CN115248734A - Private cloud multi-tenant resource quota self-adaptive adjustment method and device - Google Patents
Private cloud multi-tenant resource quota self-adaptive adjustment method and device Download PDFInfo
- Publication number
- CN115248734A CN115248734A CN202211148614.5A CN202211148614A CN115248734A CN 115248734 A CN115248734 A CN 115248734A CN 202211148614 A CN202211148614 A CN 202211148614A CN 115248734 A CN115248734 A CN 115248734A
- Authority
- CN
- China
- Prior art keywords
- resource
- quota
- tenant
- resource quota
- utilization rate
- 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.)
- Granted
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]
- G06F9/5061—Partitioning or combining of resources
Abstract
The invention discloses a resource quota self-adaptive adjustment method and device for multiple tenants in a private cloud. The device mainly comprises a resource utilization rate monitoring module and a resource quota controlling module, wherein the resource quota controlling module is used for adaptively adjusting the resource quota and periodically executing an adjusting process based on a timing task. According to the method, the actual resource utilization rate result of each tenant in the last period is calculated through the resource utilization rate monitoring data, the resource quota adjustment is performed on the tenant with an unreasonable resource quota based on the preset expected utilization rate adjustment threshold, and compared with the resource quota setting function provided by the cluster management platform, the method can adapt to the dynamic changes of factors such as the tenant group scale, the use habit and the time, the unnecessary cluster expansion is avoided, and the cluster resource utilization rate is effectively improved in a light-weight mode.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a resource quota self-adaptive adjusting device for multiple tenants in a private cloud.
Background
The traditional private cloud usually adopts a staged deployment mode, expansion and contraction capacity is not as convenient as elasticity of public cloud, in the expansion process, working items with complex operation steps are more, operations such as manual script writing and command execution are performed, efficiency is low, risk is high, and an auditing mechanism is lacked. Therefore, the maximization of the current resource utilization rate of the private cloud platform is realized, and unnecessary expansion and contraction capacity and resource waste are avoided. When a plurality of users or teams share the private cloud cluster with the fixed node number, the situation that the tenant maliciously occupies too many resources is avoided, and resource allocation based on the fairness principle is a role to be undertaken by an administrator. Application containerization deployment and management platforms such as K8S (Kubernets) and Yarn provide corresponding mechanisms to support multi-tenant resource allocation, for example, K8S provides resource quotas, which are defined by ResourceQuota objects to provide a limit to the total resource consumption of each namespace. It may limit the total upper limit on the number of objects of a certain type in the namespace, as well as the total upper limit on the computational resources that Pod in the command space can use. However, such mechanisms are performed by manually setting thresholds and parameters. The demand of platform resources changes dynamically with the change of factors such as tenant population scale, use habits, time and the like. The mode not only depends on the observation experience of management personnel on resources under tenants for regulation and control, has low flexibility, but also is used as a passive response to the use change of the resources, does not have the agility of active regulation, and easily causes two extreme phenomena of resource waste and over busy business.
Therefore, a scheme for adaptively adjusting resource quota of multiple tenants in private cloud is needed.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a resource quota self-adaptive adjusting device for multiple tenants in a private cloud, which solves the problems in the background technology.
In order to achieve the purpose, the invention is realized by the following technical scheme: a resource quota self-adaptive adjustment method for multiple tenants in a private cloud comprises the following steps:
s1: the resource quota control module sets the initialized resource quota of each tenant based on the total amount of the private cloud cluster resources and the principle of allocation according to needs;
s2: the resource utilization rate monitoring module continuously monitors and collects the resource utilization rate data of each tenant in time slices and reports the data to the resource quota control module;
s3: the resource quota control module periodically performs self-adaptive adjustment on resource quotas of each tenant according to the adjustment threshold values of the resource utilization rate data and the expected utilization rate in the last period;
s4: entering the next resource quota adjusting period or performing cluster expansion after exceeding the capacity expansion-free threshold;
s5: and (4) under the control of the timing task, entering the next resource quota adjusting period, and repeating the step S1 to the step S4.
