CN112817691B - Resource allocation method, device, equipment and medium - Google Patents
Resource allocation method, device, equipment and medium Download PDFInfo
- Publication number
- CN112817691B CN112817691B CN202110114221.1A CN202110114221A CN112817691B CN 112817691 B CN112817691 B CN 112817691B CN 202110114221 A CN202110114221 A CN 202110114221A CN 112817691 B CN112817691 B CN 112817691B
- Authority
- CN
- China
- Prior art keywords
- code
- container
- target
- resource
- execution
- 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
- 238000013468 resource allocation Methods 0.000 title claims abstract description 84
- 238000000034 method Methods 0.000 title claims abstract description 73
- 238000007726 management method Methods 0.000 claims abstract description 55
- 238000004590 computer program Methods 0.000 claims description 13
- 238000010586 diagram Methods 0.000 description 19
- 230000008569 process Effects 0.000 description 11
- 230000006870 function Effects 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000010276 construction Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 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
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
-
- 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
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/4557—Distribution of virtual machine instances; Migration and load balancing
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Stored Programmes (AREA)
Abstract
The embodiment of the invention provides a resource allocation method, a device, equipment and a medium, belongs to the technical field of computers, and aims to improve the rationality of resource allocation, wherein the method is applied to first equipment and comprises the following steps: creating a code writing container by utilizing local resources, and writing target codes through the code writing container; the first equipment utilizes local resources to create a proxy container, and receives a resource allocation request sent by a code writing container through the proxy container; applying for the target resource from the resource management server through the proxy container; the first device sends a code execution request to a first code execution container through a proxy container, the code execution request being for instructing the execution of target code through the first code execution container, the first code execution container being created by a second device providing the target resource using the target resource.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a method, an apparatus, a device, and a medium for resource allocation.
Background
With the development of computer network technology, corresponding operation resources are actually required to be provided for development of some application software and the like. Taking neural network construction as an example, to support neural network construction, a cluster of devices may be utilized to provide hardware resources, such as graphics card resources, for neural network construction. In general, these clusters of devices may be collectively referred to as cloud resources.
In the related art, part of resources are divided from cloud resources to users in a device cluster according to a certain allocation rule, so that the users can utilize the resources divided to the devices logged in by the users to perform related software development. For example, the resources are classified according to the user class, and the users with high class are classified as more resources, and the users with low class are classified as less resources. However, when such a resource division method is adopted, there are inevitably cases of more or less resources, and when the resources are more, there are wasted resources and occupied space, and when the resources are less, there are cases that the required execution environment cannot be supported.
Therefore, the resource allocation method in the related technology has the technical problems of unreasonable resource allocation and low resource utilization rate.
Disclosure of Invention
In view of the foregoing, embodiments of the present invention provide a method, apparatus, device, and medium for resource allocation, so as to overcome or at least partially solve the foregoing problems.
To solve the above-mentioned problems, in a first aspect of the present invention, there is disclosed a resource allocation method, which is applied to a first device, the method comprising:
Creating a code writing container by utilizing local resources, and writing target codes through the code writing container;
creating a proxy container by utilizing local resources, and receiving a resource allocation request sent by the code writing container through the proxy container, wherein the resource allocation request is used for requesting target resources required by executing the target code;
applying for the target resource to a resource management server through the proxy container;
the first device sends a code execution request to the first code execution container through the proxy container, the code execution request being for instructing execution of the target code by the first code execution container created by a second device providing the target resource using the target resource.
Optionally, the method further comprises:
creating a second code execution container using the local resource;
sending a code execution request to the second code execution container through the code writing container, the code execution request being for requesting execution of the object code;
and executing the target code through the second code execution container.
Optionally, the method further comprises:
Receiving a code execution result generated after the target code is executed by the first code execution container through the proxy container;
and outputting the code execution result through a local information output component.
Optionally, the method further comprises:
mounting the target code to a storage disk through the code writing container, wherein the storage disk is shared by the code writing container and a code executing container, and the code executing container is used for executing the target code;
the first device carries a storage address of the target code in the storage disk through a code execution request sent by the code writing container, so that the code execution container reads the target code from the storage disk according to the storage address, and further executes the target code.
Optionally, after writing the object code by the code writing container, the method further comprises:
obtaining a code execution mode configured by a user through the code writing container;
creating a proxy container using local resources, comprising:
when the code execution mode is remote execution, a local resource is utilized to create a proxy container, so that the target resource is applied to a resource management server through the proxy container, the first code execution container is obtained based on the target resource, and the target code is executed through the first code execution container;
Creating a second code execution container using the local resource, comprising:
and under the condition that the code execution mode is local execution, creating a second code execution container by utilizing local resources, and executing the target code through the second code execution container.
In a second aspect of the embodiments of the present application, there is provided a resource allocation method, where the method is applied to a second device, and the method includes:
receiving a resource scheduling instruction sent by a resource management server, wherein the resource scheduling instruction is generated by the resource management server when a resource allocation request sent by first equipment through a proxy container is received, and the resource allocation request is generated by a code writing container on the first equipment after the completion of target code writing is detected; wherein the proxy container and the code writing container are both created by the first device using local resources;
according to the resource scheduling instruction, a first code execution container is created by utilizing a local target resource;
and executing the target code through the first code execution container when receiving a code execution request sent by the first device through the proxy container.
Optionally, executing the object code by the first code execution container includes:
acquiring the target code from a storage disk mounted by the code writing container; wherein the storage disk is shared by the code writing container and the first code executing container;
executing the target code through the first code execution container.
Optionally, the method further comprises:
acquiring a code execution result generated after the target code is executed by the first code execution container;
and sending the code execution result to the proxy container so that the first device outputs the code execution result through an information output component local to the first device.
