CN111240836A - Computing resource management method and device, electronic equipment and storage medium - Google Patents
Computing resource management method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN111240836A CN111240836A CN202010010110.1A CN202010010110A CN111240836A CN 111240836 A CN111240836 A CN 111240836A CN 202010010110 A CN202010010110 A CN 202010010110A CN 111240836 A CN111240836 A CN 111240836A
- Authority
- CN
- China
- Prior art keywords
- user
- clusters
- candidate
- cluster
- computing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
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]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/5044—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering hardware capabilities
-
- 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
- G06F9/5072—Grid computing
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Power Sources (AREA)
Abstract
The application discloses a calculation resource management method, a calculation resource management device, electronic equipment and a storage medium, and relates to the field of deep learning, wherein the method comprises the following steps: acquiring a computing resource allocation request of any user and determining a management strategy suitable for the user; determining a cluster corresponding to the management strategy from the M constructed clusters, wherein M is a positive integer greater than one; and taking the determined clusters as candidate clusters, and allocating the computing resources of at least one of the candidate clusters to the user for use. By applying the scheme, the resource utilization rate can be improved.
Description
Technical Field
The present application relates to computer application technologies, and in particular, to a computing power resource management method and apparatus, an electronic device, and a storage medium in the field of deep learning.
Background
With the rapid development of deep learning, the types, complexity and data processing amount of networks are rapidly increasing, and the demand for deep learning computational resources is higher and higher. Computing resources may include processors, memory, bandwidth, and hard disks, wherein the processors may include Central Processing Units (CPUs), Graphics Processing Units (GPUs), and the like.
The demands of deep learning projects of different users on computing resources may also be different, but an effective computing resource management method does not exist at present.
Disclosure of Invention
In view of the above, the present application provides a computing resource management method, device, electronic device and storage medium.
A computing power resource management method, comprising:
acquiring a computing resource allocation request of any user and determining a management strategy applicable to the user;
determining a cluster corresponding to the management strategy from the M constructed clusters, wherein M is a positive integer greater than one;
and taking the determined clusters as candidate clusters, and allocating the computing resources of at least one of the candidate clusters to the user for use.
According to a preferred embodiment of the present application, the determining the management policy applicable to the user includes:
acquiring preset user information of the user;
and determining a management strategy suitable for the user according to the user information.
According to a preferred embodiment of the present application, before allocating the computing resources of at least one of the candidate clusters to the user for use, the method further includes: filtering out clusters which can not provide services from the candidate clusters.
According to a preferred embodiment of the present application, the allocating the computing resources of at least one of the candidate clusters to the user for use includes:
sorting the candidate clusters according to the principle that the performance is from high to low;
allocating the calculation resources of the clusters at the first N positions after sorting to the user for use, wherein N is a positive integer;
or the sorted clusters are distributed to the user, so that the user can select at least one computing resource of the clusters from the clusters for use.
According to a preferred embodiment of the present application, the ranking the candidate clusters includes:
for each candidate cluster, scoring the candidate cluster from at least two preset dimensions respectively, and weighting and adding the scores to obtain a comprehensive score of the candidate cluster;
and sorting the candidate clusters according to the sequence of the composite score from high to low.
According to a preferred embodiment of the present application, before acquiring the computing power resource allocation request of any user, the method further includes: allocating a calculation card with a preset limit for the user;
after the obtaining of the computing resource allocation request of any user, the method further includes: determining whether the user's calculation card has a surplus limit, if so, determining a management strategy applicable to the user, and if not, rejecting the request;
after the allocating the computing resources of at least one of the candidate clusters to the user for use, further comprising: and deducting a corresponding amount from the calculation power card according to the use condition of the user on the calculation power resource.
A computing force resource management apparatus comprising: a first management unit and a second management unit;
the first management unit is used for acquiring a computing resource allocation request of any user, determining a management strategy applicable to the user, and determining a cluster corresponding to the management strategy from M constructed clusters, wherein M is a positive integer greater than one;
and the second management unit is used for taking the determined clusters as candidate clusters and allocating the computing resources of at least one of the candidate clusters to the user for use.
According to a preferred embodiment of the present application, the first management unit obtains predetermined user information of the user, and determines a management policy applicable to the user according to the user information.
According to a preferred embodiment of the present application, the second management unit is further configured to filter out clusters that cannot provide services from the candidate clusters.
According to a preferred embodiment of the present application, the second management unit ranks the candidate clusters according to a principle that performance is from high to low, and allocates the computing power resources of the top N ranked clusters to the user for use, where N is a positive integer, or allocates each ranked cluster to the user, so that the user selects the computing power resource of at least one cluster from the computing power resources for use.
According to a preferred embodiment of the present application, the second management unit scores each candidate cluster from at least two predetermined dimensions, weights and adds the scores to obtain a composite score of the candidate clusters, and sorts the candidate clusters in order of the composite score from high to low.