Preferably, the setting of the initialized resource quota of each tenant in step S1 includes the following steps:
s1.1: the tenant management module shares all tenant management of the private cloud resource pool, including management of the life cycle of the tenant; the resource quota control module sets an expansion-free threshold, an expected utilization rate and an adjustment threshold of the expected utilization rate;
s1.2: when a private cloud adds a tenant, a resource quota control module pre-divides a quota pool of the tenant in the quota resource pool according to the current resource demand of a user; if the resource quota in the idle quota pool is still larger than the capacity expansion-free threshold value after the quota pool of the tenant is pre-divided, completing resource allocation through a resource quota setting interface of the cluster; if the resource quota in the idle quota pool is smaller than the capacity-expansion-free threshold value, triggering a one-time integral resource quota adjusting process; after the adjustment of the resource quota is completed, performing the pre-division calculation again;
preferably, the capacity expansion-free threshold includes a positive value and a negative value, and if the capacity expansion-free threshold is set to the positive value, the resource in the cluster is allowed to be in a reserved state, where the reserved state is that the resource quota allocated by the cluster is always smaller than the actual total resource quota; if the value is set to be a negative value, the resources in the cluster are in a super-allocation state, and the super-allocation state is that the allocated resource quota of the cluster can exceed the actual total amount of the resources.
Preferably, the capacity-expansion-free threshold is differentiated according to the granularity of the resource types, different capacity-expansion-free thresholds are set for different resource types, and the cluster capacity expansion is triggered when the quota total of any one resource type exceeds the capacity-expansion-free threshold.
Preferably, the resource usage rate data in step S2 has validity and invalidity, and for a newly added user in the current period, the newly added user does not have complete collected data in an adjustment period, the data is invalid, and for the invalid monitoring data, the corresponding tenant is excluded in the subsequent adjustment process, and the resource quota of the tenant is not changed.
Preferably, the adaptively adjusting resource quotas of each tenant in step S3 includes the following steps:
s3.1: the resource quota control module analyzes and processes the reported resource utilization rate data to obtain a resource utilization rate result of each tenant in one period, and the result is used as the input of a resource quota adjusting algorithm;
s3.2: inputting the utilization rate results of all types of resources of each tenant in a single period, idle quota pool data, a preset capacity expansion-free threshold value and an adjustment threshold value of an expected utilization rate into a resource quota adjustment algorithm, and obtaining the resource quota to be adjusted in the single period of each tenant through calculation and output;
s3.3: and comparing the expected resource quota and the actual resource quota of each tenant, and when the difference exceeds an adjusting threshold value, dividing the quota pool again for the cluster resource pool through the resource quota setting interface.
Preferably, the resource quota managing and controlling module in step S3.1 analyzes and processes the reported resource usage rate data, including comparing the set expected usage rate with the actual resource usage rate, and recalculating a reasonable resource quota based on the expected usage rate when the expected usage rate exceeds the adjustment threshold of the expected usage rate.
Preferably, the step S4 of performing cluster expansion after exceeding the capacity-free threshold includes comparing the total value of the resource quota with a preset capacity-free threshold, and if there is a portion of the total value of the current resource quota of a certain resource type exceeding the total value of the cluster resources, which is greater than the capacity-free threshold, triggering cluster expansion.
The invention also provides a resource quota self-adaptive adjusting device of the private cloud with multiple tenants, which comprises a tenant management module, a resource quota control module, a resource utilization monitoring module and a cluster expansion and contraction module, wherein the tenant management module shares tenant information of a private cloud resource pool to the resource quota control module, the private cloud resource pool comprises a resource quota pool and a resource utilization pool, the resource quota control module manages the resource quota pool, the resource quota pool distributes the resource utilization pool, the resource utilization monitoring module collects data of the resource utilization pool and reports the data to the resource quota control module, and when the total resource quota exceeds an expansion-free threshold value, the cluster expansion and contraction module is triggered to execute cluster expansion on the resource quota pool.