In a third aspect of the embodiments of the present application, there is provided a resource allocation apparatus, where the apparatus is deployed in a first device, including:
the creation module is used for writing a container and a proxy container by utilizing codes created by local resources; the code writing container is used for writing target codes; the agent container is used for receiving a resource allocation request sent by the code writing container, and the resource allocation request is used for requesting target resources required by executing the target code;
The resource application module is used for applying the target resource to a resource management server through the proxy container;
and the code execution module is used for sending a code execution request to the first code execution container through the proxy container, wherein the code execution request is used for instructing a second device to execute the target code through the first code execution container, and the first code execution container is created by the second device for providing the target resource by utilizing the target resource.
In a fourth aspect of the embodiments of the present application, there is provided a resource allocation apparatus, where the apparatus is deployed in a second device, including:
the resource scheduling receiving module is used for receiving a resource scheduling instruction sent by a resource management server, wherein the resource scheduling instruction is generated by the resource management server when a resource allocation request sent by first equipment through a proxy container is received, and the resource allocation request is generated by a code writing container on the first equipment after the completion of target code writing is detected; wherein the proxy container and the code writing container are both created by the first device using local resources;
the container creation module is used for creating a first code execution container by utilizing a local target resource according to the resource scheduling instruction;
And the code execution module is used for executing the target code through the first code execution container when receiving the code execution request sent by the first equipment through the proxy container.
In a fifth aspect of the embodiments of the present application, a resource allocation system is provided, where the resource allocation system is applied to a device cluster, where the device cluster includes at least a first device and a second device, and the system includes:
the first device includes: code authoring containers and proxy containers created using local resources; the second device is configured to provide a target resource required for executing the target code, including: executing a container by using the first code created by the target resource;
wherein the code writing container is for: the user writes the target code, receives the resource allocation request sent by the code writing container, and applies for the target resource required by executing the target code to the resource management server;
the agent container is used for: sending a code execution request to the first code execution container, the code execution request being for requesting execution of the object code;
the first code execution container is for: executing the target code.
Optionally, the first device further comprises: creating a second code execution container using the local resource;
The code writing container is used for sending a code execution request to the second code execution container, and the code execution request is used for requesting to execute the target code;
the second code execution container is for executing the object code.
Optionally, the first device further comprises: an information output component;
the proxy container is used for receiving a code execution result generated after the first code execution container executes the target code;
the information output component is used for outputting the code execution result.
Optionally, the system further comprises: a storage disk shared by the code writing container and a code executing container for executing the target code;
the code writing container is used for mounting the target code into the storage disk;
the first device carries a storage address of the target code in the storage disk through a code execution request sent by the code writing container, so that the code execution container reads the target code from the storage disk according to the storage address, and further executes the target code.
The embodiment of the invention also discloses an electronic device, which comprises: comprising a memory, a processor and a computer program stored on the memory and executable on the processor, which processor when executed implements a resource allocation method as described in the embodiments of the first or second aspect.
The embodiment of the invention also discloses a computer readable storage medium, and a computer program stored in the computer readable storage medium causes a processor to execute the resource allocation method according to the embodiment of the first aspect or the embodiment of the second aspect of the invention.
The embodiment of the invention has the following advantages:
in an embodiment of the present invention, a first device in a device cluster may perform the following: and creating a code writing container by utilizing a local resource, writing target codes by utilizing the code writing container, creating a proxy container by utilizing the local resource, receiving a resource allocation request sent by the code writing container by utilizing the proxy container, and then applying the target resources to a resource management server by utilizing the proxy container. In this manner, a second device providing the target resource may create a first code execution container using the target resource; then, the first device sends a code execution request to the first code execution container through the proxy container to cause the second device to execute the object code through the first code execution container.
Because in the embodiment of the application, the first device can create a code writing container by using a local resource, write the target code through the code writing container, and apply for the target resource required by the target code to the resource management server through the proxy container when the target code needs to be executed, and then create the first code executing container through the second device where the target resource is located to execute the target code. In this manner, the resources required to write the code may originate from local resources, and the resource management server may be dynamically applied for the target resources required to execute the code as the code is executed. In this way, on the one hand, the target resources required for executing the code can be not fixed to the user, so that the target resources can be allocated to other users for use before being allocated to the user, and the problem that the resources are occupied when the resources in the code editing stage and the code executing stage are both fixed resources corresponding to the user is avoided. On the other hand, the target resource required by the execution code is dynamically applied to the resource management server when the code is executed, so that the target resource is the resource required by the execution code in an adaptive manner, the problem of insufficient or excessive execution resource caused by the execution resource of the code appointed in advance is avoided, and the resource utilization rate is fully improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the description of the embodiments of the present invention will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of an implementation environment of a resource allocation method according to an embodiment of the present invention;
FIG. 2 is a flow chart of steps of a method for resource allocation in an embodiment of the present invention;
FIG. 3 is a general flow diagram of a resource allocation method shown from a first device side in the implementation of the present invention;
FIG. 4 is a flow chart of steps of yet another resource allocation method in an embodiment of the present invention;
FIG. 5 is a block diagram of a resource allocation system in the practice of the present invention;
FIG. 6 is a block diagram of yet another resource allocation system in the practice of the present invention;
FIG. 7 is a block diagram of a resource allocation device in the practice of the present invention;
fig. 8 is a block diagram of a further resource allocation device in the implementation of the present invention.
Detailed Description
In order that the above objects, features and advantages of the present invention will be readily apparent, a more particular description of embodiments of the invention will be rendered by reference to the appended drawings, which are illustrated in the appended drawings. 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 resource allocation method in the related art is as follows: for a user applying for resources in the device cluster by starting up, fixed resources are divided from the device cluster to be dedicated to the user for developing and constructing related application software. The neural network construction or software development generally comprises two stages, namely a code editing stage and a code executing stage, and when a resource allocation mode in the related technology is adopted, the resources in the code editing stage and the code executing stage are both fixedly corresponding to users.