According to a preferred embodiment of the present application, the second management unit is further configured to allocate a predetermined amount of computing power card to the user before the first management unit obtains the computing power resource allocation request of the user; after the computing power resource of at least one cluster in the candidate clusters is distributed to the user for use, deducting a corresponding amount from the computing power card according to the use condition of the user for the computing power resource;
the first management unit is further configured to, after the computing power resource allocation request of the user is obtained, determine whether a remaining amount exists in a computing power card of the user, if so, determine a management policy applicable to the user, and if not, reject the request.
An electronic device, comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform a method as described above.
A non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method as described above.
One embodiment in the above application has the following advantages or benefits: the computing power resource is platformized in a multi-cluster mode, limited computing power resources are distributed to users more reasonably in a smaller granularity and more flexible mode, and the users and the computing power resources can be decoupled through strategies, so that on one hand, the computing power resources can be conveniently expanded and contracted through increasing and decreasing servers in the clusters, increasing and decreasing the clusters and the like, the reasonable configuration of the computing power resources is realized, on the other hand, the computing power resources can be scheduled through adjusting the strategies under the condition that the total amount of the computing power resources is not changed, the computing power resources used by the users are reasonably distributed, and the flexibility of computing power resource distribution, the resource utilization rate and the like are improved; in addition, by filtering out clusters which cannot provide services, sequencing each candidate cluster and the like, computing resources with better performance are provided for users as much as possible; furthermore, the available computing power resources of the users can be limited by the computing power card, so that the problem of unreasonable resource allocation caused by monopolization of the computing power resources by part of the users is avoided, and the resource utilization rate is further improved; other effects of the above-described alternative will be described below with reference to specific embodiments.
Drawings
The drawings are included to provide a better understanding of the present solution and are not intended to limit the present application. Wherein:
FIG. 1 is a flow chart of an embodiment of a computing power resource management method of the present application;
FIG. 2 is a schematic diagram of an embodiment of a computing resource management apparatus 200 according to the present application;
FIG. 3 is a schematic diagram illustrating an interaction of the computing resource management apparatus with users and clusters according to the present application;
fig. 4 is a block diagram of an electronic device according to the method of an embodiment of the present application.
Detailed Description
The following description of the exemplary embodiments of the present application, taken in conjunction with the accompanying drawings, includes various details of the embodiments of the application for the understanding of the same, which are to be considered exemplary only. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present application. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
In addition, it should be understood that the term "and/or" herein is merely one type of association relationship that describes an associated object, meaning that three relationships may exist, e.g., a and/or B may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, the character "/" herein generally indicates that the former and latter related objects are in an "or" relationship.
Fig. 1 is a flowchart of an embodiment of a computing power resource management method according to the present application. As shown in fig. 1, the following detailed implementation is included.
In 101, a computing resource allocation request of any user is obtained, and a management policy applicable to the user is determined.
At 102, a cluster corresponding to the management policy is determined from the constructed M clusters, where M is a positive integer greater than one.
In 103, the determined clusters are used as candidate clusters, and the computing resources of at least one of the candidate clusters are allocated to the user for use.
M clusters can be constructed in advance according to a multi-dimensional containerization technology, M is a positive integer larger than one, specific values can be determined according to actual needs, and computational resource limits and the like owned by the clusters can be set. How to build a cluster is prior art.
The multi-dimensions may include regions, which may include northeast regions, etc., and processor types, which may include CPUs, GPUs, etc., among others. For example, the first cluster may correspond to a first region, the processor type is CPU, the second cluster may correspond to a second region, the processor type is GPU, the third cluster may correspond to the first region, the processor type is GPU, and the like, and the CPU and the GPU may be further subdivided, for example, the GPU is divided into P40 and V100, and the like. Which content is specifically included in the multiple dimensions may depend on the actual needs.
After a computing resource allocation request of any user is acquired, a management strategy applicable to the user can be determined. As a possible implementation manner, predetermined user information may be obtained first, and then a management policy applicable to a user may be determined according to the obtained user information.
The specific content included in the user information may also be determined according to actual needs, for example, the user information may include user attribute information, and the user attribute information may include whether the user pays a fee, and the like.
As a possible implementation manner, a corresponding relationship between the user information and the management policy may be pre-constructed, so that the management policy corresponding to the acquired user information may be determined according to the corresponding relationship, and the determined management policy may be used as the management policy applicable to the user.
Then, a cluster corresponding to the management policy applicable to the user can be determined from the M constructed clusters. For example, if the user is an unpaid user, the management policy applicable to the user is management policy 1, and management policy 1 is a processor that can only use a CPU type, then the cluster with the processor type of CPU may be used as the cluster corresponding to management policy 1. For another example, if the user is a paid user, the management policy applicable to the user is management policy 2, and the management policy 2 is a processor capable of using a GPU type, then the cluster with the processor type being a GPU may be used as the cluster corresponding to the management policy 2. Of course, the above description is only for illustration and is not intended to limit the technical solution of the present application, and the specific management policy may be determined according to actual needs.
The cluster corresponding to the management policy applicable to the user may be used as a candidate cluster. The number of the management policies applicable to the user may be one or multiple, if the number of the management policies is one, the cluster corresponding to the management policy may be determined to be a candidate cluster, and if the number of the management policies is multiple, the cluster corresponding to each management policy may be determined to be a candidate cluster. One management policy may correspond to one or more clusters, and one cluster may also correspond to one or more management policies.
