STATE MAINTENANCE AS A SERVICE

• US 20140059209A1
• Filed: 08/27/2012
• Published: 02/27/2014
• Est. Priority Date: 08/27/2012
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Alnoor, Ahmadullah
• Granted Patent: US 9,110,844 B2
• Time in Patent Office: Days
• US Class Current: 709/224
• CPC Class Codes: G06F 11/1438 Restarting or rejuvenating
• CPC Class Codes: G06F 11/1479 Generic software techniques for error detection or fault masking
• CPC Class Codes: G06F 2201/815 Virtual middleware or OS functionality using virtual machines to implement generic software techniques for error detection or fault masking G06F11/1484
• CPC Class Codes: G06F 2209/541 Client-server
• CPC Class Codes: G06F 9/54 Interprogram communication

Abstract

When a stateful service is implemented in a cloud architecture, a state service system receives a user interface definition for a service and generates a client proxy and optionally a storage proxy. The stateful service implements service instances to service requests from clients. The client proxy receives and logs messages from a client in a state service component and a storage proxy logs interactions with a data store. When a service instance serving existing clients is changed, the proxies replay the logged information to the new service instance to recover or maintain state for individual client sessions.

Claims

1. A computer-implemented method of providing a state service to a client of the service, comprising:

receiving an interface definition for the service;

generating a client proxy based on the interface definition, the client proxy being deployed between the service and the client; and

receiving a service request from the client at the client proxy, during a session between the client and the service, the client proxy providing the service request to the service and storing the service request with a state service.

2. The computer-implemented method of claim 1 and further comprising:

after receiving the service request, obtaining a service instance identifier of a first service instance to use in servicing the service request.

3. The computer-implemented method of claim 2 wherein providing the service request to the service comprises:

sending the service request from the client proxy to the first service instance; and

the client proxy sending responses from the first service instance to the client.

4. The computer implemented method of claim 3 and further comprising:

storing a pairing of the client and the first service instance for a current session between the client and the first service instance.

5. The computer-implemented method of claim 2 and further comprising:

determining that the first service instance is unavailable; and

obtaining a service instance identifier for a second service instance.

6. The computer-implemented method of claim 5 and further comprising:

restoring session state to the second service instance using the state service.

7. The computer-implemented method of claim 6 wherein restoring session state comprises:

playing back, to the second service instance, all stored service requests stored with the state service for the session by the client proxy.

8. The computer-implemented method of claim 7 and further comprising:

generating a storage proxy for the service, the storage proxy being deployed between the service and a data store; and

receiving a data store interaction, from the service, at the storage proxy, the storage proxy storing the interaction between the service and the data store, in the state service.

9. The computer-implemented method of claim 8 wherein restoring session state comprises:

playing back, to the second service instance, the data storage interactions stored by the storage proxy with the state service.

10. The computer-implemented method of claim 6 and further comprising:

after restoring session state to the second service instance, sending the service request to the second service instance; and

updating the existing pairing of the client from the first service instance to the second service instance for the session between the client and the second service instance.

11. The computer-implemented method of claim 2 and further comprising:

generating a performance parameter user interface display; and

receiving performance parameters to be monitored.

12. The computer-implemented method of claim 11 wherein obtaining the service instance identifier for the first service instance comprises:

monitoring performance of a plurality of service instances based on the performance parameters; and

selecting the first service instance to receive the service request from the client proxy based on the monitored performance.

13. The computer-implemented method of claim 12 and further comprising:

generating an actions user interface display; and

receiving action indicators indicative of actions to be performed based on control objectives.

14. The computer-implemented method of claim 13 and further comprising:

receiving the control objectives in terms of values or ranges of the performance parameters; and

allocating and de-allocating service instances based on the control objectives.

15. The computer-implemented method of claim 14 wherein receiving action indicators comprises at least one of:

receiving an allocation preemption indicator, and further comprising;

pre-empting allocation of service instances based on the allocation preemption indicator; and

receiving an allocation deferment indicator; and

further comprising;

postponing allocation of service instances based on the allocation deferment indicator.

16. The computer-implemented method of claim 15 wherein receiving action indicators comprises at least one of:

receiving a delayed de-allocation indicator, and further comprising;

delaying de-allocation of service instances based on the delayed de-allocation indicator; and

receiving a de-allocation deferment indicator; and

further comprising;

postponing de-allocation of service instances based on the de-allocation deferment indicator.

17. The computer-implemented method of claim 16 and further comprising:

exposing an application programming interface (API) with methods to receive user inputs setting for at least one of performance parameters selection, control objectives, preemption and delayed de-allocation characteristics, and performance monitoring sensitivity.

18. The computer-implemented method of claim 12 wherein obtaining a service instance identifier comprises:

ranking the service instance based on performance to obtain service instance ranking; and

storing the service instance ranking for access by the client proxy.

19. A state service system, comprising:

a hosted service implementing a plurality of service instances;

a client proxy receiving service requests from a client, during a session between the client and the hosted service, and providing them to a first service instance of the hosted service;

a storage proxy receiving database interaction requests from the first service instance of the hosted service, for the session, and providing them to a database;

a state service that stores the service requests and the database interaction requests for the session, in response to switching service instance of the session from the first service instance to a second service instance, the client proxy replaying at least the service requests, stored in the state service for the session, to the second service instance and the storage proxy replaying at least the database interaction responses, stored in the state service for the session, to the second service instance; and

a computer processor being a functional part of the system and activated by the client proxy and the storage proxy to facilitate storing the service requests and the database interaction requests in the state service.

20. A computer readable storage medium that stores computer readable instructions which, when executed by a computer, cause the computer to perform steps comprising:

receiving an interface definition for the service;

generating a client proxy based on the interface definition, the client proxy being deployed between the service and the client;

receiving a service request from the client at the client proxy, during a session between the client and the service, the client proxy logging the service request with a state service;

generating a storage proxy for the client, the storage proxy being deployed between the service and a data store;

receiving a data store interaction, from the service, at the storage proxy, the storage proxy logging the interaction between the service and the data store, in the state service;

after receiving the service request, receiving a service instance identifier of a first service instance to use in servicing the service request;

sending the service request from the client proxy to the first service instance;

determining that the first service instance is unavailable;

receiving a service instance identifier for a second service instance;

restoring state to the second service instance using the state service by playing back, to the second service instance, all logged service requests sent to the state service for the session and playing back, to the second service instance, the data storage interactions sent by the storage proxy to the state service; and

after restoring state to the second service instance, sending the service request to the second service instance.