However, in this way, the resources required by the user to perform the code when the user writes the code, which is not yet used, need to be prepared for the running code, so that the resources waiting for the running code are occupied, and the utilization rate of the resources in the device cluster is reduced.
In view of this, the applicant has proposed the following technical idea: and allocating the needed resources for the code writing from the local resources, and dynamically applying for the target resources needed by the code executing to the resource management server when the code writing is completed and the code executing is needed, so that the target resources needed by the code executing are not needed to be fixed to a user, and the target resources can be allocated to other users for use before being allocated to the user, thereby avoiding the problem that the target resources are occupied in a code editing stage and a code executing stage.
Referring to fig. 1, a schematic diagram of an implementation environment of an embodiment of the present application is shown, where, as shown in fig. 1, a device cluster may be included, where the device cluster includes a resource management server and a plurality of devices. The plurality of devices may refer to intelligent devices with graphics card resources, and the plurality of devices together form a device cluster, where the device cluster may be referred to as a cloud resource, and the resource management server may be configured to allocate resources in the cloud resource.
In this embodiment, when a user may write and run a code through one device in the device cluster, the device actually used by the user may be referred to as a first device, for example, as in fig. 1, when the user uses the device 4, the device 4 may be referred to as a first device. When the code is written and executed, the resources on other devices in the device cluster can also be used for supporting the execution of the code written by the user on the first device, and the other devices can be second devices, for example, the resources of the device 1 are used, and the device 1 is the second device. In this way, the problem of allocating resources in the device cluster to users is related, and in order to improve the resource utilization rate of the device cluster, the embodiment of the application provides a resource allocation method.
Referring to fig. 2, a schematic step flow diagram of a resource allocation method of the present application, which is specifically applied to a first device, may include the following steps:
step S201: creating a code writing container by utilizing local resources, and writing target codes through the code writing container.
In this embodiment, the first device executing the resource allocation method may refer to a device currently used by a user, and in practice, the first device may be a device having a Virtual Machine (Virtual Machine), where the Virtual Machine refers to a complete computer system that is simulated by software and has a complete hardware system function and operates in a completely isolated environment. In this way, when the user uses the first device for code encoding and code execution, this may be done in the virtual machine of the first device.
In particular implementations, the first device may provide an IDE (Integrated Development Environment ) portal for a user to access the virtual machine, where the IDE portal may be JupyterLab. Thus, when the user uses the virtual machine of the first device for code writing, it can be realized by JupyterLab.
When the user uses the virtual machine of the first device to write the code, the implementation mode of the virtual machine is a container, and in this embodiment, the code writing can be implemented through the code writing container, where the resources required by the container can be derived from the first device, that is, the code writing container is created by using the resources local to the first device. The resources local to the first device may be CPU resources or graphics card resources, and specifically, which resource is specifically applicable may be determined according to the service corresponding to the code writing. For example, when the resources required for the service are more, the local graphics card resources may be allocated to the user as the resources of the virtual machine, or when the resources required for the service are less, the local CPU resources may be allocated to the user as the resources of the virtual machine.
Wherein after creating a code writing container using a local resource, the code writing container can be used to write the object code. In particular, a code writing container may be understood as a computer environment in which a user writes object code, and in the environment provided by the code writing container, the user may be understood as writing object code in a view interface.
Step S202: and creating a proxy container by utilizing local resources, and receiving a resource allocation request sent by the code writing container through the proxy container, wherein the resource allocation request is used for requesting target resources required by executing the target code.
In this embodiment, after the target code is written, the target code may be executed. Wherein the resources required to execute the object code may be dynamically acquired by the first device, in particular the first device may create a proxy container using the local resources, the proxy container communicating with the code writing container to receive the resource allocation request sent by the code writing container. That is, the code composition container may send a resource allocation request to the proxy container upon detecting that the user has completed composition of the target code, thereby requesting the target resources required to execute the target code.
The target resource may be a resource derived from the local of the first device, or may be a resource derived from other devices in the device cluster, and may specifically be determined according to the size of the resource required for executing the target code.
Step S203: and applying the target resource to a resource management server through the proxy container.
In this embodiment, a resource management server may be located in the device cluster, where the resource management server may be configured to schedule resources of each device in the device cluster. In this manner, the proxy container may apply for the target resource from the resource management server, and in particular, the proxy container may send a resource request to the resource management server, which may carry the size of the required resource or the configuration of the required resource. In this manner, the resource management server may schedule to the target resource from the cluster of devices according to the size or configuration of the target resource.
The target resource may be a graphics card resource or a memory resource.
Step S204: and sending, by the proxy container, a code execution request to the first code execution container, the code execution request being for instructing a second device to execute the target code by the first code execution container, the first code execution container being created by the second device providing the target resource using the target resource.
In this embodiment, when the target resource is a resource on the second device, the scheduling of the target resource by the resource management server may be the following manner: and sending a resource scheduling request to the second device so that the second device creates a first code execution container by using the target resource, wherein the target code can be carried in the resource scheduling request.
Wherein the second device is a different device than the first device, the second device creating the first code execution container using the target resource may be understood as: the second device creates a first code execution container for executing the code in the software and hardware environment provided by the target resource, for example, the first code execution container is set in the graphics card resource of the second device, so that the code is executed by using the graphics card resource, that is, the target code is executed in the graphics card resource, and the first code execution container can be understood as a process environment for executing the code.