In addition, clusters that cannot provide services can be filtered out of the candidate clusters, for example, clusters that reach the computational resource limit can be filtered out of the candidate clusters, and the subsequent processing is continued for the remaining candidate clusters.
The obtained candidate clusters are usually multiple, and further, the candidate clusters can be sorted according to the principle that the performance is from high to low.
As a possible implementation manner, for each candidate cluster, the candidate cluster may be scored from at least two predetermined dimensions, and the scores may be added in a weighted manner, so that a composite score of the candidate cluster may be obtained, and then the candidate clusters may be sorted in order of the composite score from high to low.
The at least two dimensions may include, but are not limited to, the following:
a) regional relevance of users and computing resources: the area where the user is located can be sensed through an intelligent Domain Name System (DNS), the area where the user is located can be determined by combining historical operation information of deep learning items of the user, the area corresponding to each candidate cluster is known, and for any candidate cluster, if the area where the user is located is the same as the area corresponding to the candidate cluster, the area where the user is located can be considered to be related, otherwise, the area where the user is located can be considered to be unrelated. If relevant, a higher score may be assigned, if not relevant, a lower score may be assigned, and the cluster speed for accessing the same area will generally be faster.
b) Computing resource performance: the better the processor and/or memory and/or storage performance of the cluster, the higher the score may be.
c) Cluster load: the computing power resource limit of the computing power resource/cluster being used by the cluster, "/" indicates the quotient divided by, the larger the quotient, the lower the score may be.
d) Cluster bandwidth load: the average of the occupied bandwidths/the maximum bandwidth of the cluster within a certain duration may be the latest predetermined duration, for example, the latest hour, and the larger the obtained quotient is, the lower the score may be.
e) And (4) project operation statistical information: the average starting speed of the items in the cluster within a certain time length may refer to the latest preset time length, and the faster the average starting speed is, the higher the score may be.
For any candidate cluster, after the scores of the dimensions are respectively obtained, the scores of the dimensions can be added in a weighted manner, so that the comprehensive score of the candidate cluster is obtained. The weight corresponding to each dimension can be determined according to actual needs.
Correspondingly, the candidate clusters can be sorted according to the sequence of the composite score from high to low, and the computing power resources of the first N-bit cluster after sorting can be allocated to the user for use, where N is a positive integer, and a specific value can be determined according to actual needs, such as 1, or each sorted cluster can be allocated to the user, so that the user can select at least one computing power resource of a cluster from the ranked clusters for use.
For example, the following steps are carried out: the number of the candidate clusters is 4, the candidate clusters are respectively a cluster 1, a cluster 2, a cluster 3 and a cluster 4, the comprehensive scores of the candidate clusters are respectively obtained, and the candidate clusters are ranked according to the sequence of the comprehensive scores from high to low, so that the ranked clusters are as follows: cluster 2, cluster 1, cluster 4, and cluster 3, the computing resources of cluster 2 may be allocated to the user for use, or each candidate cluster after ranking may be allocated to the user, and the user may select the computing resources of cluster 2 for use, or select the computing resources of cluster 2 and cluster 1 for use, etc.
In the solution described in this embodiment, for different users, a certain amount of force calculating cards may be allocated to the users respectively. The credit lines of the credit cards corresponding to different users can be different, and the specific credit line can be determined according to actual needs. The computing power card can be used for limiting available computing power resources of the user, so that the problems of unreasonable resource allocation and the like caused by monopolization of the computing power resources by part of users are solved, and the resource utilization rate can be improved.
Correspondingly, in this embodiment, after the computing power resource allocation request of the user is acquired, it may be first determined whether a remaining amount exists in the computing power card of the user, if so, it may be determined that a management policy and the like are applicable to the user, that is, the subsequent processing is continued, and if not, the request may be rejected.
In addition, after the computing power resource of at least one cluster in the candidate clusters is allocated to the user for use, the corresponding credit can be deducted from the computing power card according to the use condition of the user for the computing power resource.
The specific form of the force calculating card is not limited, for example, the time length of the force calculating resource can be expressed, and the corresponding time length can be deducted from the force calculating card according to the time length of the force calculating resource used by the user. If the amount of the computing power card becomes zero in the process of using the computing power resource, the user can not be allowed to continue using the computing power resource any more, namely, the allocated computing power resource is recovered after the computing power card is expired. The user may then be reassigned a force card, etc., if desired.
It should be noted that for simplicity of description, the aforementioned method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present application is not limited by the order of acts, as some steps may occur in other orders or concurrently depending on the application. Further, those skilled in the art should also appreciate that the embodiments described in the specification are preferred embodiments and that the acts and modules referred to are not necessarily required in this application.
