System for application server messaging with multiple dispatch pools

Abstract

An embodiment of the invention provides an ability to configure the dispatch policy in an application or transaction server. For servlets and JSP's this is configured on a per-URL (Uniform Resource Locator) basis, RMI and EJB it is configured on a per-method basis. Available dispatch options include the ability to execute the request in the same thread that read the request or to enqueue it on a queue which feeds a pool of worker threads. Using this configuration one can control various quality of service parameters of the requests. Requests may be expedited by executing them directly in the thread which read them, thus skipping the queue. Alternatively, some requests may be throttled by assigning them to a queue which is tended by only a limited number of threads.

1 . A system for receiving and handling requests from a client application to a server resource, comprising: 
 a server request interface that receives requests from a client to access a server resource;    a plurality of execute queues, configured to queue said requests for access to the server resource, wherein each execute queue has a pool of threads assigned to pull requests from said execute queue and execute the requests against the server resource; and    a dispatch policy configured to queue specific requests on a selected one of said execute queues.    
     
     
         2 . The system of  claim 1  wherein the request specifies a quality of service to be provided to the client application in accessing the resource.  
     
     
         3 . The system of  claim 1  wherein multiple execute queues can be configured for each resource.  
     
     
         4 . The system of  claim 1  wherein the server specifies by the dispatch policy either direct dispatch or queued dispatch.  
     
     
         5 . The system of  claim 1  wherein the resource is a servlet.  
     
     
         6 . The system of  claim 1  wherein the resource is a Java Servlet Page.  
     
     
         7 . The system of  claim 1  wherein the resource is associated with a particular execute queue name.  
     
     
         8 . The system of  claim 1  wherein the pool of threads for each execute queue can be configured to allow for prioritization of requests associated with that execute queue.  
     
     
         9 . A method for receiving and handling requests from a client application to a server resource, comprising the steps of: 
 receiving requests from a client to access a server resource;    dispatching the requests to a plurality of execute queues using a dispatch policy for said server resource, wherein each execute queue is associated with a pool of threads;    pulling the requests from each execute queue; and    executing the requests against said server resource.    
     
     
         10 . The method of  claim 9  wherein the request specifies a quality of service to be provided to the client application in accessing the resource.  
     
     
         11 . The method of  claim 9  wherein multiple execute queues can be configured for each resource.  
     
     
         12 . The method of  claim 9  wherein the server specifies by the dispatch policy either direct dispatch or queued dispatch.  
     
     
         13 . The method of  claim 9  wherein the resource is a servlet.  
     
     
         14 . The method of  claim 9  wherein the resource is a Java Servlet Page.  
     
     
         15 . The method of  claim 9  wherein the resource is associated with a particular execute queue name.  
     
     
         16 . The method of  claim 9  wherein the pool of threads for each execute queue can be configured to allow for prioritization of requests associated with that execute queue.  
     
     
         17 . A computer readable medium including instructions stored thereon which when executed cause the computer to perform the steps of: 
 receiving requests from a client to access a server resource;    dispatching the requests to a plurality of execute queues using a dispatch policy for said server resource, wherein each execute queue is associated with a pool of threads;    pulling the requests from each execute queue; and    executing the requests against said server resource.