Preferably, the resource quota control module comprises monitoring data processing, a quota adjusting algorithm, a quota setting interface and timing task control, the monitoring data processing analyzes and processes the data of the resource utilization monitoring module and outputs the result to the quota adjusting algorithm, the quota adjusting algorithm performs adaptive adjustment on the resource quota, when the difference between the expected resource quota and the actual resource quota exceeds an adjusting threshold value, the resource quota pool is divided again through the quota setting interface, and the timing task controls the operation of the management device; the resource quota pool comprises tenant quota data and an idle quota pool, and the resource usage pool comprises tenant usage data.
According to the method, the regulation process is periodically executed based on the timing task, the actual resource utilization rate result of each tenant in the last period is calculated through the resource utilization rate monitoring data, the self-adaptive regulation is carried out on the tenant with unreasonable resource quota in the private cloud resource pool based on the preset expected utilization rate regulation threshold value, and compared with the resource quota setting function provided by a cluster management platform, the method can adapt to the dynamic change of factors such as tenant group scale, use habits and time, the unnecessary cluster capacity expansion is avoided, and the cluster resource utilization rate is effectively improved in a light-weight manner.
Drawings
Fig. 1 is a general flowchart of a resource quota adaptive adjustment method according to the present invention;
fig. 2 is a schematic structural diagram of a resource quota adaptive adjustment apparatus according to the present invention;
fig. 3 is a resource quota adjusting flowchart in embodiment 1 of the present invention;
fig. 4 is a schematic diagram of a resource quota adjustment result in embodiment 1 of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments.
As shown in fig. 1, a resource quota adaptive adjustment method for multiple tenants in a private cloud includes the following implementation steps:
s1: the resource quota control module sets the initialized resource quota of each tenant based on the total private cloud cluster resource and the principle of allocation according to needs;
this step S1 operates as follows:
s1.1: the tenant management module shares all tenant management of the private cloud resource pool, including management of a tenant life cycle; the resource quota control module sets an expansion-free threshold, an expected utilization rate and an adjustment threshold of the expected utilization rate;
s1.2: when a private cloud adds a tenant, a resource quota control module pre-divides a quota pool of the tenant in the quota resource pool according to the current resource demand of a user; if the resource quota in the idle quota pool is still larger than the capacity expansion-free threshold value after the quota pool of the tenant is pre-divided, completing resource allocation through a resource quota setting interface of the cluster; if the resource quota in the idle quota pool is smaller than the capacity-expansion-free threshold value, triggering a primary integral resource quota adjusting process; after the adjustment of the resource quota is completed, the pre-division calculation is carried out again;
the capacity expansion free threshold comprises a positive value and a negative value, if the capacity expansion free threshold is set to the positive value, the resources in the cluster are allowed to be in a reserved state, and the reserved state is that the allocated resource quota of the cluster is always smaller than the actual total resource quota, so that higher stability can be provided for the application; if the value is set to be a negative value, the resources in the cluster are in an over-allocation state, the over-allocation state is that the allocated resource quota of the cluster can exceed the actual total amount of the resources, and the resource utilization rate is high.
The capacity-expansion-free threshold is divided according to the granularity of the resource types, different capacity-expansion-free thresholds are set for different resource types, and the cluster capacity expansion is triggered when the quota total of any one resource type exceeds the capacity-expansion-free threshold.
S2: the resource utilization rate monitoring module continuously monitors and controlsThe method comprises the following steps of collecting resource utilization rate data of each tenant according to time slices, reporting the data to a resource quota control module, and having the data format characteristics that: tenant
Is
Resource type
In that
The time is given by
。
The resource usage rate data in step S2 has validity and invalidity, and for a newly added user in the current period, the newly added user does not have complete collected data in an adjustment period, and the data is invalid, and for the invalid monitoring data, the corresponding tenant is excluded in the subsequent adjustment process, and the resource quota of the tenant is not changed.
