System and method for supporting transaction affinity based request handling in a middleware environment

• US 10,127,122 B2
• Filed: 11/08/2016
• Issued: 11/13/2018
• Est. Priority Date: 10/21/2014
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Somogyi, Alexander, Subramanyam, Sindhu, Felts, Stephen
• Primary Examiner(s): Dao, Thuy
• Assistant Examiner(s): Wang, Ben C
• Application Number:
  US15/346,135
• Publication Number:
  US 20170052855A1
• Time in Patent Office: 735 Days
• CPC Class Codes: G06F 11/2007 using redundant communication media
• CPC Class Codes: G06F 2201/85 Active fault masking without idle spares
• CPC Class Codes: G06F 8/20 Software design
• CPC Class Codes: G06F 8/71 Version control security arrangements therefor G06F21/57; Configuration management
• CPC Class Codes: G06F 9/466 Transaction processing
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals
• CPC Class Codes: G06F 9/5033 considering data affinity
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: H04L 67/34 involving the movement of software or configuration parameters network booting or remote initial program loading [RIPL] G06F9/4416

Abstract

A system and method can support transaction processing in a middleware environment. A processor, such as a remote method invocation stub in the middleware environment, can be associated with a transaction, wherein the transaction is from a first cluster. Then, the processor can handle a transactional request that is associated with the transaction, wherein the transactional request is to be sent to the first cluster. Furthermore, the processor can route the transactional request to a said cluster member in the first cluster, which is an existing participant of the transaction.

Claims

1. A method for supporting transaction processing in a computing cluster including a plurality of application servers, the method comprising:

initiating a first global transaction from an application running on a host server external to the cluster by transmitting a first RMI request from the host server to said cluster;

receiving at said cluster the first remote method invocation (RMI) request associated with the first global transaction;

determining that no RMI request associated with said first global transaction has been previously received by said cluster;

selecting a first application server of said plurality of application servers for processing said first RMI request;

routing said first RMI request to said first application server for processing;

processing on the first application server, a first method invoked by the first RMI request associated with the first global transaction;

receiving at said cluster a second RMI request;

determining whether the second RMI request is associated with the first global transaction; and

selecting the first application server for processing said second RMI request if the second RMI request is associated with the first global transaction, and using a load balancing method for selecting a second application server of said plurality of application servers for processing said second RMI request if the second RMI request is associated with a second global transaction.

2. The method of claim 1, further comprising:

selecting the first application server for processing all subsequent RMI requests associated with the first global transaction received at said cluster.

3. The method of claim 1, wherein the second RMI request is associated with the second global transaction, the method further comprising:

selecting the second application server for processing all subsequent RMI requests associated with the second global transaction received at said cluster.

4. The method of claim 1, further comprising:

providing an indicator of transaction affinity routing in said first RMI request sent from said host server to said cluster.

5. The method of claim 1, further comprising:

committing the first global transaction using a two-phase commit protocol.

6. The method of claim 1, further comprising:

providing an interceptor in said cluster for intercepting RMI requests; and

using a handler in said cluster to select application servers for processing said RMI requests and routing said RMI requests to application servers selected.

7. The method of claim 1, further comprising:

using a cluster manager to configure the cluster, to require transaction affinity based request handling.

8. The method of claim 1, further comprising:

performing a failover operation to select another application server of the plurality of application servers, if the first application server is not available for processing a subsequent RMI request associated with the first global transaction.

9. The method of claim 1, wherein the load-balancing method comprises a round robin load balancing method for selecting an application server of the plurality of application servers.

10. The method of claim 1, further comprising:

allowing a user to specify an application server in the first cluster to handle a particular RMI request.

11. A system for supporting transaction processing comprising:

a computing cluster including a plurality of application servers, each application server comprising one or more microprocessors;

an application running on a host server external to the cluster which initiates a first global transaction and transmits a first RMI request from the host server in the cluster;

an interceptor and a handler running on said plurality of application servers in said cluster;

wherein said interceptor and handler are configured to perform steps comprising;

intercepting the first remote method invocation (RMI) request associated with the first global transaction received at said cluster;

determining that no RMI request associated with said first global transaction has been previously received by said cluster;

selecting a first application server of said plurality of application servers for processing said first RMI request;

routing said first RMI request to said first application server for processing;

intercepting a second RMI request received at the cluster;

determining whether the second RMI request is associated with the first global transaction; and

selecting the first application server for processing said second RMI request if the second RMI request is associated with the first global transaction, and using a load balancing method for selecting a second application server of said plurality of application servers for processing said second RMI request if the second RMI request is associated with a second global transaction.

12. The system according to claim 11, wherein:

the interceptor and handler are further configured to select the first application server for processing all subsequent RMI requests associated with the first global transaction received at said cluster.

13. The system of claim 11, wherein:

the interceptor and handler are configured select the second application server for processing all subsequent RMI requests associated with the second global transaction received at said cluster when the second RMI request is associated with the second global transaction.

14. The system of claim 11, wherein:

said application running on the host server includes an indicator for transaction affinity routing in said first RMI request.

15. The system of claim 11, wherein:

said application running on the host server is configured to commit the first global transaction using a two-phase commit protocol.

16. The system of claim 11, wherein:

the interceptor and handler are configured to allow a user to specify an application server in the first cluster to handle a particular RMI request.

17. The system of claim 11, further comprising:

a cluster manager adapted to configure the first cluster to require transaction affinity based request handling.

18. The system of claim 11, wherein:

the interceptor and handler are further configured to perform a failover operation to select another application server of the plurality of application servers, if the first application server is not available for processing a subsequent RMI request associated with the first global transaction.

19. The system of claim 11, wherein:

the load-balancing method comprises a round robin load balancing method for selecting an application server of the plurality of application servers.

20. A non-transitory machine readable storage medium having instructions stored thereon for supporting transaction processing in a computing cluster including a plurality of application servers, which instructions, when executed, cause said cluster to perform steps comprising:

initiating a first global transaction from an application running on a host server external to the cluster by transmitting a first RMI request from the host server to said cluster;

receiving at said cluster the first remote method invocation (RMI) request associated with the first global transaction;

determining that no RMI request associated with said first global transaction has been previously received by said cluster;

selecting a first application server of said plurality of application servers for processing said first RMI request;

routing said first RMI request to said first application server for processing;

processing on the first application server, a first method invoked by the first RMI request associated with the first global transaction;

receiving at said cluster a second RMI request;

determining whether the second RMI request is associated with the first global transaction; and

selecting the first application server for processing said second RMI request if the second RMI request is associated with the first global transaction, and

using a load balancing method for selecting a second application server of said plurality of application servers for processing said second RMI request if the second RMI request is associated with a second global transaction.