In this embodiment, when the resource management server dispatches to the target resource and the second device creates the first code execution container, the resource management server may return the details of the target resource to the proxy container, so that the proxy container may send a code execution request to the first code execution container according to the details of the target resource, and the first code execution container may execute the target code carried in the resource dispatch request.
By adopting the technical scheme of the embodiment of the application, the first equipment can utilize the local resource to create a code writing container, write the target code through the code writing container, dynamically apply for the target resource required by the target code to the resource management server through the proxy container when the target code needs to be executed, and further create the first code executing container through the second equipment where the target resource is located so as to execute the target code. In this manner, the resources required to write the code may originate from local resources, and the resource management server may be dynamically applied for the target resources required to execute the code as the code is executed. In this way, on the one hand, the target resources required for executing the code can be not fixed to the user, so that the target resources can be allocated to other users for use before being allocated to the user, and the problem that the resources are occupied when the resources in the code editing stage and the code executing stage are both resources fixed to the user is avoided. On the other hand, the target resource required by the execution code is dynamically applied to the resource management server when the code is executed, so that the target resource is the resource required by the execution code in an adaptive manner, the problem of insufficient or excessive execution resource caused by the execution resource of the code appointed in advance is avoided, and the resource utilization rate is fully improved.
In one embodiment, referring to FIG. 3, a schematic diagram of a process for encoding and executing code from the perspective of a first device is shown, where the first device may store written code in a mounted memory disk when writing the code, and may read target code from the memory disk to execute the target code when executing the target code, as shown in FIG. 3. The target code can be executed by using a second code execution container created by the local resource of the first device, or by using a first code execution container created by the resource of the second device. Wherein the execution result of the object code can be output through the information output component,
in the following, in connection with fig. 3, a detailed description is first given of how the first device of the present application supports code writing:
specifically, in one example, a first device may mount object code into a storage disk through a code writing container, which may be shared by the code writing container and a code execution container for executing the object code;
the first device carries a storage address of the target code in the storage disk through a code execution request sent by the code writing container, so that the code execution container reads the target code from the storage disk according to the storage address, and further executes the target code.
In this embodiment, the storage disk may be mounted on the code writing container, where the storage disk mounted on the code writing container may include a data disk and a rootfs disk, or may be only one of the rootfs disk and the data disk, and may specifically be determined according to code writing requirements.
The data disk and the rootfs disk are both in network storage form, so that the same storage disk can be shared by multiple containers, in this embodiment, the storage disk can be shared by a code writing container and a code executing container, where the code executing container specifically refers to a container of a user executing an object code, and as shown in fig. 3, the code executing container specifically may be a first code executing container or a second code executing container local to the first device. Wherein the creation process and conditions of the second code execution container local to the first device are set forth in the subsequent embodiments.
Wherein the storage disk for the code writing container and the code executing container to share may refer to: the code writing container may store the written code to a storage disk, and the code execution container may read the written code from the storage disk.
Accordingly, in the target code executing stage, the code executing request sent by the second device through the proxy container may include the storage address of the target code in the storage disc, so that the code executing container may read the target code from the storage disc according to the storage address, and execute the target code.
When the implementation mode is adopted, the storage disk can be shared by the code writing container and the code executing container, so that the consistency of environment data can be ensured, and the data in the storage disk mounted by the code writing container and the code executing container are kept consistent.
Next, a detailed description is given of how the first device of the embodiment of the present application executes the code:
in one example, the target code may also be executed with the resources of the first device when less resources are required to execute the code. Specifically, the first device may create a second code execution container using a local resource, and send a code execution request to the second code execution container through the code writing container, the code execution request being for requesting execution of the target code; in this manner, the first device may execute the object code through the second code execution container.
The resources local to the first device may refer to resources of a virtual machine of the first device, which may be CPU resources or graphics card resources, and if the first device uses the local resources to create the second code execution container, the first device may not apply for the target resources to the resource management server, so that the second code execution container is a container running locally. In this case, the first device may send a code execution request to the second code execution container through the code writing container, where the code execution request may be used to request the second code execution container to execute the target code.
In this manner, the object code may also be executed using local resources.
In yet another example, as shown in fig. 3, after the first device writes the target code through the code writing container, the first device may obtain the code execution mode configured by the user through the code writing container, so that it may select whether to execute the target code by using the local resource or using the remote resource according to the code execution mode configured by the user, so as to implement flexible configuration of the target resource.
In a specific implementation manner, the first device may provide an optional code execution manner to the user through the IDE interface, and further, the first device may apply for the target resource to the resource management server or create the second code execution container by using the local resource according to the code execution manner selected by the user.
For example, the following 3 code implementations are provided: different resource allocation modes can be applied to different resources, for example, a user selects a code execution mode of the display card x1, the first device can create a second code execution container by using local resources, and if the user selects the code execution modes of the display card x2 and the display card x4, the first device can apply for target resources to a resource management server by using a proxy container, namely, request for the target resources to other devices in the device cluster.
When the code executing mode is remote execution, the first device may create a proxy container by using a local resource, apply the target resource to the resource management server through the proxy container, obtain the first code executing container based on the target resource, and execute the target code through the first code executing container.
In the case that the code execution mode is local execution, the first device may create a second code execution container by using a local resource, and execute the target code through the second code execution container.
The process of creating the second code execution container by using the local resource and executing the object code by the second code execution container may be described in the above embodiment, and will not be described herein. Accordingly, the process of executing the target code by the first code execution container may refer to the above-mentioned steps S203-S206 by applying the proxy container for the target resource to the resource management server, which is not described herein.
With this embodiment, the proxy container of the first device corresponds to a proxy service of the first device applying for the remote target resource, and is used for transferring information between the first execution code container created by the remote target resource and the first device. Therefore, proper resources can be configured for the execution target code, the problem that the execution resources are insufficient or too many when the code execution resources are designated in advance is avoided, and the resource utilization rate is fully improved.