S3: the resource quota control module periodically performs self-adaptive adjustment on resource quotas of each tenant according to the adjustment threshold values of the resource utilization rate data and the expected utilization rate in the last period;
this step S3 operates as follows:
s3.1: the resource quota control module analyzes and processes the reported resource utilization rate data to obtain a resource utilization rate result of each tenant in one period: obtaining tenants
Is
Current quota of a resource type
Suppose there is one regulation period T
With data collected in close proximity
After the time domain of the system is integrated, the average value is obtained to obtain the actual average utilization rate in a period. Tenant
The actual utilization rate of each resource adopts the following average value algorithm:
this result is used as input to the resource quota adjustment algorithm;
s3.2: inputting the utilization rate results of all types of resources of each tenant in a single period, idle quota pool data, a preset capacity expansion-free threshold value and an adjustment threshold value of an expected utilization rate into a resource quota adjustment algorithm, and obtaining the resource quota to be adjusted in the single period of each tenant through calculation and output;
s3.3: and comparing the expected resource quotas and the actual resource quotas of all the tenants, and when the difference exceeds the adjustment threshold of the expected utilization rate, carrying out quota pool division on the cluster resource pool again by taking the expected utilization rate as the reference through the resource quota setting interface. The adjusted threshold of expected usage includes an adjusted threshold of expected usage that is greater than the adjusted threshold of expected usage or less than the adjusted threshold of expected usage and an adjusted threshold of expected usage.
The expected utilization rate is a preset expected value for each resource type, represents a utilization rate reference which can be achieved by the resource type in the expected cluster, and comprises a down-regulation threshold value
And an up-regulation threshold
. If it is
If the resource quota of the resource type of the tenant is smaller than the down-regulation threshold value in the current period, the resource quota of the resource type of the tenant is reduced to meet the utilization rate reference, and waste caused by resource idleness is avoided; if it is
If the utilization rate of the resources in the current period of the tenant is greater than the up-regulation threshold value, the corresponding resource quota is increased, and the over-busy service is avoided; if it is
Then there is no need to adjust the resource quota of the tenant.
S4: entering the next resource quota adjusting period or performing cluster expansion after exceeding the capacity expansion-free threshold;
the step S4 of performing cluster expansion after exceeding the capacity-free threshold includes comparing the total value of the resource quota with a preset capacity-free threshold, and if there is a portion of the total value of the current resource quota of a certain resource type exceeding the total amount of the cluster resources, which is greater than the capacity-free threshold, it indicates that resource quota adjustment is insufficient to meet the current tenant demand, and further triggers the cluster expansion module to perform cluster expansion on the resource quota pool.
S5: and (4) under the control of the timing task, entering the next resource quota adjusting period, and repeating the step S1 to the step S4.
As shown in fig. 2, a resource quota self-adaptive adjusting device for multiple tenants in a private cloud includes a tenant management module, a resource quota management module, a resource usage monitoring module, and a cluster expansion module, where the tenant management module shares tenant information of a private cloud resource pool to the resource quota management module, the private cloud resource pool includes a resource quota pool and a resource usage pool, the resource quota management module manages the resource quota pool, the resource quota pool allocates the resource usage pool, the resource usage monitoring module collects data of the resource usage pool and reports the data to the resource quota management module, and when a total resource quota exceeds an expansion-free threshold, the cluster expansion module is triggered to perform cluster expansion on the resource quota pool.
The resource quota control module comprises monitoring data processing, a quota adjusting algorithm, a quota setting interface and timing task control, wherein the monitoring data processing analyzes and processes data of the resource utilization rate monitoring module and outputs a result to the quota adjusting algorithm, the quota adjusting algorithm performs self-adaptive adjustment on the resource quota, when the difference between an expected resource quota and an actual resource quota exceeds an adjusting threshold value, a resource quota pool is divided again through the quota setting interface, and the timing task controls the operation of the management device; the resource quota pool comprises tenant quota data and an idle quota pool, and the resource usage pool comprises tenant usage data.
Example 1
First, the term of the term related to example 1 in the present specification is explained.
K8s: kubernets, often abbreviated as K8s, is an open source system for automatically deploying, extending, and managing "containerized applications". The system was designed by Google and donated to Cloud Native Computing Foundation (now the Linux Foundation) for use. It aims to provide a platform for automatic deployment, extension and running of application containers across host clusters.
