes { node 1,node2,…,nodeN } in the server cluster, where M is less than N; the M groups of terminals are accessed to the M nodes for operation, and a group of terminal equipment is operated on each node;
wherein M is more than 1 and less than K; m, N, K are integers greater than 1;
The current available resource space size { Mem chose1,Memchose2,…,MemchoseM } of the selected M nodes { node chose1,nodechose2,…,nodechoseM } satisfies the following condition:
Wherein, min { … } and max { … } respectively represent the minimum and maximum values in a plurality of values, and M is less than N;
Determining a performance mode for each application based on the current available resource space size Mem chose of the selected node chose and the operating resource requirement size of a set of applications running on node chose, at least one application being in a low performance mode;
loading a corresponding mirror layer of the application program in a corresponding performance mode;
the loading and running of the mirror layer are two different phases; when the available resources are sufficient, the loading and running phases are continuously completed; when the available resources are insufficient, loading the mirror image layer, and executing an operation stage when waiting for the sufficient resources;
the executable mirror image comprises a plurality of mirror image layers which are mutually independent, and the resource requirements of different mirror image layers are different; the image feature visualization mirror layer needs display card resources; the image feature matrixing mirror layer requires GPU resources; CPU resources are needed for the image feature preprocessing mirror layer; when all the mirror layers are loaded, the application program is in a standard performance mode; when all the mirror layers are loaded and run, the application program is in a high-performance mode; when the partial mirror layer is loaded, the application is placed in a low performance mode.
5. An interactive control system for a shared space as claimed in claim 4, wherein: and each group of terminal equipment accesses a shared space corresponding to at least one node in the server cluster.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310241663.1A CN116974744B (en) | 2023-03-14 | 2023-03-14 | Interactive control system and interactive control method for shared space |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310241663.1A CN116974744B (en) | 2023-03-14 | 2023-03-14 | Interactive control system and interactive control method for shared space |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116974744A CN116974744A (en) | 2023-10-31 |
| CN116974744B true CN116974744B (en) | 2024-05-17 |
Family
ID=88477298
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310241663.1A Active CN116974744B (en) | 2023-03-14 | 2023-03-14 | Interactive control system and interactive control method for shared space |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116974744B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105183561A (en) * | 2015-09-02 | 2015-12-23 | 浪潮(北京)电子信息产业有限公司 | Resource distribution method and resource distribution system |
| CN105786619A (en) * | 2016-02-24 | 2016-07-20 | 中国联合网络通信集团有限公司 | Virtual machine distribution method and device |
| CN110308993A (en) * | 2019-06-27 | 2019-10-08 | 大连理工大学 | A kind of cloud computing resources distribution method based on improved adaptive GA-IAGA |
| CN112346861A (en) * | 2020-10-27 | 2021-02-09 | 北京天融信网络安全技术有限公司 | Resource allocation method, device, equipment and super-fusion system |
| WO2022105589A1 (en) * | 2020-11-20 | 2022-05-27 | 上海连尚网络科技有限公司 | Resource scheduling method and apparatus, electronic device and computer readable medium |
| CN114706596A (en) * | 2022-04-11 | 2022-07-05 | 中国电信股份有限公司 | Container deployment method, resource scheduling method, device, medium and electronic equipment |
| CN115599513A (en) * | 2022-04-07 | 2023-01-13 | 荣耀终端有限公司（Cn） | Resource scheduling method and electronic equipment |
-
2023
- 2023-03-14 CN CN202310241663.1A patent/CN116974744B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105183561A (en) * | 2015-09-02 | 2015-12-23 | 浪潮(北京)电子信息产业有限公司 | Resource distribution method and resource distribution system |
| CN105786619A (en) * | 2016-02-24 | 2016-07-20 | 中国联合网络通信集团有限公司 | Virtual machine distribution method and device |
| CN110308993A (en) * | 2019-06-27 | 2019-10-08 | 大连理工大学 | A kind of cloud computing resources distribution method based on improved adaptive GA-IAGA |
| CN112346861A (en) * | 2020-10-27 | 2021-02-09 | 北京天融信网络安全技术有限公司 | Resource allocation method, device, equipment and super-fusion system |
| WO2022105589A1 (en) * | 2020-11-20 | 2022-05-27 | 上海连尚网络科技有限公司 | Resource scheduling method and apparatus, electronic device and computer readable medium |
| CN115599513A (en) * | 2022-04-07 | 2023-01-13 | 荣耀终端有限公司（Cn） | Resource scheduling method and electronic equipment |
| CN114706596A (en) * | 2022-04-11 | 2022-07-05 | 中国电信股份有限公司 | Container deployment method, resource scheduling method, device, medium and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116974744A (en) | 2023-10-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10949237B2 (en) | Operating system customization in an on-demand network code execution system | |
| US11113782B2 (en) | Dynamic kernel slicing for VGPU sharing in serverless computing systems | |
| CN111966500B (en) | Resource scheduling method and device, electronic equipment and storage medium | |
| US8930731B2 (en) | Reducing power consumption in data centers having nodes for hosting virtual machines | |
| US9052932B2 (en) | Hybrid virtual machine configuration management | |
| US7945913B2 (en) | Method, system and computer program product for optimizing allocation of resources on partitions of a data processing system | |
| CN110098946B (en) | Method and device for deploying virtualized network element equipment | |
| US20100115510A1 (en) | Virtual graphics device and methods thereof | |
| US20180331927A1 (en) | Resource Coordinate System for Data Centers | |
| CN112767135B (en) | Configuration method and device of rule engine, storage medium and computer equipment | |
| CN112764938B (en) | Cloud server resource management method, cloud server resource management device, computer equipment and storage medium | |
| CN109739614A (en) | Virtual machine rebuilding method, device and equipment | |
| CN116541134A (en) | Method and device for deploying containers in multi-architecture cluster | |
| CN111475256A (en) | Resource allocation method, device and system | |
| CN112905317A (en) | Task scheduling method and system under rapid reconfigurable signal processing heterogeneous platform | |
| US7627655B2 (en) | Increased performance of grid applications | |
| CN108829516B (en) | Resource virtualization scheduling method for graphic processor | |
| EP1949228B1 (en) | Asynchronous just-in-time compilation | |
| CN116974744B (en) | Interactive control system and interactive control method for shared space | |
| CN116578416B (en) | Signal-level simulation acceleration method based on GPU virtualization | |
| CN109614242B (en) | Computing capacity sharing method, device, equipment and medium | |
| CN111400051A (en) | Resource scheduling method, device and system | |
| CN111381831B (en) | Application deployment method and server | |
| CN116680078A (en) | Cloud computing resource scheduling method, device, equipment and computer storage medium | |
| US11954534B2 (en) | Scheduling in a container orchestration system utilizing hardware topology hints |
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 | ||
| GR01 | Patent grant |