CN117851035A - Resource cluster scheduling method based on multi-tenant sharing, computer and storage medium - Google Patents
Resource cluster scheduling method based on multi-tenant sharing, computer and storage medium Download PDFInfo
- Publication number
- CN117851035A CN117851035A CN202311702652.5A CN202311702652A CN117851035A CN 117851035 A CN117851035 A CN 117851035A CN 202311702652 A CN202311702652 A CN 202311702652A CN 117851035 A CN117851035 A CN 117851035A
- Authority
- CN
- China
- Prior art keywords
- tenant
- resource
- quota
- resources
- scheduling
- 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
- 238000000034 method Methods 0.000 title claims abstract description 41
- 235000003642 hunger Nutrition 0.000 claims abstract description 49
- 230000037351 starvation Effects 0.000 claims abstract description 30
- 238000013468 resource allocation Methods 0.000 claims description 11
- 238000004891 communication Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000002035 prolonged effect Effects 0.000 description 2
- 208000003443 Unconsciousness Diseases 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The invention discloses a resource cluster scheduling method based on multi-tenant sharing, a computer and a storage medium, and relates to the technical field of computers, wherein the method comprises the following steps: determining the starvation degree of each tenant resource scheduling request; determining a first tenant according to the starvation degree of each tenant resource scheduling request; allocating the requested amount of resources to the first tenant and the second tenant; the second tenant is a tenant with a resource scheduling request starvation degree smaller than that of the first tenant; allocating the residual resources to a third tenant; the remaining resources are the resources remaining after the total resources are removed from the resources allocated to the first tenant and the second tenant, and the third tenant is the tenant with the starvation degree of the resource scheduling request being greater than that of the first tenant. Through the method, the invention ensures the strong fairness and high scheduling performance of resource scheduling.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a method for scheduling resource clusters based on multi-tenant sharing, a computer, and a storage medium.
Background
With the increasing size of data processing, traditional stand-alone computing modes have failed to meet the increasing demands for information services. Clusters (clusters) are a group of independent computers interconnected through a high-speed network that form a computing group and can be managed in a unified manner. The cluster can realize high operation speed, complete calculation of large operation amount, has higher response capability, and can reduce the overall operation and maintenance cost, so that more and more applications are obtained.
Because of the relatively high overall cost of the clusters, particularly large-scale clusters, it is impractical to deploy a separate cluster for each application. In other cases, the application has own resource utilization peaks and troughs, namely, the application cannot guarantee higher cluster resource utilization rate at any time, and the cluster sharing mode can save cost and simplify management.
While shared clusters can save cost, they face the problem of fair scheduling, which is more challenging to achieve while guaranteeing fairness, especially in very large-scale clusters. The fair scheduling is that a scheduler guarantees that the ratio of resource allocation to quota among all tenants is close at a certain moment, and if the scheduler needs to guarantee fairness among the tenants every time of scheduling, the application is defined as strong fairness; if the scheduler only guarantees fairness between tenants at certain times, the present application is defined as weak fairness or final fairness.
Weak fairness sacrifices partial fairness but improves scheduling performance. Such as batch scheduling, that is, resources are mainly used by a tenant at a certain moment, but from the perspective of time stretch, the resource allocation rate of all tenants is close to the quota ratio between tenants.
The strong fairness indicates that the scheduler needs to find the optimal solution of all the tenant allocation resources every time the scheduler performs a scheduling decision, which tends to result in a decrease in scheduling performance, but absolute fairness is guaranteed. For example, a request queue is designed for each tenant, a request to be scheduled is extracted from all tenant queues each time to calculate the resource allocation and quota ratio of the tenant, and the tenant request scheduling with highest hunger level is selected. This is a very mainstream design, but it should be noted that once the request schedule of a certain tenant is selected, the time from this time to the next scheduling decision may be unfair to other tenants, and although the time is prolonged and eventually each tenant is relatively fair, it is not fair to all tenants from every point of view.