Naming space, which is referred to herein as the naming space in k8s, provides a mechanism to divide resources in the same cluster into isolated groups, and the names of the resources in the same naming space are unique.
The resource quota refers to the resource quota in k8s, is defined by a resource quota object, and provides a limit for the total resource consumption of each namespace. It may limit the total upper limit on the number of objects of a certain type in the namespace, as well as the total upper limit on the computational resources that may be used by the set of containers in the command space.
Promethus is an open monitoring solution from Borgmon, google, who added the CNCF foundation 5 months after k8s in 2016. Prometheus has the characteristics of easiness in management, high efficiency, expandability and easiness in integration.
As shown in fig. 3, which is a flowchart for adjusting the multi-tenant resource quota based on k8s, first, the flow for adjusting the multi-tenant resource quota is triggered by two situations: (1) creating a new tenant; and (2) timing tasks. The first mode is relatively simple, a new naming space is created in a k8s cluster and is mapped with a newly created tenant identity; and then, setting resource quotas of various types to be initialized under the name space directly according to the requirements of the tenants. In the second mode, firstly, monitoring data of the utilization rates of other object resources, such as a container group and the like, in each namespace are acquired, and the acquisition of the monitoring data is not limited to the Promethus API described in the embodiment; and calculating the actual utilization rate of each resource type of each tenant according to the mapping relation between the namespace and the tenant identity. Further, the resource usage of the nth tenant is determined, and the resource quota of a certain type of resource of the tenant is recalculated when | the actual usage-the expected usage | > is greater than the threshold, and the same manner is applicable to the (n + 1) th user. After the resource quotas to be set for each type of each tenant are obtained through calculation in the two modes, whether the difference value of the total resource quotas and the total resource quotas is larger than a capacity-free threshold value or not is further judged, and if the difference value is larger than the capacity-free threshold value, a k8s cluster capacity expansion mechanism is triggered; otherwise, directly creating or setting the attribute of the resource quota object in the namespace.
Fig. 4 shows a result diagram of adaptive adjustment of the resource allocation pool implemented once by the basic k8s, assuming that total resources of the k8s cluster are respectively 98-core CPU, memory 6T and 88-card GPU. At T 1 In the period, the cluster has 2 tenants, and a namespace a and a namespace b are respectively allocated to the tenants; according to the initial requirements of the user and the resource quota adjusting result in the pre-period, the current resource quota pool is distributed as follows: in a namespace a belonging to tenant 1, the resource quota object attribute of k8s comprises CPU quota of 64, memory quota of 2T, and quota of GPU of 32; similarly, in namespace b belonging to tenant 2, the resource quota object attribute includes a CPU quota of 32, a memory quota of 2T, and a GPU quota of 64. In addition, the idle quota pool size includes,2 cores of CPU,2T memory and-8 cards of GPU. In a T2 period, a newly added tenant 3 is added to the cloud platform, and the initial quota requirements are 2 cores of CPU,2T of memory and 8 cards of GPU. Because the GPU quota of the idle quota pool is smaller than or equal to the capacity-free threshold value, a primary adaptive resource quota adjusting process is triggered. Let T be assumed here 1 The average value of the utilization rate of the resource utilization pool in the period is shown in the figure, resource quota adjustment is performed based on 75% of a preset expected utilization rate down-regulation threshold value and 90% of an up-regulation threshold value, the utilization rate of the tenant 1 is lower than the down-regulation threshold value, therefore, the resource quota values of various types under the corresponding namespace a are down-regulated, the adjusted CPU quota is 3, the memory quota is 170G, and the GPU quota is 3. And the resource utilization rate of the tenant 2 is between 75% and 90%, so that the current situation is maintained and no adjustment is made. After the resource quota adjustment is completed, the resource quota of the tenant 3 is set, the CPU quota is 2, the memory quota is 2T, and the GPU quota is 8. The calculated idle quota pool is configured to 61-core CPUs, 1878G memories and 13-card GPUs after being adjusted, the idle quota of the GPU is larger than the capacity-expansion-free threshold value, namely the allocation of the current resource quota meets the requirement, the cluster can maintain the current situation, and the waste of capacity-expansion resources is avoided.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.
