Mechanism for Automatic Scaling of Application Resources in a Multi-Tenant Platform-as-a-Service (PaaS) Environment in a Cloud Computing System

• US 20140068611A1
• Filed: 09/06/2012
• Published: 03/06/2014
• Est. Priority Date: 09/06/2012
• Status: Active Grant

• Current Assignee: Red Hat, Inc. (International Business Machines Corporation)
• Original Assignee: Red Hat, Inc. (International Business Machines Corporation)
• Inventors: McGrath, Michael P., Hicks, Matthew
• Granted Patent: US 8,978,035 B2
• Time in Patent Office: Days
• US Class Current: 718/1
• CPC Class Codes: G06F 9/45533 Hypervisors; Virtual machine monitors
• CPC Class Codes: G06F 9/5072 Grid computing
• CPC Class Codes: G06F 9/5077 Logical partitioning of resources; Management or configuration of virtualized resources specific details on emulation or internal functioning of virtual machines G06F9/455
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system

Abstract

A mechanism for automatic scaling of application resources in a multi-tenant Platform-as-a-Service (PaaS) environment in a cloud computing system is disclosed. A method includes monitoring, by a load balancer executing on a computing device, an overall load measurement of containers associated with an application executed on virtual machines (VMs) in a multi-tenant Platform-as-a-Service (PaaS) system, wherein the overall load measurement is a summation of individual current loads at the containers, comparing, by the load balancer, the overall load measurement to an overall load threshold for the containers, and when the overall load measurement exceeds the overall load threshold, causing resources to be added to execute functionality of the application in the VMs of the multi-tenant PaaS system.

Claims

1. A computer-implemented method, comprising:

monitoring, by a load balancer executing on a computing device, an overall load measurement of containers associated with an application executed on virtual machines (VMs) in a multi-tenant Platform-as-a-Service (PaaS) system, wherein the overall load measurement is a summation of individual current loads at the containers;

comparing, by the load balancer, the overall load measurement to an overall load threshold for the containers; and

when the overall load measurement exceeds the overall load threshold, causing resources to be added to execute functionality of the application in the VMs of the multi-tenant PaaS system.

2. The computer-implemented method of claim 1, wherein each of the containers in the VMs comprise a resource-constrained process space of each VM to execute a component of the application, wherein the application is one of a plurality of applications executed in the by the VMs and having different owners.

3. The computer-implemented method of claim 1, wherein a load of the overall load measurement comprises a work request directed to the application, the work request comprising at least one of a network traffic request or a database request.

4. The computer-implemented method of claim 1, wherein the overall load threshold comprises a percentage of a maximum overall load capacity of the monitored containers.

5. The computer-implemented method of claim 4, wherein the maximum overall load capacity is a summation of a maximum total load capacity configured for each of the monitored containers.

6. The computer-implemented method of claim 5, further comprising receiving, from a resource control module of the multi-tenant PaaS system, data comprising the maximum total load capacity configured for each of the monitored containers.

7. The computer-implemented method of claim 1, wherein the causing resources to be added further comprises requesting a resource control module of the multi-tenant PaaS system to provision one or more containers and associated cartridges for the application in one or more of the VMs of the multi-tenant PaaS system.

8. The computer-implemented method of claim 7, wherein the resource control module operates on a broker layer of the multi-tenant PaaS system, wherein the multi-tenant PaaS system is implemented in a cloud computing system.

9. The computer-implemented method of claim 1, wherein the application executes in one or more of the VMs distributed across one or more districts of the multi-tenant PaaS system, wherein the one or more districts comprise a collection of the VMs that is given a dedicated set of Internet Protocol (IP) addresses, ports, and security labels to share when executing the containers within the VMs of the district.

10. A system, comprising:

a memory;

a processing device communicably coupled to the memory; and

a resource control module executable from the memory by the processing device, the resource control module operable to;

provide capacity load configuration data to a load balancer associated with an application executed on virtual machines (VMs) in a multi-tenant Platform-as-a-Service (PaaS) system of the resource control module;

initialize resources for the application in response to an add request sent by the load balancer, the add request sent when an overall load measurement of containers of the application exceeds a maximum overall load threshold for the containers; and

remove resources from the application in response to a remove request sent by the load balancer, the remove request sent when the overall load measurement of the containers of the application falls below a minimum overall load threshold for the containers and when thrashing protections for the overall load measurement are satisfied.

11. The system of claim 10, wherein each of the containers in the VMs comprise a resource-constrained process space of each VM to execute a component of the application, wherein the application is one of a plurality of applications executed in the by the VMs and having different owners.

12. The system of claim 10, wherein the thrashing protections comprise at least one of a number of contiguous previous checks that the overall load measurement fell below the minimum overall load threshold, a time interval that resources have not been added to the application, or a time interval that resources have not been removed from the application.

13. The system of claim 10, wherein a load of the overall load measurement comprises a work request directed to the application, the work request comprising at least one of a network traffic request or a database request.

14. The system of claim 10, wherein the overall load measurement is a summation of individual current load at each monitored container.

15. The system of claim 10, wherein the resource control module operates on a broker layer of the multi-tenant PaaS system, and wherein the multi-tenant PaaS system is implemented in a cloud computing system.

16. A non-transitory machine-readable storage medium including data that, when accessed by a machine, cause the machine to perform operations comprising:

monitoring, by a load balancer executing on a computing device, an overall load measurement of containers associated with an application executed on virtual machines (VMs) in a multi-tenant Platform-as-a-Service (PaaS) system, wherein the overall load measurement is a summation of individual current loads at the containers;

comparing, by the load balancer, the overall load measurement to an overall load threshold for the containers;

when the overall load measurement falls below the overall load threshold, determining whether thrashing protections are satisfied for the overall load measurement; and

when the thrashing protections are satisfied, causing resources to be removed from the application in the VMs of the multi-tenant PaaS system.

17. The non-transitory machine-readable storage medium of claim 16, wherein each of the containers in the VMs comprise a resource-constrained process space of each VM to execute a component of the application, wherein the application is one of a plurality of applications executed in the by the VMs and having different owners.

18. The non-transitory machine-readable storage medium of claim 16, wherein the thrashing protections comprise at least one of a number of contiguous previous checks that the overall load measurement fell below the overall load threshold, a time interval that resources have not been added to the application, or a time interval that resources have not been removed from the application.

19. The non-transitory machine-readable storage medium of claim 16, wherein causing resources to be removed from the application further comprises requesting a resource control module of the multi-tenant PaaS system to remove one or more containers and associated cartridges from one or more of the VMs executing the application in the multi-tenant PaaS system.

20. The non-transitory machine-readable storage medium of claim 16, wherein the resource control module operates on a broker layer of a Platform-as-a-Service (PaaS) system that includes the VM, the PaaS system implemented in a cloud computing system.