SHARING WITH PERFORMANCE ISOLATION BETWEEN TENANTS IN A SOFTWARE-AS-A SERVICE SYSTEM

• US 20120011518A1
• Filed: 07/08/2010
• Published: 01/12/2012
• Est. Priority Date: 07/08/2010
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Duan, Ning, Natarajan, Ramesh, Ghoting, Amol
• Granted Patent: US 8,539,078 B2
• Time in Patent Office: Days
• US Class Current: 718/104
• CPC Class Codes: G06F 9/505 considering the load
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources

Abstract

An apparatus hosting a multi-tenant software-as-a-service (SaaS) system maximizes resource sharing capability of the SaaS system. The apparatus receives service requests from multiple users belonging to different tenants of the multi-tenant SaaS system. The apparatus partitions the resources in the SaaS system into different resource groups. Each resource group handles a category of the service requests. The apparatus estimates costs of the service requests of the users. The apparatus dispatches service requests to resource groups according to the estimated costs, whereby the resources are shared, among the users, without impacting each other.

Claims

1. A method for real-time sharing of resources in a multi-tenant software-as-a-service (SaaS) system, the method comprising:

receiving service requests from multiple users belonging to different tenants of the multi-tenant SaaS system;

partitioning the resources in the SaaS system into different resource groups, each resource group handling a category of the service requests;

estimating costs of the service requests; and

dispatching the service requests to the resource groups according to the estimated costs, whereby the resources are shared, among the users, without impacting each other.

2. The method according to claim 1, further comprising:

analyzing the service requests to acquire user IDs of the users, tenant IDs of the tenants, and requested functions of the service requests.

3. The method according to claim 1, further comprising:

isolating a mutual performance impact between the different tenants.

4. The method according to claim 3, wherein a resource group handles heavyweight service requests and another resource group handles regular service requests.

5. The method according to claim 3, wherein the estimation includes steps of:

segmenting tenants according to tenants'"'"' characteristics;

modeling a variability of the cost estimation for each segment of the tenants;

modeling a variability of the cost estimation for each tenant within each segment; and

modeling a variability of the cost estimation for each user within each tenant.

6. The method according to claim 5, wherein the modelings include at least one regression technique.

7. The method according to claim 6, further comprising:

using a modeling based on the regression technique with model parameters obtained by maximum likelihood estimation or Bayesian estimation methods.

8. An apparatus hosting a multi-tenant software-as-a-service (SaaS) system to maximize resource sharing capability of the SaaS system, the apparatus comprising:

a memory device; and

a processor being connected to the memory device,

wherein the processor performs steps of;

receiving service requests from multiple users belonging to different tenants of the multi-tenant SaaS system;

partitioning resources in the SaaS system into different resource groups, each resource group handling a category of the service requests;

estimating costs of the service requests; and

dispatching the service requests to the resource groups according to the estimated costs, whereby the resources are shared, among the users, without impacting each other.

9. The apparatus according to claim 8, wherein the processor further performs a step of:

analyzing the service requests to acquire user IDs of the users, tenant IDs of the tenants, and requested functions of the service requests.

10. The apparatus according to claim 8, wherein the processor further performs a step of:

isolating a mutual-performance impact between the different tenants.

11. The apparatus according to claim 10, wherein a resource group handles heavyweight service requests and another resource group handles regular service requests.

12. The apparatus according to claim 8, wherein the estimation includes steps of:

segmenting tenants according to tenants'"'"' characteristics;

modeling a variability of the cost estimation for each segment of the tenants;

modeling a variability of the cost estimation for each tenant within each segment;

modeling a variability of the cost estimation for each user within each tenant;

13. The apparatus according to claim 12, wherein the modelings include at least one regression technique.

14. The apparatus according to claim 13, wherein the processor further performs a step of:

using a modeling based on the regression technique with model parameters obtained by maximum likelihood estimation or Bayesian estimation methods.

15. A computer program product for sharing resources in a multi-tenant software-as-a-service (SaaS) system, the computer program product comprising a storage medium readable by a processing circuit and storing instructions run by the processing circuit for performing a method, the method comprising:

receiving service requests from multiple users belonging to different tenants of the multi-tenant SaaS system;

partitioning resources in the SaaS system into different resource groups, each resource group handling a category of the service requests;

estimating costs of the service requests; and

dispatching the service requests to the resource groups according to the estimated costs, whereby the resources are shared, among the users, without impacting each other.

16. The computer program product according to claim 15, wherein the method further comprises:

analyzing the service requests to acquire user IDs of the users, tenant IDs of the tenants, and requested functions of the service requests.

17. The computer program product according to claim 15, wherein the method further comprise:

isolating a mutual performance impact the different tenants.

18. The computer program product according to claim 17, wherein a resource group handles heavyweight service requests and another resource group handles regular service requests.

19. The computer program product according to claim 17, wherein the estimation includes steps of:

segmenting tenants according to tenants'"'"' characteristics;

modeling a variability of the cost estimation for each segment of the tenants;

modeling a variability of the cost estimation for each tenant within each segment; and

modeling a variability of the cost estimation for each user within each tenant.

20. The computer program product according to claim 19, wherein the modelings include at least one regression technique.

21. The computer program product according to claim 20, wherein the method further comprises:

using a modeling based on the regression technique with model parameters obtained by maximum likelihood estimation or Bayesian estimation methods.