Claims (10)
1. A resource quota self-adaptive adjustment method for multiple tenants in a private cloud is characterized by comprising the following steps:
s1: the resource quota control module sets the initialized resource quota of each tenant based on the total amount of the private cloud cluster resources and the principle of allocation according to needs;
s2: the resource utilization rate monitoring module continuously monitors and collects the resource utilization rate data of each tenant in a time slice mode and reports the data to the resource quota control module;
s3: the resource quota control module periodically performs self-adaptive adjustment on resource quotas of each tenant according to the adjustment threshold values of the resource utilization rate data and the expected utilization rate in the last period;
s4: entering the next resource quota adjusting period or performing cluster expansion after exceeding the capacity expansion-free threshold;
s5: and (4) under the control of the timing task, entering the next resource quota adjusting period, and repeating the step S1 to the step S4.
2. The method for adaptively adjusting resource quotas of multiple tenants in a private cloud according to claim 1, wherein the setting of the initialized resource quotas of each tenant in step S1 comprises the following steps:
s1.1: the tenant management module shares all tenant management of the private cloud resource pool, including management of a tenant life cycle; the resource quota management and control module sets a capacity expansion-free threshold value, an expected utilization rate and an adjustment threshold value of the expected utilization rate;
s1.2: when a private cloud adds a tenant, a resource quota control module pre-divides a quota pool of the tenant in the quota resource pool according to the current resource demand of a user; if the resource quota in the idle quota pool is still larger than the capacity expansion-free threshold value after the quota pool of the tenant is pre-divided, completing resource allocation through a resource quota setting interface of the cluster; if the resource quota in the idle quota pool is smaller than the capacity-expansion-free threshold value, triggering a primary integral resource quota adjusting process; after the adjustment of the resource quota is completed, the pre-division calculation is performed again.
3. The method for adaptively adjusting resource quotas of multiple tenants in a private cloud according to claim 2, wherein: the capacity expansion-free threshold comprises a positive value and a negative value, and if the capacity expansion-free threshold is set to the positive value, the resource in the cluster is allowed to be in a reserved state; if the setting value is negative, the resources in the cluster are in a super-configuration state.
4. The method for adaptively adjusting resource quotas of multiple tenants in a private cloud according to claim 3, wherein: the capacity-expansion-free threshold is divided according to the granularity of the resource types, different capacity-expansion-free thresholds are set for different resource types, and the cluster capacity expansion is triggered when the quota total of any one resource type exceeds the capacity-expansion-free threshold.
5. The method for adaptively adjusting resource quotas of multiple tenants in a private cloud according to claim 1, wherein: the resource utilization rate data in the step S2 has validity and invalidity, for newly added users in the current period, the newly added users do not have complete acquisition data in an adjustment period, the data is invalid, and for invalid monitoring data, the corresponding tenants are excluded in the subsequent adjustment process, and the resource quota of the tenants is not changed.
6. The self-adaptive resource quota adjusting method for the private cloud with multiple tenants according to claim 2, characterized in that: in step S3, the self-adaptive adjustment of the resource quota of each tenant includes the following steps:
s3.1: the resource quota control module analyzes and processes the reported resource utilization rate data to obtain a resource utilization rate result of each tenant in one period, and the result is used as the input of a resource quota adjusting algorithm;
s3.2: inputting the utilization rate results of all types of resources of each tenant in a single period, idle quota pool data, a preset capacity expansion-free threshold value and an adjustment threshold value of an expected utilization rate into a resource quota adjustment algorithm, and obtaining the resource quota to be adjusted in the single period of each tenant through calculation and output;
s3.3: and comparing the expected resource quota and the actual resource quota of each tenant, and when the difference exceeds an adjusting threshold value, dividing the quota pool again for the cluster resource pool through the resource quota setting interface.
7. The self-adaptive resource quota adjusting method for the private cloud with multiple tenants according to claim 6, wherein the method comprises the following steps: in the step S3.1, the resource quota management and control module analyzes and processes the reported resource utilization rate data, including comparing the set expected utilization rate with the actual resource utilization rate, and recalculating a reasonable resource quota based on the expected utilization rate when the expected utilization rate exceeds the adjustment threshold of the expected utilization rate.