In this embodiment, as shown in fig. 3, after the first code execution container finishes executing the target code, the first device may receive, through the proxy container, a code execution result generated after the first code execution container executes the target code, and output the code execution result through the local information output component.
As shown in fig. 3, wherein the information output component communicates with the second code execution container and the agent container. When the target code is executed by the first code execution container created by the remote target resource, the first code execution container may feed back an execution result of the execution target code to the proxy container, and the proxy container transmits the execution result to the information output component, so that the information output component outputs the code execution result.
Accordingly, when the target code is executed by the second code execution container created by the resource local to the first device, the second code execution container may also feed back an execution result of the execution target code to the information output component, and further, the information output component outputs the code execution result.
With such an embodiment, the remote code execution results may be fed back to the local via the proxy container, thereby simulating the effect of dynamically deploying computing resources locally (i.e., the first device).
Based on the same inventive concept, the embodiments of the present application also provide a resource allocation system, which may be applied to a device cluster, where the device cluster includes at least a first device and a second device.
Based on the same inventive concept, the embodiment of the present application further provides a resource allocation method, and referring to fig. 4, a step flowchart of another resource allocation method is shown, and as shown in fig. 4, the method may be applied to a second device in a device cluster, and specifically may include the following steps:
step S401: and receiving a resource scheduling instruction sent by the resource management server.
The resource scheduling instruction is generated by the resource management server when a resource allocation request sent by the first device through the proxy container is received, and the resource allocation request is generated by a code writing container on the first device after the completion of target code writing is detected.
Wherein the proxy container and the code writing container are both created by the first device using local resources.
In this embodiment, the process of creating, by the first device, the proxy container and the code writing container by using the local resources, specifically, sending the resource allocation request through the code writing container, receiving, by the proxy container, the resource allocation request sent by the code writing container, and applying, by the proxy container, the target resource to the resource management server may be described in steps S201 to S203 of the foregoing embodiment, which is not described herein again.
When the resource management server sends a resource allocation request according to the proxy container, a proper target resource can be scheduled from the device cluster, and a resource scheduling instruction is sent to a second device where the target resource is located, so that the second device can receive the resource scheduling instruction sent by the resource management server.
Step S402: and creating a first code execution container by utilizing the local target resource according to the resource scheduling instruction.
In this embodiment, the second device may create the first code execution container by using a local target resource according to a resource scheduling instruction sent by the resource management server, where the target resource may be a CPU resource or a graphics card resource. The second device creating the first code execution container using the target resource can be understood as: the second device creates a first code execution container for executing the code in the software and hardware environment provided by the target resource, for example, the first code execution container is set in the graphics card resource of the second device, so that the code is executed by using the graphics card resource, that is, the target code is executed in the graphics card resource, and the first code execution container can be understood as a process environment for executing the code.
Step S403: and executing the target code through the first code execution container when receiving a code execution request sent by the first device through the proxy container.
In this embodiment, the second device may receive the code execution request sent by the agent container of the first device, and further the second device may execute the target code through the created first code execution container.
Wherein in one embodiment, the target code may be obtained from a storage disk on which a code writing container is mounted when the target code is executed by the first code execution container; further, the object code is executed by the first code execution container.
Wherein the storage disk is shared by the code writing container and the first code execution container.
In this embodiment, the code execution request received by the second device may include the address of the storage disk, so that the second device may obtain the written target code from the storage disk according to the address of the storage disk. Further, the first code container is started to execute the object code.
The object code may be edited by the user in a code writing container and stored in a storage disk as described in step S201 above.
In one embodiment, after the first code execution container executes the target code, the second device may feed back the code execution result of the first code execution container executing the target code to the proxy container, and further, the first device may output the code execution result through an information output component local to the first device, so as to be convenient for the user to view.
By adopting the technical scheme of the embodiment of the application, the second device can establish the first code execution container for executing the target code by utilizing the local target resource according to the resource scheduling of the resource management server, and further can remotely execute the target code written by the first device. And the creation of the object code is completed by the first device using the code writing container created by the local resource, and the resource scheduling request is requested by the agent container on the first device when the first device finishes writing the object code. As such, the resources required to write the code may originate from the first device's local resources and, upon execution of the code, may be executed remotely by the second device using the local target resources. In this way, on the one hand, the target resources required for executing the code can be not fixed to the user, so that the target resources can be allocated to other users for use before being allocated to the user, and the problem that the resources are occupied when the resources in the code editing stage and the code executing stage are both resources fixed to the user is avoided. On the other hand, the target resource required by the execution code is dynamically applied to the resource management server when the code is executed, so that the target resource is the resource required by the execution code in an adaptive manner, the problem of insufficient or excessive execution resource caused by the execution resource of the code appointed in advance is avoided, and the resource utilization rate is fully improved.
Referring to FIG. 5, a schematic diagram of a framework of a resource allocation system is shown, as shown in FIG. 5, which may include, in particular, a code writing container and a proxy container at a first device, and a first code execution container at a second device.
Wherein the second device is configured to provide a target resource required for executing the target code.
Wherein the code writing container may be for: writing target codes for users;
wherein the proxy container may be for: the code writing container is used for receiving a resource allocation request sent by the code writing container, applying for target resources required by executing the target code to a resource management server, and sending a code execution request to the first code execution container, wherein the code execution request is used for requesting to execute the target code;
wherein the first code execution container may be for: executing the target code.
The resource allocation system provided in this embodiment may further include an IDE interface provided for a user, where when the user writes a code by using a first device through the IDE interface, the code may be written through a writing environment provided by a code writing container, and further after the code writing is completed, the proxy container may apply, to the resource management server, a target resource required for executing the target code, and then, a second device where the target resource is located may create a first code execution container for executing the target code, where the first code execution container provides an execution environment for executing the target code.