In a word, by adopting the scheme of the embodiment of the method, the computational power resources can be platformized in a multi-cluster mode, the limited computational power resources can be more reasonably distributed to the users in a smaller granularity and a more flexible mode, and the users and the computational power resources can be decoupled through strategies, so that on one hand, the expansion and contraction capacity of the computational power resources can be conveniently performed by increasing and decreasing servers in the clusters, increasing and decreasing the clusters and the like, the reasonable configuration of the computational power resources is realized, on the other hand, the computational power resources can be scheduled through adjusting the strategies under the condition that the total amount of the computational power resources is not changed, and the computational power resources used by the users are reasonably distributed, so that the flexibility of the computational power resource distribution, the resource utilization rate and the like are improved; in addition, by filtering out clusters which cannot provide services, sequencing each candidate cluster and the like, computing resources with better performance are provided for users as much as possible; in addition, the available computing power resources of the users can be limited by the computing power card, so that the problem of unreasonable resource allocation caused by monopolization of the computing power resources by part of the users is solved, and the resource utilization rate is further improved.
The above is a description of method embodiments, and the embodiments of the present application are further described below by way of apparatus embodiments.
Fig. 2 is a schematic structural diagram of an embodiment of a computing resource management device 200 according to the present application. As shown in fig. 2, includes: a first management unit 201 and a second management unit 202.
The first management unit 201 is configured to obtain a computing resource allocation request of any user, determine a management policy applicable to the user, and determine a cluster corresponding to the management policy from M clusters that are constructed, where M is a positive integer greater than one.
And the second management unit 202 is configured to use the determined cluster as a candidate cluster, and allocate the computing resources of at least one of the candidate clusters to the user for use.
M clusters can be constructed in advance according to a multi-dimensional containerization technology, M is a positive integer larger than one, and computational resource limits and the like owned by the clusters can be set.
The multi-dimensions may include regions, which may include northeast regions, etc., and processor types, which may include CPUs, GPUs, etc., among others.
After acquiring the computing power resource allocation request of any user, the first management unit 201 may determine a management policy applicable to the user. Specifically, predetermined user information may be first acquired, and then a management policy applicable to a user may be determined according to the acquired user information. As a possible implementation manner, a corresponding relationship between the user information and the management policy may be pre-constructed, so that the management policy corresponding to the acquired user information may be determined according to the corresponding relationship, and the determined management policy may be used as the management policy applicable to the user. Then, the first management unit 201 may further determine, from the M clusters that are constructed, a cluster corresponding to the management policy applicable to the user.
The second management unit 202 may use the cluster corresponding to the management policy applicable to the user as a candidate cluster, and may filter out a cluster that cannot provide the service from the candidate cluster, for example, filter out a cluster that reaches the computational resource limit from the candidate cluster, and perform subsequent processing on the remaining candidate cluster.
Further, the second management unit 202 may rank the candidate clusters according to a principle that performance is from high to low, for example, for each candidate cluster, score the candidate cluster from at least two predetermined dimensions, add the scores in a weighted manner to obtain a composite score of the candidate cluster, and rank the candidate clusters according to a sequence that the composite score is from high to low.
After the sorting is completed, the second management unit 202 may allocate the computing resources of the first N-bit sorted clusters to the user for use, where N is a positive integer, or allocate each sorted cluster to the user, so that the user selects at least one computing resource of the cluster from the sorted clusters for use.
In addition, the second management unit 202 may further allocate a predetermined quota of the computing power card to the user before the first management unit 201 obtains the computing power resource allocation request of the user, and after allocating the computing power resource of at least one of the candidate clusters to the user for use, may further deduct a corresponding quota from the computing power card according to the use condition of the user for the computing power resource. Correspondingly, after acquiring the calculation resource allocation request of the user, the first management unit 201 may further determine whether a remaining amount exists in the calculation card of the user, if so, may determine a management policy applicable to the user, that is, continue the subsequent processing, and if not, may reject the request.
The specific form of the force calculating card is not limited, for example, the time length of the force calculating resource can be expressed, and the corresponding time length can be deducted from the force calculating card according to the time length of the force calculating resource used by the user. If the amount of the computing power card becomes zero in the process of using the computing power resource, the user can not be allowed to continue using the computing power resource any more, namely, the allocated computing power resource is recovered after the computing power card is expired. The user may then be reassigned a force card, etc., if desired.
For a specific work flow of the embodiment of the apparatus shown in fig. 2, reference is made to the related description in the foregoing method embodiment, and details are not repeated.
In addition, based on the above description, fig. 3 is a schematic diagram illustrating an interaction manner between the computing power resource management apparatus and the user and each cluster. As shown in fig. 3, M clusters may be constructed in advance, and a predetermined amount of force cards may be allocated to the user. After a user sends a calculation resource allocation request, if the remaining amount in the calculation card of the user is determined, the preset user information can be obtained, and a management strategy suitable for the user can be determined according to the obtained user information. And then, determining a cluster corresponding to the management policy from the M constructed clusters to serve as a candidate cluster, and filtering out clusters which cannot provide services from the candidate cluster. Then, the candidate clusters can be sorted according to the principle that the performance is from high to low, for example, for each candidate cluster, the candidate cluster can be respectively scored from at least two predetermined dimensions, and the scores are weighted and added, so that the composite score of the candidate cluster is obtained, and then the candidate clusters can be sorted according to the sequence that the composite score is from high to low. The at least two dimensions may include, but are not limited to: the regional relevance of users and computing resources, computing resource performance, cluster load, cluster bandwidth load, project operation statistical information and the like can acquire relevant information by interacting with the clusters and the like. The computing power resources of the first N ranked clusters may be allocated to the user for use, where N is a positive integer, or each ranked cluster may be allocated to the user, so that the user may select at least one computing power resource from the clusters for use, for example, deep learning may be performed using the computing power resources. In addition, corresponding quota can be deducted from the power card according to the use condition of the user on the power resource.