8. The self-adaptive resource quota adjusting method for the private cloud with multiple tenants according to claim 1, characterized in that: the step S4 of performing cluster expansion after exceeding the expansion-free threshold includes comparing the total value of the resource quota with a preset expansion-free threshold, and if there is a portion of the total value of the current resource quota of a certain resource type exceeding the total value of the cluster resources, which is greater than the expansion-free threshold, triggering cluster expansion.
9. A private cloud multi-tenant resource quota self-adaptive adjusting device is characterized in that: the system comprises a tenant management module, a resource quota control module, a resource utilization rate monitoring module and a cluster expansion and contraction module, wherein the tenant management module shares tenant information of a private cloud resource pool to the resource quota control module, the private cloud resource pool comprises a resource quota pool and a resource utilization pool, the resource quota control module manages the resource quota pool, the resource quota pool distributes the resource utilization pool, the resource utilization rate monitoring module collects data of the resource utilization pool and reports the data to the resource quota control module, and when the total resource quota exceeds a capacity expansion-free threshold value, the cluster expansion and contraction module is triggered to execute cluster expansion on the resource quota pool.
10. The device according to claim 9, wherein the device for adaptively adjusting resource quotas of multiple tenants in a private cloud comprises: the resource quota control module comprises monitoring data processing, a quota adjusting algorithm, a quota setting interface and timed task control, wherein the monitoring data processing analyzes and processes data of the resource utilization rate monitoring module and outputs a result to the quota adjusting algorithm, the quota adjusting algorithm performs adaptive adjustment on a resource quota, when the difference between an expected resource quota and an actual resource quota exceeds an adjusting threshold value, a resource quota pool is divided again through the quota setting interface, and the timed task controls the operation of the management device; the resource quota pool comprises tenant quota data and an idle quota pool, and the resource usage pool comprises tenant usage data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211148614.5A CN115248734B (en) | 2022-09-21 | 2022-09-21 | Private cloud multi-tenant resource quota self-adaptive adjustment method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211148614.5A CN115248734B (en) | 2022-09-21 | 2022-09-21 | Private cloud multi-tenant resource quota self-adaptive adjustment method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115248734A true CN115248734A (en) | 2022-10-28 |
| CN115248734B CN115248734B (en) | 2022-12-13 |
Family
ID=83699993
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211148614.5A Active CN115248734B (en) | 2022-09-21 | 2022-09-21 | Private cloud multi-tenant resource quota self-adaptive adjustment method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115248734B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114978998A (en) * | 2021-02-26 | 2022-08-30 | 中移(苏州)软件技术有限公司 | Flow control method, device, terminal and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111399986A (en) * | 2020-03-24 | 2020-07-10 | 中国建设银行股份有限公司 | Method and device for configuring Pod resource quota |
| CN112527504A (en) * | 2020-12-17 | 2021-03-19 | 金蝶蝶金云计算有限公司 | Multi-tenant resource quota management method and device, and computer equipment |
| WO2021129862A1 (en) * | 2019-12-26 | 2021-07-01 | 华为技术有限公司 | Method and apparatus for managing container cluster node resource pool |
| CN113127134A (en) * | 2019-12-30 | 2021-07-16 | 北京懿医云科技有限公司 | Container cluster, multi-tenant deployment method based on container cluster and electronic equipment |
| US20210288918A1 (en) * | 2020-03-16 | 2021-09-16 | NEC Laboratories Europe GmbH | Multi-resource and autonomous hierarchical brokering platform to enable slice resource exchange among heterogeneous network tenants |