The resource allocation system provided by the embodiment can provide local resources for code writing, and dynamically apply for required target resources in the code execution stage, so that the problems of resource occupation, insufficient or excessive resources caused by allocation of fixed resources when a user starts to use the first device are avoided, and the resource utilization rate of the device cluster is improved.
Referring to FIG. 6, a schematic diagram of a framework of a resource allocation system is shown in greater detail, as shown in FIG. 6, including a code writing container and a proxy container at a first device, a first code execution container at a second device, and a second code execution container and an information output component at the first device.
Wherein the code writing container is in communication with the agent container and the second code execution container, respectively, the agent container is in communication with the first code execution container, and the information output component is in communication with the second code execution container and the agent container simultaneously.
In this case, a code writing container is configured to send a code execution request to the second code execution container, the code execution request being configured to request execution of the target code; the second code execution container is for executing the object code.
When the method is adopted, the code writing container can determine whether to apply for remote resources or local resources according to the size of the resources required by executing the target code or according to the code execution mode required by a user for executing the code, so that the dynamic allocation of the resources is realized. Specifically, if a remote resource is applied, the code writing container applies for a target resource on the second device through the proxy container, and if a local resource is applied, the code writing container sends a code execution request to the second code execution container, so that the second code execution container executes the target code.
As shown in fig. 6, the system may further include an information output component located at the first device, in which case the proxy container may be further configured to receive a code execution result generated after the first code execution container executes the target code; the information output component is used for outputting the code execution result.
Specifically, the information output component may further receive and output a code execution result generated after the second code execution container executes the object code when the object code is executed by the second code execution container created by the resource local to the first device.
As shown in fig. 6, the system may further include a storage disk.
Wherein the storage disk may be shared by a code writing container and a code executing container for executing the object code; in this embodiment, the code execution container may be a first code execution container or a second code execution container.
The code writing container is used for mounting the target code into the storage disk;
the first device can carry a storage address of the target code in the storage disk through a code execution request sent by the code writing container, so that the code execution container reads the target code from the storage disk according to the storage address, and further executes the target code.
The storage disks may include data disks and rootfs disks, which may be shared by code writing containers and code execution containers.
Referring to fig. 7, a schematic structural diagram of a resource allocation apparatus according to an embodiment of the present application is shown, where the apparatus is applied to a device cluster, and the device cluster includes at least a first device, where the apparatus is disposed on the first device, and may specifically include the following modules:
an encoding container creating module 701, configured to create a code writing container by using a local resource, and write an object code through the code writing container;
A proxy container creation module 702, configured to create a proxy container by using a local resource, and receive, through the proxy container, a resource allocation request sent by the code writing container, where the resource allocation request is used to request execution of the target code;
a resource application module 703, configured to apply, through the proxy container, for the target resource to a resource management server;
a code execution module 704, configured to send, through the proxy container, a code execution request to the first code execution container, where the code execution request is used to instruct the second device to execute the target code through the first code execution container, and the first code execution container is created by the second device that provides the target resource using the target resource.
Optionally, the apparatus may further include the following modules:
a first creation module for creating a second code execution container using a local resource;
a request execution module, configured to send a code execution request to the second code execution container through the code writing container, where the code execution request is used to request execution of the target code;
and the first execution module is used for executing the target code through the second code execution container.
Optionally, the apparatus may further include the following modules:
the result receiving module is used for receiving a code execution result generated after the first code execution container executes the target code through the proxy container;
and the output module is used for outputting the code execution result through a local information output component.
Optionally, the apparatus may further include the following modules:
the mounting module is used for mounting the target code into a storage disk through the code writing container, the storage disk is shared by the code writing container and a code executing container, and the code executing container is used for executing the target code;
the sent code execution request carries the storage address of the target code in the storage disk, so that the code execution container reads the target code from the storage disk according to the storage address, and further executes the target code.
Optionally, the apparatus may further include the following modules:
the execution mode acquisition module is used for acquiring a code execution mode configured by a user through the code writing container;
the proxy container creation module 702 creates a proxy container by using local resources, specifically:
When the code execution mode is remote execution, a local resource is utilized to create a proxy container, so that the target resource is applied to a resource management server through the proxy container, the first code execution container is obtained based on the target resource, and the target code is executed through the first code execution container;
the first creating module is specifically configured to create a second code execution container by using a local resource when the code execution mode is local execution, and execute the target code through the second code execution container.
Referring to fig. 8, a schematic structural diagram of a resource allocation apparatus according to an embodiment of the present application is shown, where the apparatus is applied to a device cluster, and the device cluster includes at least a second device, and the apparatus is deployed on the second device, and may specifically include the following modules:
a resource scheduling receiving module 801, configured to receive a resource scheduling instruction sent by a resource management server, where the resource scheduling instruction is generated by the resource management server when a resource allocation request sent by a first device through a proxy container is received, and the resource allocation request is generated by a code writing container on the first device after completion of target code writing is detected; wherein the proxy container and the code writing container are both created by the first device using local resources;
A container creation module 802, configured to create a first code execution container by using a local target resource according to the resource scheduling instruction;
and a code executing module 803, configured to execute, when receiving a code execution request sent by the first device through the proxy container, the target code through the first code execution container.
Optionally, the code execution module 803 may specifically include the following units:
the code acquisition unit is used for acquiring the target code from a storage disk mounted on the code writing container; wherein the storage disk is shared by the code writing container and the first code executing container;
and the code execution unit is used for executing the target code through the first code execution container.
