Application-based elastic resource provisioning in disaggregated computing systems

• US 10,063,493 B2
• Filed: 05/16/2016
• Issued: 08/28/2018
• Est. Priority Date: 05/16/2016
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Li, Min, Bivens, John A., Mahindru, Ruchi, Ramasamy, HariGovind V., Ruan, Yaoping, Salapura, Valentina, Schenfeld, Eugen
• Primary Examiner(s): Tang, Karen
• Application Number:
  US15/155,560
• Publication Number:
  US 20170331763A1
• Time in Patent Office: 834 Days
• Field of Search: 709224, 709231, 709232
• CPC Class Codes: H04L 41/0893 Assignment of logical groups to network elements
• CPC Class Codes: H04L 41/5025 by proactively reacting to service quality change, e.g. by reconfiguration after service quality degradation or upgrade
• CPC Class Codes: H04L 41/5096 wherein the managed service relates to distributed or central networked applications
• CPC Class Codes: H04L 47/2425 for supporting services specification, e.g. SLA
• CPC Class Codes: H04L 47/803 Application aware
• CPC Class Codes: H04L 47/805 QOS or priority aware
• CPC Class Codes: H04L 47/83 based on usage prediction
• CPC Class Codes: H04L 67/1012 based on compliance of requirements or conditions with available server resources
• CPC Class Codes: H04L 67/1023 based on a hash applied to IP addresses or costs

Abstract

Various embodiments for elastic resource provisioning in a disaggregated cloud computing environment, by a processor device, are provided. Respective members of pools of hardware resources within the disaggregated cloud computing environment are provisioned to a tenant according to an application-level service level agreement (SLA). Upon detecting a potential violation of the application-level SLA, additional respective members of the pools of hardware resources are provisioned on a component level to the tenant to avoid a violation of the SLA by one of a scale-up process and a scale-out process.

Claims

1. A method for elastic resource provisioning in a disaggregated cloud computing environment, by a processor device, comprising:

provisioning respective members of pools of hardware resources within the disaggregated cloud computing environment to a tenant according to an application-level service level agreement (SLA); and

upon detecting a potential violation of the application-level SLA, provisioning additional respective members of the pools of hardware resources on a component level to the tenant to avoid a violation of the SLA by each of a scale-up process and a scale-out process based upon workload characteristics to ensure a tenant receives a service-level as specified within the SLA while retaining an overall cost and performance efficiency model utilized in the disaggregated cloud computing environment;

wherein;

the scale-up process is used for scale-up friendly workloads, and comprises provisioning the additional respective members of the pools of hardware resources according to a quantity of the respective members of the pools of hardware resources; and

the scale-out process is used scale-out friendly workloads, and comprises provisioning the additional respective members of the pools of hardware resources according to a CPU interconnect bandwidth and off-chip memory bandwidth between the respective members of the pool of hardware resources.

2. The method of claim 1,

further including de-provisioning respective members of the pools of hardware resources to the tenant according to one of a scale-down process and a scale-in process, based upon the workload characteristics.

3. The method of claim 2, wherein the scale-down process and the scale-in process are used upon detecting an underutilization of the provisioned respective members of the pools of hardware resources, and each comprise prompting the tenant to de-provision certain respective members of the pools of hardware resources to reduce an operating cost.

4. The method of claim 2, wherein detecting the potential violation of the application-level SLA comprises monitoring the workload characteristics of the tenant, the tenant specifying the application-level SLA upon request to perform the workload.

5. The method of claim 1, wherein the provisioning is based upon history information maintained in an SLA database and a predicted performance of a given application from the tenant.

6. The method of claim 1, wherein the respective members of the pools of hardware resources include at least processors, memory devices, graphical processing units (GPUs), and accelerators;

and

provisioning the additional respective members of the pools of hardware resources on the component level to the tenant comprises allocating individual ones of the processors, the memory devices, the GPUs, and the accelerators to the tenant.

7. A system for elastic resource provisioning in a disaggregated cloud computing environment, the system comprising:

at least one processor device, wherein the at least one processor device;

provisions respective members of pools of hardware resources within the disaggregated cloud computing environment to a tenant according to an application-level service level agreement (SLA); and

upon detecting a potential violation of the application-level SLA, provisions additional respective members of the pools of hardware resources on a component level to the tenant to avoid a violation of the SLA by each of a scale-up process and a scale-out process based upon workload characteristics to ensure a tenant receives a service-level as specified within the SLA while retaining an overall cost and performance efficiency model utilized in the disaggregated cloud computing environment;

wherein;

the scale-up process is used for scale-up friendly workloads, and comprises provisioning the additional respective members of the pools of hardware resources according to a quantity of the respective members of the pools of hardware resources; and

the scale-out process is used scale-out friendly workloads, and comprises provisioning the additional respective members of the pools of hardware resources according to a CPU interconnect bandwidth and off-chip memory bandwidth between the respective members of the pool of hardware resources.

8. The system of claim 7,

wherein the at least one processor device de-provisions respective members of the pools of hardware resources to the tenant according to one of a scale-down process and a scale-in process, based upon the workload characteristics.

9. The system of claim 8, wherein the scale-down process and the scale-in process are used upon detecting an underutilization of the provisioned respective members of the pools of hardware resources, and each comprise prompting the tenant to de-provision certain respective members of the pools of hardware resources to reduce an operating cost.

10. The system of claim 8, wherein detecting the potential violation of the application-level SLA comprises monitoring the workload characteristics of the tenant, the tenant specifying the application-level SLA upon request to perform the workload.

11. The system of claim 7, wherein the provisioning is based upon history information maintained in an SLA database and a predicted performance of a given application from the tenant.

12. The system of claim 7, wherein the respective members of the pools of hardware resources include at least processors, memory devices, graphical processing units (GPUs), and accelerators;

and

provisioning the additional respective members of the pools of hardware resources on the component level to the tenant comprises allocating individual ones of the processors, the memory devices, the GPUs, and the accelerators to the tenant.

13. A computer program product for elastic resource provisioning in a disaggregated cloud computing environment by at least one processor device, the computer program product embodied on a non-transitory computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions comprising:

an executable portion that provisions respective members of pools of hardware resources within the disaggregated cloud computing environment to a tenant according to an application-level service level agreement (SLA); and

an executable portion that, upon detecting a potential violation of the application-level SLA, provisions additional respective members of the pools of hardware resources on a component level to the tenant to avoid a violation of the SLA by each of a scale-up process and a scale-out process based upon workload characteristics to ensure a tenant receives a service-level as specified within the SLA while retaining an overall cost and performance efficiency model utilized in the disaggregated cloud computing environment;

wherein;

the scale-up process is used for scale-up friendly workloads, and comprises provisioning the additional respective members of the pools of hardware resources according to a quantity of the respective members of the pools of hardware resources; and

the scale-out process is used scale-out friendly workloads, and comprises provisioning the additional respective members of the pools of hardware resources according to a CPU interconnect bandwidth and off-chip memory bandwidth between the respective members of the pool of hardware resources.

14. The computer program product of claim 13,

further including an executable portion that de-provisions respective members of the pools of hardware resources to the tenant according to one of a scale-down process and a scale-in process, based upon the workload characteristics.

15. The computer program product of claim 14, wherein the scale-down process and the scale-in process are used upon detecting an underutilization of the provisioned respective members of the pools of hardware resources, and each comprise prompting the tenant to de-provision certain respective members of the pools of hardware resources to reduce an operating cost.

16. The computer program product of claim 14, wherein detecting the potential violation of the application-level SLA comprises monitoring the workload characteristics of the tenant, the tenant specifying the application-level SLA upon request to perform the workload.

17. The computer program product of claim 13, wherein the provisioning is based upon history information maintained in an SLA database and a predicted performance of a given application from the tenant.

18. The computer program product of claim 13, wherein the respective members of the pools of hardware resources include at least processors, memory devices, graphical processing units (GPUs), and accelerators;

and

provisioning the additional respective members of the pools of hardware resources on the component level to the tenant comprises allocating individual ones of the processors, the memory devices, the GPUs, and the accelerators to the tenant.