In a word, by adopting the scheme of the embodiment of the application device, the computing power resource can be platformized in a multi-cluster mode, the limited computing power resource can be more reasonably distributed to the user in a smaller granularity and more flexible mode, and the user and the computing power resource can be decoupled through a strategy, so that on one hand, the expansion and contraction capacity of the computing power resource can be conveniently performed by increasing and decreasing servers in the cluster, increasing and decreasing the cluster and the like, the reasonable configuration of the computing power resource is realized, on the other hand, the computing power resource can be scheduled through adjusting the strategy under the condition that the total amount of the computing power resource is not changed, the computing power resource used by the user is reasonably distributed, and the flexibility of the distribution of the computing power resource, the resource utilization rate and the like are; in addition, by filtering out clusters which cannot provide services, sequencing each candidate cluster and the like, computing resources with better performance are provided for users as much as possible; in addition, the available computing power resources of the users can be limited by the computing power card, so that the problem of unreasonable resource allocation caused by monopolization of the computing power resources by part of the users is solved, and the resource utilization rate is further improved.
According to an embodiment of the present application, an electronic device and a readable storage medium are also provided.
Fig. 4 is a block diagram of an electronic device according to the method of the embodiment of the present application. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the present application that are described and/or claimed herein.
As shown in fig. 4, the electronic apparatus includes: one or more processors Y01, a memory Y02, and interfaces for connecting the various components, including a high speed interface and a low speed interface. The various components are interconnected using different buses and may be mounted on a common motherboard or in other manners as desired. The processor may process instructions for execution within the electronic device, including instructions stored in or on the memory to display graphical information for a graphical user interface on an external input/output device (such as a display device coupled to the interface). In other embodiments, multiple processors and/or multiple buses may be used, along with multiple memories and multiple memories, as desired. Also, multiple electronic devices may be connected, with each device providing portions of the necessary operations (e.g., as a server array, a group of blade servers, or a multi-processor system). In fig. 4, one processor Y01 is taken as an example.
Memory Y02 is a non-transitory computer readable storage medium as provided herein. Wherein the memory stores instructions executable by at least one processor to cause the at least one processor to perform the methods provided herein. The non-transitory computer readable storage medium of the present application stores computer instructions for causing a computer to perform the methods provided herein.
Memory Y02, which is a non-transitory computer readable storage medium, may be used to store non-transitory software programs, non-transitory computer executable programs, and modules, such as program instructions/modules corresponding to the methods in the embodiments of the present application (e.g., xx module X01, xx module X02, and xx module X03 shown in fig. X). The processor Y01 executes various functional applications of the server and data processing, i.e., implements the method in the above-described method embodiments, by executing non-transitory software programs, instructions, and modules stored in the memory Y02.
The memory Y02 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to use of the electronic device, and the like. Additionally, the memory Y02 may include high speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, memory Y02 may optionally include memory located remotely from processor Y01, which may be connected to the electronic device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The electronic device may further include: an input device Y03 and an output device Y04. The processor Y01, the memory Y02, the input device Y03 and the output device Y04 may be connected by a bus or other means, and the bus connection is exemplified in fig. 4.
The input device Y03 may receive input numeric or character information and generate key signal inputs related to user settings and function control of the electronic device, such as a touch screen, keypad, mouse, track pad, touch pad, pointer, one or more mouse buttons, track ball, joystick, or other input device. The output device Y04 may include a display device, an auxiliary lighting device, a tactile feedback device (e.g., a vibration motor), and the like. The display device may include, but is not limited to, a liquid crystal display, a light emitting diode display, and a plasma display. In some implementations, the display device can be a touch screen.
Various implementations of the systems and techniques described here can be realized in digital electronic circuitry, integrated circuitry, application specific integrated circuits, computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.
These computer programs (also known as programs, software applications, or code) include machine instructions for a programmable processor, and may be implemented using high-level procedural and/or object-oriented programming languages, and/or assembly/machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and/or device (e.g., magnetic discs, optical disks, memory, programmable logic devices) used to provide machine instructions and/or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal used to provide machine instructions and/or data to a programmable processor.
To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a cathode ray tube or a liquid crystal display monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local area networks, wide area networks, and the internet.
The computer system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.
It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present application may be executed in parallel, sequentially, or in different orders, and the present invention is not limited thereto as long as the desired results of the technical solutions disclosed in the present application can be achieved.
The above-described embodiments should not be construed as limiting the scope of the present application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made in accordance with design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims (14)
1. A computing power resource management method, comprising:
acquiring a computing resource allocation request of any user and determining a management strategy applicable to the user;
determining a cluster corresponding to the management strategy from the M constructed clusters, wherein M is a positive integer greater than one;
and taking the determined clusters as candidate clusters, and allocating the computing resources of at least one of the candidate clusters to the user for use.