| CN114666333A (en) * | 2022-04-02 | 2022-06-24 | 国网江苏省电力有限公司信息通信分公司 | Control method for cloud computing resource scheduling problem based on multi-tenant theory |
| CN115048260A (en) * | 2021-03-09 | 2022-09-13 | 中核武汉核电运行技术股份有限公司 | Cloud computing-based nuclear power plant PaaS platform resource quota monitoring method and system |
-
2022
- 2022-09-21 CN CN202211148614.5A patent/CN115248734B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021129862A1 (en) * | 2019-12-26 | 2021-07-01 | 华为技术有限公司 | Method and apparatus for managing container cluster node resource pool |
| CN113127134A (en) * | 2019-12-30 | 2021-07-16 | 北京懿医云科技有限公司 | Container cluster, multi-tenant deployment method based on container cluster and electronic equipment |
| US20210288918A1 (en) * | 2020-03-16 | 2021-09-16 | NEC Laboratories Europe GmbH | Multi-resource and autonomous hierarchical brokering platform to enable slice resource exchange among heterogeneous network tenants |
| CN111399986A (en) * | 2020-03-24 | 2020-07-10 | 中国建设银行股份有限公司 | Method and device for configuring Pod resource quota |
| CN112527504A (en) * | 2020-12-17 | 2021-03-19 | 金蝶蝶金云计算有限公司 | Multi-tenant resource quota management method and device, and computer equipment |
| CN115048260A (en) * | 2021-03-09 | 2022-09-13 | 中核武汉核电运行技术股份有限公司 | Cloud computing-based nuclear power plant PaaS platform resource quota monitoring method and system |
| CN114666333A (en) * | 2022-04-02 | 2022-06-24 | 国网江苏省电力有限公司信息通信分公司 | Control method for cloud computing resource scheduling problem based on multi-tenant theory |
Non-Patent Citations (2)
| Title |
|---|
| 何子龙等: "TenantCPUMan:基于负载分析的多租户动态CPU资源调整", 《计算机应用与软件》 * |
| 谢文舟等: "基于Kubernetes负载特征的资源预测模型研究", 《网络安全技术与应用》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114978998A (en) * | 2021-02-26 | 2022-08-30 | 中移(苏州)软件技术有限公司 | Flow control method, device, terminal and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115248734B (en) | 2022-12-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101938416B (en) | Cloud computing resource scheduling method based on dynamic reconfiguration virtual resources | |
| CN107038069B (en) | Dynamic label matching DLMS scheduling method under Hadoop platform | |
| CN110888714B (en) | Scheduling method, scheduling device and computer readable storage medium for containers | |
| WO2018120991A1 (en) | Resource scheduling method and device | |
| WO2018113472A1 (en) | Method for scheduling resource, and server | |
| CN109582448B (en) | Criticality and timeliness oriented edge calculation task scheduling method | |
| CN109783225B (en) | Tenant priority management method and system of multi-tenant big data platform | |
| CN111861412B (en) | Completion time optimization-oriented scientific workflow scheduling method and system | |
| CN111225050B (en) | Cloud computing resource allocation method and device | |
| CN114787830A (en) | Machine learning workload orchestration in heterogeneous clusters | |
| CN115248734B (en) | Private cloud multi-tenant resource quota self-adaptive adjustment method and device | |
| WO2022001086A1 (en) | Efficient gpu resource allocation optimization method and system | |
| CN105511950A (en) | Dispatching management method for task queue priority of large data set | |
| CN113867959A (en) | Training task resource scheduling method, device, equipment and medium | |
| CN111694669A (en) | Task processing method and device | |
| CN109960591A (en) | A method of the cloud application resource dynamic dispatching occupied towards tenant's resource | |
| CN107070965B (en) | Multi-workflow resource supply method under virtualized container resource | |
| CN107168805A (en) | A kind of resource regulating method based on virtual machine | |
| CN107423134B (en) | Dynamic resource scheduling method for large-scale computing cluster | |
| CN110825212B (en) | Energy-saving scheduling method and device and computer storage medium | |
| CN109189581B (en) | Job scheduling method and device | |
| CN110347477B (en) | Service self-adaptive deployment method and device in cloud environment | |
| CN115562841B (en) | Cloud video service self-adaptive resource scheduling system and method | |
| Li et al. | On scheduling of high-throughput scientific workflows under budget constraints in multi-cloud environments | |
| CN110457130B (en) | Distributed resource elastic scheduling model, method, 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |