Workload balancing using dynamically allocated virtual servers

• US 7,080,378 B1
• Filed: 05/17/2002
• Issued: 07/18/2006
• Est. Priority Date: 05/17/2002
• Status: Expired due to Term

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Storage Technology Corp (Oracle Corporation)
• Inventors: Milligan, Charles A., Hodge, Leslie K., Noland, Thomas Nelson
• Primary Examiner(s): Bullock, Jr., Lewis A.
• Assistant Examiner(s): Vo, Lilian
• Application Number:
  US10/150,174
• Time in Patent Office: 1,523 Days
• Field of Search: 718/100, 718/104, 718/105, 709/224, 709/226
• US Class Current: 718/104
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system

Abstract

A method for scaling resources according to workload among virtual servers running on a mainframe computer is provided. The invention comprises monitoring the number of service requests received by a cluster of virtual servers and determining if the service requests exceed a specified service limit for the servers. If the number of service requests exceeds the specified service limit, a new virtual server is automatically deployed by a software solution. This additional virtual server performs identical services as the other virtual servers in the cluster. This process is repeated until there are a sufficient number of servers to handle the workload. Service requests are then allocated among the cluster of virtual servers, until the number of requests falls below a certain threshold, at which point the extra servers are automatically deactivated.

Claims

1. A computer implemented method for balancing workload among virtual servers configured in a virtual server cluster, the method comprising the computer implemented steps of:

monitoring requests directed to the virtual server cluster and determining if an amount of work represented by the requests exceeds a specified limit;

if the amount of work represented by the requests exceeds the specified limit, automatically deploying a new at least one additional virtual server to the virtual server cluster such that the at least one additional virtual server is added to and operates in tandem with the virtual server cluster, wherein the at least one additional virtual server performs similar functions as the virtual server cluster; and

allocating subsequent requests between the virtual servers and the at least one additional server that is newly deployed such that the virtual servers and the at least one additional virtual server operate in tandem as the virtual server cluster, wherein the virtual servers are deployed on a mainframe computer as a plurality of model images executing in a memory of the mainframe computer, and the at least one additional virtual server is deployed by propagating one of the model images into a portion of the memory usable by the at least one additional virtual server to create a new model image, and updating the new model image with local identification parameters of the at least one additional virtual server.

2. The method according to claim 1, wherein the step of automatically deploying at least one additional virtual server comprises deploying a plurality of additional virtual servers.

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

if an amount of work represented by subsequent requests falls below the specified limit, automatically deactivating an appropriate subset of the plurality of additional virtual servers.

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

if any of the plurality of additional virtual servers remains idle for a specified length of time, automatically deactivating the idle virtual servers.

5. The method according to claim 1, wherein the additional virtual server has a subset of a capability of the first virtual server.

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

if an amount of work represented by subsequent requests falls below the specified limit, automatically deactivating the additional virtual server.

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

if the additional virtual server remains idle for a specified length of time, automatically deactivating the additional virtual server.

8. The method according to claim 1, wherein the step of allocating subsequent requests among the first and additional virtual servers is performed by a control program operating on a mainframe hypervisor.

9. The method according to claim 8, wherein the controller program creates a request queue for the virtual servers.

10. A data processing system for balancing workload among virtual servers configured in a virtual server cluster, the data process system comprising at least one hardware data processor and further comprising:

a means for monitoring requests directed to a first virtual server and determining if an amount of work represented by the requests exceeds a specified limit;

a means for automatically deploying a new at least one additional virtual server to the virtual server cluster if the amount of work represented by the requests exceeds the specified limit, such that the at least one additional virtual server is added to and operates in tandem with other of the virtual servers of the virtual server cluster, wherein the additional virtual server performs similar functions as the first viral server; and

a means for allocating subsequent requests between the first and additional virtual servers that are newly deployed such that both the first server and the newly deployed additional virtual servers operate as a part of the virtual server cluster, wherein the virtual servers are deployed on a mainframe computer as a plurality of model images executing in a memory of the mainframe computer, and the at least one additional virtual server is deployed by propagating one of the model images into a portion of the memory usable by the at least one additional virtual server to create a new model image, and updating the new model image with local identification parameters of the at least one additional virtual server.

11. The data processing system according to claim 10, wherein the means for automatically deploying at least one additional virtual server further comprises means for deploying a plurality of additional virtual servers.

12. The data processing system according to claim 11, further comprising:

if an amount of work represented by subsequent requests falls below the specified limit, a means for automatically deactivating an appropriate subset of the plurality of additional virtual server.

13. The data processing system according to claim 11, further comprising:

if any of the plurality of additional virtual servers remain idle for a specified length of time, means for automatically deactivating the idle virtual servers.

14. The data processing system according to claim 10, wherein the additional virtual server has a subset of a capability of the first virtual server.

15. The data processing system according to claim 10, further comprising:

if an amount of work represented by subsequent requests falls below the specified limit, a means for automatically deactivating the additional virtual server.

16. The data processing system according to claim 10, further comprising:

if the additional virtual server remains idle for a specified length of time, means for automatically deactivating the additional virtual server.