2. The method of claim 1,
the determining the management policy applicable to the user comprises:
acquiring preset user information of the user;
and determining a management strategy suitable for the user according to the user information.
3. The method of claim 1,
before the allocating the computing resources of at least one of the candidate clusters to the user for use, further comprises: filtering out clusters which can not provide services from the candidate clusters.
4. The method of claim 1,
the allocating computational resources of at least one of the candidate clusters to the user for use comprises:
sorting the candidate clusters according to the principle that the performance is from high to low;
allocating the calculation resources of the clusters at the first N positions after sorting to the user for use, wherein N is a positive integer;
or the sorted clusters are distributed to the user, so that the user can select at least one computing resource of the clusters from the clusters for use.
5. The method of claim 4,
the ranking of the candidate clusters includes:
for each candidate cluster, scoring the candidate cluster from at least two preset dimensions respectively, and weighting and adding the scores to obtain a comprehensive score of the candidate cluster;
and sorting the candidate clusters according to the sequence of the composite score from high to low.
6. The method of claim 1,
before the obtaining of the computing resource allocation request of any user, the method further includes: allocating a calculation card with a preset limit for the user;
after the obtaining of the computing resource allocation request of any user, the method further includes: determining whether the user's calculation card has a surplus limit, if so, determining a management strategy applicable to the user, and if not, rejecting the request;
after the allocating the computing resources of at least one of the candidate clusters to the user for use, further comprising: and deducting a corresponding amount from the calculation power card according to the use condition of the user on the calculation power resource.
7. A computing force resource management apparatus, comprising: a first management unit and a second management unit;
the first management unit is used for acquiring a computing resource allocation request of any user, determining a management strategy applicable to the user, and determining a cluster corresponding to the management strategy from M constructed clusters, wherein M is a positive integer greater than one;
and the second management unit is used for taking the determined clusters as candidate clusters and allocating the computing resources of at least one of the candidate clusters to the user for use.
8. The apparatus of claim 7,
the first management unit acquires the preset user information of the user and determines the management strategy applicable to the user according to the user information.
9. The apparatus of claim 7,
the second management unit is further configured to filter out clusters that cannot provide services from the candidate clusters.
10. The apparatus of claim 7,
and the second management unit ranks the candidate clusters according to a principle that the performance is from high to low, and allocates the computing power resources of the first N ranked clusters to the user for use, wherein N is a positive integer, or allocates each ranked cluster to the user so that the user can select at least one computing power resource of a cluster for use.
11. The apparatus of claim 10,
and the second management unit scores the candidate clusters from at least two preset dimensions respectively aiming at each candidate cluster, weights and adds the scores to obtain the comprehensive score of the candidate clusters, and sorts the candidate clusters according to the sequence of the comprehensive score from high to low.
12. The apparatus of claim 7,
the second management unit is further used for distributing a calculation capacity card with a preset limit for the user before the first management unit acquires the calculation capacity resource distribution request of the user; after the computing power resource of at least one cluster in the candidate clusters is distributed to the user for use, deducting a corresponding amount from the computing power card according to the use condition of the user for the computing power resource;
the first management unit is further configured to, after the computing power resource allocation request of the user is obtained, determine whether a remaining amount exists in a computing power card of the user, if so, determine a management policy applicable to the user, and if not, reject the request.
13. An electronic device, comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.
14. A non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method of any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010010110.1A CN111240836A (en) | 2020-01-06 | 2020-01-06 | Computing resource management method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010010110.1A CN111240836A (en) | 2020-01-06 | 2020-01-06 | Computing resource management method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111240836A true CN111240836A (en) | 2020-06-05 |
Family
ID=70865274