The existing multi-tenant cluster scheduling system has the following problems:
poor fairness: adopting a simple first-come first-get or sorting strategy according to priority, so that some malicious tenants occupy a large amount of cluster resources;
good fairness but low performance: the method has better fairness, but the overall scheduling throughput rate is not high, and the method is suitable for online service scheduling, but becomes unconscious for high-concurrency offline service.
From the above, how to ensure strong fairness and high scheduling becomes a problem to be solved in resource scheduling.
Disclosure of Invention
The purpose of the application is to provide a resource cluster scheduling method, a computer and a storage medium based on multi-tenant sharing, so as to ensure the strong fairness and high scheduling performance of resource scheduling.
In order to achieve the above object, the present invention provides the following solutions:
in one aspect, the present application provides a method for scheduling a resource cluster based on multi-tenant sharing, where the method includes:
at each resource scheduling period:
determining the starvation degree of each tenant resource scheduling request, wherein the starvation degree represents the degree to which the tenant resource scheduling request is not satisfied;
determining a first tenant according to the starvation degree of each tenant resource scheduling request;
allocating the requested amount of resources to the first tenant and the second tenant; the second tenant is a tenant with a resource scheduling request starvation degree smaller than that of the first tenant;
allocating the residual resources to a third tenant; and the third tenant is a tenant with a higher starvation degree of the resource scheduling request than the first tenant.
Optionally, the determining the first tenant according to the starvation degree of each tenant resource scheduling request specifically includes:
determining a target tenant meeting the target condition; wherein the target condition is: after the requested resource amount is distributed to all tenants with the hunger degree smaller than or equal to the target tenant, the resource request amount still cannot be satisfied after all tenants with the hunger degree larger than the target tenant distribute the residual resources according to the quota weight; the target tenant may be determined by a dichotomy, that is, based on the target condition, the target tenant is determined by a dichotomy.
And determining the first tenant from the target tenants.
Optionally, after allocating the remaining resources to the third tenant, the method further comprises:
determining the resource request quantity of the fourth tenant at the current moment; the fourth tenant is a tenant with a resource request quantity smaller than the self quota guarantee at the last moment;
for each fourth tenant:
if the resource request quantity at the current moment is larger than the resource request quantity at the last moment, performing quota allocation on the newly added resource request quantity in the self quota guarantee range based on the priority allocation; the priority allocation right is right to preempt the remaining quota, and the remaining quota is an excessive scheduling quota formed by that the resource request amount of the tenant at the current moment is smaller than the resource allocation amount at the last moment.
For the above-mentioned resource allocation decision of the fourth tenant, the following is considered: in each resource scheduling period, only one resource scheduling calculation is performed, but in one resource scheduling period, the resource request amount of each tenant is changed. Accordingly, the present application provides the above-described resource allocation decision for the amount of resource requests that vary within one resource scheduling period.
In another aspect, the present application provides a computer device, where the computer includes a processor and a memory, where at least one instruction is stored in the memory, where the instruction is loaded and executed by the processor to implement the resource cluster scheduling method described above.
In another aspect, the present application further provides a storage medium, where at least one instruction is stored, where the instruction is loaded and executed by a processor to implement the resource cluster scheduling method described above.
According to a specific embodiment provided by the application, the application discloses the following technical effects: according to the multi-tenant sharing-based resource cluster scheduling method, the computer and the storage medium, the first tenant is determined according to the starvation degree of the tenants, the requested resource quantity is distributed to the first tenant and the second tenant with the starvation degree smaller than that of the first tenant, and the residual resources are distributed to the third tenant with the starvation degree larger than that of the first tenant, so that the strong fairness and the high scheduling performance of resource scheduling are guaranteed.
In addition, the method determines the target tenant based on the principle that after the requested resource quantity is distributed to all tenants with the hunger degree smaller than or equal to the target tenant, the resource request quantity of all tenants with the hunger degree larger than the target tenant still cannot be met after the residual resources are distributed according to the quota weight, and selects the tenant with the largest hunger degree from the target tenant as the first tenant, so that the strong fairness and the high scheduling performance of the resource scheduling are further ensured.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flow chart of a resource cluster scheduling method based on multi-tenant sharing according to an embodiment of the present application;
fig. 2 is a schematic structural diagram of a computer device in an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention aims to provide a resource cluster scheduling method, a computer and a storage medium based on multi-tenant sharing, so as to ensure the strong fairness and high scheduling performance of resource scheduling.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
The technical terms referred to in this application are explained: the quota of each tenant consists of a quota total amount and a quota guarantee:
quota Weight (Weight of quote, hereinafter WQ): the quota weight represents the weight occupied by the upper limit of the resources which can be used by the tenant, the quota weight has no dimension, and the quota weight takes on any natural number. The quota weights can be used to calculate which tenant can allocate more resources, e.g., the scheduler counts the sum of the quota weights of all tenants, and calculates the duty ratio of each tenant quota weight, and the tenant with the high weight duty ratio can theoretically allocate more resources.
Quota guarantee (hereinafter, simply referred to as GQ): the quota guarantee is the amount of resources which can be obtained by the tenant from the cluster under any condition, and can avoid the condition that some low-priority and quota-weighted small user requests are starved. The sum of quota guarantee values of all tenants cannot be larger than the total amount of cluster resources.
The Quota guarantee is a basic property of the tenant, and in any case, the scheduler allocates a resource amount not less than the Quota guarantee as long as the tenant applies for a resource (RQ for short hereinafter). It should be noted that if the user applies that the resource amount is smaller than the quota guarantee, the remaining quota guarantee resource amount may be used by other tenants, so the scheduler can ensure that the tenant obtains the resource amount not smaller than Min (GQ, RQ) at each moment.
The definition of strong fairness in the application refers to that the decision of each scheduling period of the cluster resource scheduler is fair for all tenants, and is mainly characterized in that each scheduling period must traverse the resource request amount and quota amount of all tenants, and the resource amount which can be allocated by each tenant is calculated fairly. To achieve strong fairness, the present application divides cluster resource scheduling into two phases:
inter-tenant resource allocation: the scheduler needs to calculate the amount of resources to be allocated to each tenant, and the specific method is as follows:
when RQ < = GQ, the resources allocated by the tenant are RQ;
when RQ > GQ, the tenant may allocate resources as gq+total cluster resources, and the remaining amount of the total resources of all tenants Min (RQ, GQ) is calculated according to the ratio of the tenant quota weights, where the following formula is shown:
the above formula can calculate the amount of resources that each tenant can allocate, but the tenant application amount of resources is uncertain. If the amount of resources applied by the tenant is smaller than the amount of allocated resources, part of the resources are wasted, and if the amount of resources applied by the tenant is larger than the amount of allocated resources, unused resources of other tenants cannot be reused, which sacrifices the sharing property of cluster resources.
The application further upgrades the formula, provides a starvation degree concept, the starvation degree expresses the satisfaction degree of a tenant resource request, and the larger the starvation degree is, the more resources are needed by the tenant (the tenant can be understood to have more queued requests).
Based on the above concept, the resource cluster scheduling method provided by the application is shown in fig. 1, and includes the following steps:
at each resource scheduling period:
step 101: and determining the starvation degree of each tenant resource scheduling request, wherein the starvation degree represents the degree to which the tenant resource scheduling request is not satisfied.
Step 102: and determining the first tenant according to the starvation degree of each tenant resource scheduling request.
Step 103: allocating the requested amount of resources to the first tenant and the second tenant; the second tenant is a tenant with a resource scheduling request starvation degree smaller than that of the first tenant.
Step 104: allocating the residual resources to a third tenant; for example, the remaining resources may be allocated to each third tenant according to the quota weight of each third tenant. The remaining resources are the resources remaining after the total resources are removed from the resources allocated to the first tenant and the second tenant, and the third tenant is the tenant with the starvation degree of the resource scheduling request being greater than that of the first tenant.
Wherein step 103 and step 104 have no temporal sequence.
In an example embodiment, the determinants of hunger include the quota request amount, quota guarantees, and quota weights of the tenant. Specifically, the formula for calculating the hunger level may be as follows:
where h represents starvation, RQ represents resource request amount, GQ represents quota guarantee, and WQ represents quota weight.
In an example embodiment, step 102 is specifically as follows:
and determining target tenants meeting the target conditions, and determining a first tenant from the target tenants. Wherein, the target condition is: after the requested resource amount is distributed to all tenants with the hunger degree smaller than or equal to that of the target tenant, the resource request amount still cannot be satisfied after all tenants with the hunger degree larger than that of the target tenant distribute the residual resources according to the quota weight. The first tenant may be the most hungry tenant of the target tenants.
The specific operation can be as follows:
sequencing hunger of all tenants from small to large, and finding that tenant x meets the following conditions:
1. how many resources are allocated for all tenants with hunger less than or equal to tenant x apply for how many resources;
2. all tenants with hunger degree larger than tenant x still cannot meet the request amount of the tenants after the rest resources are allocated according to the weight ratio;
tenant x is the tenant with the highest hunger satisfying the conditions 1 and 2.
Tenant x, i.e. the first tenant in the above, can be located quickly by dichotomy.
The method realizes fairness, sharing and high efficiency:
fairness: resources outside of all tenants Min (RQ, GQ) can be assigned per tenant quota weight.
Shareability: the quota that the tenant cannot use currently can be used by other tenants, and the quota is distributed according to fairness.
High efficiency: and calculating the resources which can be allocated by all tenants with lower time complexity.
After the tenant allocates the resource, the request in the tenant can be dispatched, and the request of the tenant can be dispatched in parallel among the tenants because the tenants are not coupled. By using the method, the resources which can be allocated by each dispatching tenant are preferentially calculated, and the complex flow that each dispatching period is required to calculate the optimal solution among a plurality of tenants is avoided. And as the request scheduling between tenants is not coupled, the concurrent scheduling can be realized, and the scheduling efficiency is provided.
Since the resource scheduling calculation is performed only once in each resource scheduling period, the resource request amount of each tenant is changed in one resource scheduling period. In order to further ensure the strong fairness and high scheduling of the resource scheduling, the application provides that the resource quantity except the GQ allocated by all tenants is shared, and other tenants do not have preemption; resources within the GQ are exclusive to the tenant, having preemption. In particular, in the exemplary embodiment, for the resource request amount that changes in one resource scheduling period, after step 103 and step 104, the following operations are further included:
determining the resource request quantity of the fourth tenant at the current moment; the fourth tenant is the tenant with the resource request quantity smaller than the self quota guarantee at the last moment;
for each fourth tenant:
if the resource request quantity at the current moment is larger than the resource request quantity at the last moment, performing quota allocation on the newly added resource request quantity in the self quota guarantee range based on the priority allocation; the priority allocation right is the right of preempting the residual quota, and the residual quota is an excessive scheduling quota formed by the fact that the resource request quantity of the tenant at the current moment is smaller than the resource allocation quantity at the last moment.
Such as: GQ of tenant a is 5, and initially because the resource that is not much allocated by tenant a is 2, then the 3 guarantee quotas of tenant guarantee a are not shared by other tenants for use by themselves. When the request amount of the tenant A reaches 5, the amount of resources allocated to the tenant A by the scheduler is also 5, at this time, the scheduler can schedule 3 requests of the tenant A again, and the 3 requests of the resource can preempt the scheduled requests of the reduced resource amounts of other tenants. When the request amount of the tenant A reaches 7, the resource allocation amount of the tenant A is 6, and the newly added 1 resource amount quota is the shared resource quota, so that the preemption is not available, and other tenants cannot be preempted to reduce the requests of which the quota is scheduled.
Compared with the prior art, the application has the following advantages:
strong fairness: the amount of resources available between tenants at any time (with the scheduling period as the sampling granularity) is fair. In the existing mainstream method, one scheduling queue is used for each tenant, one/batch of request scheduling is selected for each scheduling period, the time is prolonged, and finally each tenant is fair, but each scheduling period is unfair for all tenants.
Shareability: the guaranteeing quota can not be used by the tenant, and idle resources of the cluster can be distributed to other tenants for use according to the weight equal ratio, so that the utilization rate of the cluster resources is improved.
Efficient scheduling performance: the method of utilizing the two division of the hunger index is introduced to rapidly calculate the resource allocation amount of each tenant, and parallel scheduling requests are carried out among the tenants. There are also some existing methods for calculating optimal results according to tenant request amount, quota amount and resource total amount, for example, the calculated amount is far higher than that of the method mentioned herein.
The method and the device can be applied to the technical field of cluster scheduling with multi-tenant demands, such as container arrangement, big data, cloud computing and the like. The method is particularly suitable for cloud computing and big data scenes, resources which each tenant should allocate are calculated fairly in each scheduling, and then scheduling requests are concurrent among the tenants, so that a fair and efficient scheduling strategy is realized.
The present application further provides a computer device, fig. 2 illustrates a physical structure schematic diagram of the computer device, and as shown in fig. 2, the device may include: processor 610, communication interface (Communications Interface) 620, memory 630, and communication bus 640, wherein processor 610, communication interface 620, and memory 630 communicate with each other via communication bus 640. The processor 610 may invoke logic instructions in the memory 630 to implement the resource cluster scheduling method described above.
Further, the logic instructions in the memory 630 may be implemented in the form of software functional units and stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method of the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, randomAccess Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the above-described resource cluster scheduling method.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. The resource cluster scheduling method based on multi-tenant sharing is characterized by comprising the following steps:
at each resource scheduling period:
determining the starvation degree of each tenant resource scheduling request, wherein the starvation degree represents the degree to which the tenant resource scheduling request is not satisfied;
determining a first tenant according to the starvation degree of each tenant resource scheduling request;
allocating the requested amount of resources to the first tenant and the second tenant; the second tenant is a tenant with a resource scheduling request starvation degree smaller than that of the first tenant;
allocating the residual resources to a third tenant; and the residual resources are the resources which are remained after the total resources are removed from the resources distributed to the first tenant and the second tenant, and the third tenant is the tenant with the starvation degree of the resource scheduling request being greater than that of the first tenant.
2. The method for scheduling resource clusters based on multi-tenant sharing according to claim 1, wherein the determining the first tenant according to the starvation degree of each tenant resource scheduling request specifically comprises:
determining a target tenant meeting the target condition; wherein the target condition is: after the requested resource amount is distributed to all tenants with the hunger degree smaller than or equal to the target tenant, the resource request amount still cannot be satisfied after all tenants with the hunger degree larger than the target tenant distribute the residual resources according to the quota weight;
and determining the first tenant from the target tenants.
3. The method for scheduling resource clusters based on multi-tenant sharing according to claim 2, wherein the determining the first tenant from the target tenants specifically comprises:
and determining the tenant with the largest hunger degree in the target tenants as the first tenant.
4. The multi-tenant sharing-based resource cluster scheduling method of claim 2, wherein the determining the target tenant that meets the target condition specifically includes:
and determining the target tenant by adopting a dichotomy based on the target condition.
5. The multi-tenant sharing-based resource cluster scheduling method of claim 1, wherein the hunger determinants include a quota request amount, a quota guarantee, and a quota weight of a tenant.
6. The multi-tenant shared-based resource cluster scheduling method of claim 5, wherein the starvation is calculated according to the following formula:
where h represents starvation, RQ represents resource request amount, GQ represents quota guarantee, and WQ represents quota weight.
7. The multi-tenant shared resource cluster scheduling method of claim 1, wherein the allocating the remaining resources to the third tenant specifically comprises:
and distributing the residual resources to each third tenant according to the quota weight of each third tenant.
8. The multi-tenant shared-based resource cluster scheduling method of claim 1, further comprising, after the allocating the remaining resources to the third tenant:
determining the resource request quantity of the fourth tenant at the current moment; the fourth tenant is a tenant with a resource request quantity smaller than the self quota guarantee at the last moment;
for each fourth tenant:
if the resource request quantity at the current moment is larger than the resource request quantity at the last moment, performing quota allocation on the newly added resource request quantity in the self quota guarantee range based on the priority allocation; the priority allocation right is right to preempt the remaining quota, and the remaining quota is an excessive scheduling quota formed by that the resource request amount of the tenant at the current moment is smaller than the resource allocation amount at the last moment.
9. A computer comprising a processor and a memory having at least one instruction stored therein, the instruction being loaded and executed by the processor to implement the resource cluster scheduling method of any one of claims 1 to 8.
10. A storage medium having stored therein at least one instruction that is loaded and executed by a processor to implement the resource cluster scheduling method of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311702652.5A CN117851035A (en) | 2023-12-12 | 2023-12-12 | Resource cluster scheduling method based on multi-tenant sharing, computer and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311702652.5A CN117851035A (en) | 2023-12-12 | 2023-12-12 | Resource cluster scheduling method based on multi-tenant sharing, computer and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117851035A true CN117851035A (en) | 2024-04-09 |
Family
ID=90527921
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311702652.5A Pending CN117851035A (en) | 2023-12-12 | 2023-12-12 | Resource cluster scheduling method based on multi-tenant sharing, computer and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117851035A (en) |
-
2023
- 2023-12-12 CN CN202311702652.5A patent/CN117851035A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108345501B (en) | Distributed resource scheduling method and system | |
| CN109564528B (en) | System and method for computing resource allocation in distributed computing | |
| WO2016078178A1 (en) | Virtual cpu scheduling method | |
| WO2018120991A1 (en) | Resource scheduling method and device | |
| CN106445675B (en) | B2B platform distributed application scheduling and resource allocation method | |
| CN109542608B (en) | Cloud simulation task scheduling method based on hybrid queuing network | |
| CN108123980B (en) | Resource scheduling method and system | |
| WO2022068697A1 (en) | Task scheduling method and apparatus | |
| US20130061233A1 (en) | Efficient method for the scheduling of work loads in a multi-core computing environment | |
| CN107203422B (en) | Job scheduling method for high-performance computing cloud platform | |
| CN109992418B (en) | SLA-aware resource priority scheduling method and system for multi-tenant big data platform | |
| CN105718316A (en) | Job scheduling method and apparatus | |
| CN114787830A (en) | Machine learning workload orchestration in heterogeneous clusters | |
| US8090903B2 (en) | Fair and dynamic disk input/output bandwidth distribution | |
| CN112749002A (en) | Method and device for dynamically managing cluster resources | |
| Jia et al. | A deadline constrained preemptive scheduler using queuing systems for multi-tenancy clouds | |
| CN112463044A (en) | Method and system for ensuring tail reading delay of server side of distributed storage system | |
| CN111221631A (en) | Task scheduling method, device and storage medium | |
| Pereira et al. | Increasing the efficiency of fog nodes through of priority-based load balancing | |
| CN112925616A (en) | Task allocation method and device, storage medium and electronic equipment | |
| CN112817726B (en) | Priority-based virtual machine grouping resource scheduling method in cloud environment | |
| CN107423134B (en) | Dynamic resource scheduling method for large-scale computing cluster | |
| Dubey et al. | QoS driven task scheduling in cloud computing | |
| Wu et al. | Abp scheduler: Speeding up service spread in docker swarm | |
| CN104731662B (en) | A kind of resource allocation methods of variable concurrent job |
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 |