Polling-based mechanism for improved rpc timeout handling

Abstract

The present invention provides for a method and computer program product for handling timeout in a standard RPC connection. First, a client establishes a connection with a server with unique identification. After submitting an RPC request, the client system will periodically make secondary requests to the server to determine if the server is still actively processing the primary RPC request. If the secondary request is processed successfully and the server indicates that the primary request is still in progress, the client will continue to wait until either the primary request completes or enough time elapses to warrant another secondary request. The success of the secondary request hinges on finding a match of identification for the primary and secondary requests. If the secondary request fails, this failure is treated as a sign that there is either a network or a server problem, and the client is triggered into taking appropriate corrective action. To provide for a reasonably graceful failure mechanism, this polling protocol can be modified to require a predetermined number of successive secondary poll failures before signaling a failure of the primary RPC request.