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010010110.1A Pending CN111240836A (en) | 2020-01-06 | 2020-01-06 | Computing resource management method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111240836A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111901862A (en) * | 2020-07-07 | 2020-11-06 | 西安交通大学 | User clustering and power distribution method, device and medium based on deep Q network |
| CN112085419A (en) * | 2020-09-25 | 2020-12-15 | 中国建设银行股份有限公司 | Resource acquisition method, device and equipment |
| CN113641670A (en) * | 2021-07-09 | 2021-11-12 | 北京百度网讯科技有限公司 | Data storage and data retrieval method and device, electronic equipment and storage medium |
| WO2021244343A1 (en) * | 2020-06-01 | 2021-12-09 | 杭州嘉楠耘智信息科技有限公司 | Cloud computing power allocation method, user terminal, cloud computing power platform and system |
| CN113810438A (en) * | 2020-06-12 | 2021-12-17 | 中国移动通信有限公司研究院 | Scheduling and requesting methods of service computing resources, node equipment and terminal |
| CN113840317A (en) * | 2020-06-08 | 2021-12-24 | 中国移动通信有限公司研究院 | Calculation capacity reporting method, calculation capacity obtaining method, calculation capacity network element and calculation capacity sensing control network element |
| WO2022016466A1 (en) * | 2020-07-23 | 2022-01-27 | 北京小米移动软件有限公司 | Resource request information processing method and apparatus, and communication device and storage medium |
| CN115904663A (en) * | 2022-12-02 | 2023-04-04 | 滨州心若网络科技有限公司 | Information disaster tolerance method and system based on database and cloud platform |
| US20230129548A1 (en) * | 2021-10-26 | 2023-04-27 | Dell Products L.P. | Datacenter efficiency management system |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101938504A (en) * | 2009-06-30 | 2011-01-05 | 深圳市融创天下科技发展有限公司 | Cluster server intelligent dispatching method and system |
| CN103324535A (en) * | 2012-03-23 | 2013-09-25 | 百度在线网络技术（北京）有限公司 | Method and device for allocating computing resources |
| CN106101074A (en) * | 2016-05-31 | 2016-11-09 | 北京大学 | A kind of sacurity dispatching method based on user's classification towards big data platform |
| CN107145395A (en) * | 2017-07-04 | 2017-09-08 | 北京百度网讯科技有限公司 | Method and apparatus for handling task |
| CN107168788A (en) * | 2016-03-07 | 2017-09-15 | 阿里巴巴集团控股有限公司 | The dispatching method and device of resource in distributed system |
| CN107169586A (en) * | 2017-03-29 | 2017-09-15 | 北京百度网讯科技有限公司 | Resource optimization method, device and storage medium based on artificial intelligence |
| CN107609084A (en) * | 2017-09-06 | 2018-01-19 | 华中师范大学 | One kind converges convergent resource correlation method based on gunz |
| CN108667940A (en) * | 2018-05-22 | 2018-10-16 | 深信服网络科技（深圳）有限公司 | Resource allocation methods, device and the computer readable storage medium of cloud platform |
| CN108989418A (en) * | 2018-07-11 | 2018-12-11 | 国云科技股份有限公司 | A kind of resource amount method of mixed cloud object storage common authentication |
| CN109684065A (en) * | 2018-12-26 | 2019-04-26 | 北京云联万维技术有限公司 | A kind of resource regulating method, apparatus and system |
| CN109783237A (en) * | 2019-01-16 | 2019-05-21 | 腾讯科技（深圳）有限公司 | A kind of resource allocation method and device |
| CN110221915A (en) * | 2019-05-21 | 2019-09-10 | 新华三大数据技术有限公司 | Node scheduling method and apparatus |
| CN110351384A (en) * | 2019-07-19 | 2019-10-18 | 深圳前海微众银行股份有限公司 | Big data platform method for managing resource, device, equipment and readable storage medium storing program for executing |
| CN110442449A (en) * | 2019-07-09 | 2019-11-12 | 北京云和时空科技有限公司 | A kind of resource regulating method and device |
| CN110532092A (en) * | 2019-08-21 | 2019-12-03 | 云湾科技(嘉兴)有限公司 | Reso urce matching method, device and equipment |
-
2020
- 2020-01-06 CN CN202010010110.1A patent/CN111240836A/en active Pending
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101938504A (en) * | 2009-06-30 | 2011-01-05 | 深圳市融创天下科技发展有限公司 | Cluster server intelligent dispatching method and system |
| CN103324535A (en) * | 2012-03-23 | 2013-09-25 | 百度在线网络技术（北京）有限公司 | Method and device for allocating computing resources |
| CN107168788A (en) * | 2016-03-07 | 2017-09-15 | 阿里巴巴集团控股有限公司 | The dispatching method and device of resource in distributed system |
| CN106101074A (en) * | 2016-05-31 | 2016-11-09 | 北京大学 | A kind of sacurity dispatching method based on user's classification towards big data platform |
| CN107169586A (en) * | 2017-03-29 | 2017-09-15 | 北京百度网讯科技有限公司 | Resource optimization method, device and storage medium based on artificial intelligence |
| CN107145395A (en) * | 2017-07-04 | 2017-09-08 | 北京百度网讯科技有限公司 | Method and apparatus for handling task |
| CN107609084A (en) * | 2017-09-06 | 2018-01-19 | 华中师范大学 | One kind converges convergent resource correlation method based on gunz |
| CN108667940A (en) * | 2018-05-22 | 2018-10-16 | 深信服网络科技（深圳）有限公司 | Resource allocation methods, device and the computer readable storage medium of cloud platform |
| CN108989418A (en) * | 2018-07-11 | 2018-12-11 | 国云科技股份有限公司 | A kind of resource amount method of mixed cloud object storage common authentication |
| CN109684065A (en) * | 2018-12-26 | 2019-04-26 | 北京云联万维技术有限公司 | A kind of resource regulating method, apparatus and system |
| CN109783237A (en) * | 2019-01-16 | 2019-05-21 | 腾讯科技（深圳）有限公司 | A kind of resource allocation method and device |
| CN110221915A (en) * | 2019-05-21 | 2019-09-10 | 新华三大数据技术有限公司 | Node scheduling method and apparatus |
| CN110442449A (en) * | 2019-07-09 | 2019-11-12 | 北京云和时空科技有限公司 | A kind of resource regulating method and device |