Optionally, the apparatus may further include the following modules:
the result acquisition module is used for acquiring a code execution result generated after the first code execution container executes the target code;
and the result sending module is used for sending the code execution result to the proxy container so that the first device outputs the code execution result through an information output component local to the first device.
It should be noted that, for simplicity of description, the method embodiments are shown as a series of acts, but it should be understood by those skilled in the art that the embodiments are not limited by the order of acts, as some steps may occur in other orders or concurrently in accordance with the embodiments. Further, those skilled in the art will appreciate that the embodiments described in the specification are presently preferred embodiments, and that the acts are not necessarily required by the embodiments of the invention.
The embodiment of the invention also provides an electronic device which can be used for executing the resource allocation method, and can comprise a memory, a processor and a computer program which is stored on the memory and can be run on the processor, wherein the processor is configured to execute the resource allocation method.
The embodiment of the invention also provides a computer readable storage medium, and a stored computer program causes a processor to execute the resource allocation method according to the embodiment of the invention.
In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described by differences from other embodiments, and identical and similar parts between the embodiments are all enough to be referred to each other.
It will be apparent to those skilled in the art that embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the invention may take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiment and all such alterations and modifications as fall within the scope of the embodiments of the invention.
Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or terminal device comprising the element.
The foregoing has described in detail a method, apparatus, device and storage medium for resource allocation, according to the present invention, wherein specific examples are employed to illustrate the principles and embodiments of the present invention, and the above examples are only used to help understand the method and core idea of the present invention; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present invention, the present description should not be construed as limiting the present invention in view of the above.
Claims (16)
1. A method of resource allocation, the method being applied to a first device, the method comprising:
creating a code writing container by utilizing local resources, and writing target codes through the code writing container;
creating a proxy container by utilizing local resources, and receiving a resource allocation request sent by the code writing container through the proxy container, wherein the resource allocation request is used for requesting target resources required by executing the target code;
applying for the target resource to a resource management server through the proxy container;
and sending, by the proxy container, a code execution request to a first code execution container, the code execution request being for instructing execution of the target code by the first code execution container, the first code execution container being created by a second device that provides the target resource using the target resource.
2. The method according to claim 1, wherein the method further comprises:
creating a second code execution container using the local resource;
sending a code execution request to the second code execution container through the code writing container, the code execution request being for requesting execution of the object code;
And executing the target code through the second code execution container.
3. The method according to claim 1, wherein the method further comprises:
receiving a code execution result generated after the target code is executed by the first code execution container through the proxy container;
and outputting the code execution result through a local information output component.
4. The method according to claim 1 or 2, characterized in that the method further comprises:
mounting the target code to a storage disk through the code writing container, wherein the storage disk is shared by the code writing container and a code executing container, and the code executing container is used for executing the target code;
the sent code execution request carries the storage address of the target code in the storage disk, so that the code execution container reads the target code from the storage disk according to the storage address, and further executes the target code.
5. The method of claim 2, wherein after writing object code by the code writing container, the method further comprises:
obtaining a code execution mode configured by a user through the code writing container;
Creating a proxy container using local resources, comprising:
when the code execution mode is remote execution, a local resource is utilized to create a proxy container, so that the target resource is applied to a resource management server through the proxy container, the first code execution container is obtained based on the target resource, and the target code is executed through the first code execution container;
creating a second code execution container using the local resource, comprising:
and under the condition that the code execution mode is local execution, creating a second code execution container by utilizing local resources, and executing the target code through the second code execution container.
6. A method of resource allocation, the method being applied to a second device, the method comprising:
receiving a resource scheduling instruction sent by a resource management server, wherein the resource scheduling instruction is generated by the resource management server when a resource allocation request sent by first equipment through a proxy container is received, and the resource allocation request is generated by a code writing container on the first equipment after the completion of target code writing is detected; wherein the proxy container and the code writing container are both created by the first device using local resources;
According to the resource scheduling instruction, a first code execution container is created by utilizing a local target resource;
and executing the target code through the first code execution container when receiving a code execution request sent by the first device through the proxy container.
7. The method of claim 6, wherein executing the object code through the first code execution container comprises:
acquiring the target code from a storage disk mounted by the code writing container; wherein the storage disk is shared by the code writing container and the first code executing container;
executing the target code through the first code execution container.
8. The method according to claim 6 or 7, characterized in that the method further comprises:
acquiring a code execution result generated after the target code is executed by the first code execution container;
and sending the code execution result to the proxy container so that the first device outputs the code execution result through an information output component local to the first device.
9. A resource allocation apparatus, the apparatus deployed at a first device, comprising:
The code container creation module is used for creating a code writing container by utilizing local resources and writing target codes through the code writing container;
the agent container creation module is used for creating an agent container by utilizing local resources, and receiving a resource allocation request sent by the code writing container through the agent container, wherein the resource allocation request is used for requesting target resources required by executing the target code;
the resource application module is used for applying the target resource to a resource management server through the proxy container;
and the code execution module is used for sending a code execution request to a first code execution container through the proxy container, wherein the code execution request is used for instructing the first code execution container to execute the target code, and the first code execution container is created by a second device for providing the target resource by using the target resource.
10. A resource allocation apparatus, the apparatus deployed at a second device, comprising:
the resource scheduling receiving module is used for receiving a resource scheduling instruction sent by a resource management server, wherein the resource scheduling instruction is generated by the resource management server when a resource allocation request sent by first equipment through a proxy container is received, and the resource allocation request is generated by a code writing container on the first equipment after the completion of target code writing is detected; wherein the proxy container and the code writing container are both created by the first device using local resources;
The container creation module is used for creating a first code execution container by utilizing a local target resource according to the resource scheduling instruction;
and the code execution module is used for executing the target code through the first code execution container when receiving the code execution request sent by the first equipment through the proxy container.