| CN110351384A (en) * | 2019-07-19 | 2019-10-18 | 深圳前海微众银行股份有限公司 | Big data platform method for managing resource, device, equipment and readable storage medium storing program for executing |
| CN110532092A (en) * | 2019-08-21 | 2019-12-03 | 云湾科技(嘉兴)有限公司 | Reso urce matching method, device and equipment |
Non-Patent Citations (1)
| Title |
|---|
| 赵帅: "《网络服务计算基础》", 北京邮电大学出版社, pages: 169 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021244343A1 (en) * | 2020-06-01 | 2021-12-09 | 杭州嘉楠耘智信息科技有限公司 | Cloud computing power allocation method, user terminal, cloud computing power platform and system |
| CN113840317A (en) * | 2020-06-08 | 2021-12-24 | 中国移动通信有限公司研究院 | Calculation capacity reporting method, calculation capacity obtaining method, calculation capacity network element and calculation capacity sensing control network element |
| CN113810438A (en) * | 2020-06-12 | 2021-12-17 | 中国移动通信有限公司研究院 | Scheduling and requesting methods of service computing resources, node equipment and terminal |
| CN111901862B (en) * | 2020-07-07 | 2021-08-13 | 西安交通大学 | User clustering and power distribution method, device and medium based on deep Q network |
| CN111901862A (en) * | 2020-07-07 | 2020-11-06 | 西安交通大学 | User clustering and power distribution method, device and medium based on deep Q network |
| CN114258729A (en) * | 2020-07-23 | 2022-03-29 | 北京小米移动软件有限公司 | Resource request information processing method and device, communication equipment and storage medium |
| WO2022016466A1 (en) * | 2020-07-23 | 2022-01-27 | 北京小米移动软件有限公司 | Resource request information processing method and apparatus, and communication device and storage medium |
| CN112085419A (en) * | 2020-09-25 | 2020-12-15 | 中国建设银行股份有限公司 | Resource acquisition method, device and equipment |
| CN113641670A (en) * | 2021-07-09 | 2021-11-12 | 北京百度网讯科技有限公司 | Data storage and data retrieval method and device, electronic equipment and storage medium |
| CN113641670B (en) * | 2021-07-09 | 2023-08-11 | 北京百度网讯科技有限公司 | Data storage and data retrieval method and device, electronic equipment and storage medium |
| US20230129548A1 (en) * | 2021-10-26 | 2023-04-27 | Dell Products L.P. | Datacenter efficiency management system |
| US11915061B2 (en) * | 2021-10-26 | 2024-02-27 | Dell Products L.P. | Datacenter efficiency management system for migrating workload across nodes based on workload performance efficiency ranking |
| CN115904663A (en) * | 2022-12-02 | 2023-04-04 | 滨州心若网络科技有限公司 | Information disaster tolerance method and system based on database and cloud platform |
| CN115904663B (en) * | 2022-12-02 | 2024-01-05 | 中雄世纪征信有限公司 | Information disaster recovery method and system based on database and cloud platform |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111240836A (en) | Computing resource management method and device, electronic equipment and storage medium | |
| CN109783237B (en) | Resource allocation method and device | |
| US9946579B2 (en) | Systems and methods for load balancing of time-based tasks in a distributed computing system | |
| JP6241300B2 (en) | Job scheduling apparatus, job scheduling method, and job scheduling program | |
| US9531636B2 (en) | Extending processing capacity of server | |
| CN111694646B (en) | Resource scheduling method, device, electronic equipment and computer readable storage medium | |
| US11496413B2 (en) | Allocating cloud computing resources in a cloud computing environment based on user predictability | |
| JP7119082B2 (en) | Application Prioritization for Automatic Diagonal Scaling in Distributed Computing Environments | |
| CN111880914A (en) | Resource scheduling method, resource scheduling apparatus, electronic device, and storage medium | |
| KR20210040299A (en) | Method and apparatus for processing transaction requests in blockchain, device, medium, and program | |
| US11190581B1 (en) | Job allocation support system and method | |
| CN111506401A (en) | Automatic driving simulation task scheduling method and device, electronic equipment and storage medium | |
| Singh et al. | Cloud resource management optimization: taxonomy and research challenges | |
| CN114090223A (en) | Memory access request scheduling method, device, equipment and storage medium | |
| CN112565356A (en) | Data storage method and device and electronic equipment | |
| JP7182836B2 (en) | Automatic Diagonal Scaling of Workloads in Distributed Computing Environments | |
| CN110688229A (en) | Task processing method and device | |
| US20060149611A1 (en) | Peer to peer resource negotiation and coordination to satisfy a service level objective | |
| CN111796940B (en) | Resource allocation method and device and electronic equipment | |
| CN112527509A (en) | Resource allocation method and device, electronic equipment and storage medium | |
| CN111858030A (en) | Job resource processing method and device, electronic equipment and readable storage medium | |
| CN111176838A (en) | Method and device for distributing embedded vectors for nodes in bipartite graph | |
| CN116647560A (en) | Method, device, equipment and medium for coordinated optimization control of Internet of things computer clusters | |
| CN111598571A (en) | Transaction processing method, device, equipment and storage medium of block chain | |
| CN115604269A (en) | Load balancing method and device of server, electronic equipment 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 |