11. A resource allocation system, characterized by being applied to a cluster of devices, said cluster of devices comprising at least a first device and a second device, said system comprising:
the first device includes: code authoring containers and proxy containers created using local resources; the second device is configured to provide target resources required for executing target code, including: executing a container by using the first code created by the target resource;
wherein the code writing container is for: writing the target code by a user;
the agent container is used for: receiving a resource allocation request sent by the code writing container, applying for target resources required by executing the target code to a resource management server, and sending a code execution request to the first code execution container, wherein the code execution request is used for requesting to execute the target code;
The first code execution container is for: executing the target code.
12. The system of claim 11, wherein the first device further comprises: creating a second code execution container using the local resource;
wherein the code writing container is for: sending a code execution request to the second code execution container, the code execution request being for requesting execution of the object code;
the second code execution container is for: executing the target code.
13. The system of claim 11, wherein the first device further comprises: an information output component;
the agent container is used for: receiving a code execution result generated after the first code execution container executes the target code;
the information output component is used for: and outputting the code execution result.
14. The system according to claim 11 or 12, characterized in that the system further comprises: a storage disk shared by the code writing container and a code executing container for executing the target code;
the code writing container is used for: mounting the target code to the storage disk;
The first device carries a storage address of the target code in the storage disk through a code execution request sent by the code writing container, so that the code execution container reads the target code from the storage disk according to the storage address, and further executes the target code.
15. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor when executed implementing a resource allocation method according to any one of claims 1-5 or 6-8.
16. A computer readable storage medium storing a computer program for causing a processor to perform the resource allocation method according to any one of claims 1-5 or 6-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110114221.1A CN112817691B (en) | 2021-01-27 | 2021-01-27 | Resource allocation method, device, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110114221.1A CN112817691B (en) | 2021-01-27 | 2021-01-27 | Resource allocation method, device, equipment and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112817691A CN112817691A (en) | 2021-05-18 |
| CN112817691B true CN112817691B (en) | 2024-02-02 |
Family
ID=75860039
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110114221.1A Active CN112817691B (en) | 2021-01-27 | 2021-01-27 | Resource allocation method, device, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112817691B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2969812A1 (en) * | 2014-12-05 | 2016-06-09 | Amazon Technologies, Inc. | Automatic management of resource sizing |
| CN108958930A (en) * | 2018-06-19 | 2018-12-07 | 北京百度网讯科技有限公司 | Processing method, device and the equipment of container resource |
| CN108958927A (en) * | 2018-05-31 | 2018-12-07 | 康键信息技术（深圳）有限公司 | Dispositions method, device, computer equipment and the storage medium of container application |
| CN112214330A (en) * | 2020-11-04 | 2021-01-12 | 腾讯科技（深圳）有限公司 | Method and device for deploying master nodes in cluster and computer-readable storage medium |
-
2021
- 2021-01-27 CN CN202110114221.1A patent/CN112817691B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2969812A1 (en) * | 2014-12-05 | 2016-06-09 | Amazon Technologies, Inc. | Automatic management of resource sizing |
| CN108958927A (en) * | 2018-05-31 | 2018-12-07 | 康键信息技术（深圳）有限公司 | Dispositions method, device, computer equipment and the storage medium of container application |
| CN108958930A (en) * | 2018-06-19 | 2018-12-07 | 北京百度网讯科技有限公司 | Processing method, device and the equipment of container resource |
| CN112214330A (en) * | 2020-11-04 | 2021-01-12 | 腾讯科技（深圳）有限公司 | Method and device for deploying master nodes in cluster and computer-readable storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 密码服务器运算资源可扩展管理研究;容晓峰;周利华;;电信科学(第08期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112817691A (en) | 2021-05-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105988874B (en) | Resource processing method and device | |
| US10154089B2 (en) | Distributed system and data operation method thereof | |
| CN106056529B (en) | Method and equipment for training convolutional neural network for picture recognition | |
| CN106293893B (en) | Job scheduling method and device and distributed system | |
| CN103067468B (en) | Cloud dispatching method and system thereof | |
| CN109324903B (en) | Display resource scheduling method and device for embedded system | |
| KR20210095690A (en) | Resource management method and apparatus, electronic device and recording medium | |
| CN112463375A (en) | Data processing method and device | |
| CN106897299B (en) | Database access method and device | |
| CN113419846B (en) | Resource allocation method and device, electronic equipment and computer readable storage medium | |
| CN104144202B (en) | Access method, system and the device of Hadoop distributed file system | |
| CN115048216B (en) | Resource management scheduling method, device and equipment of artificial intelligent cluster | |
| CN111679911A (en) | Management method, device, equipment and medium for GPU (graphics processing Unit) card in cloud environment | |
| KR20190028210A (en) | Cloud service method and system for deployment of artificial intelligence application using container | |
| CN104518897A (en) | Resource management optimization processing method and resource management optimization processing device for virtual firewalls | |
| CN113296926A (en) | Resource allocation method, computing device and storage medium | |
| CN113535087B (en) | Data processing method, server and storage system in data migration process | |
| CN114490062A (en) | Local disk scheduling method and device, electronic equipment and storage medium | |
| CN111813541B (en) | Task scheduling method, device, medium and equipment | |
| CN105677481A (en) | Method and system for processing data and electronic equipment | |
| CN113204425A (en) | Method and device for process management internal thread, electronic equipment and storage medium | |
| CN112817691B (en) | Resource allocation method, device, equipment and medium | |
| US12028269B2 (en) | Method for optimal resource selection based on available GPU resource analysis in large-scale container platform | |
| CN111796932A (en) | GPU resource scheduling method | |
| CN112114958A (en) | Resource isolation method, distributed platform, computer device, and storage medium |
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 |