Abstract: Embodiments include a system for releasing unused files. The files may be held open initially by a virtual machine, class loader, application or similar program. Embodiments register the access times to files with a timeout manager program. After a predetermined time period where no access has been made to a file, the file may be closed. This allows an update module to replace or modify files that have been closed while keeping open files that are frequently used to minimize the processing time for accessing frequently used files.

Abstract: Within a cluster, when a new server node joins the cluster, an indication of the new server node has joined is received by at least one server node in the cluster. Global naming operations are disabled within the cluster. While the global naming operations are disabled, the global naming objects are replicated to the new server node. Once the replication is complete, global naming is reenabled for the cluster. In one embodiment of the invention, subsequent naming data from subsequent global naming operations is replicated in all server nodes in the cluster. One embodiment of the invention redirects naming operations directed to a down server node to a different server node within the cluster. One embodiment of the invention performs security checks responsive to a first access to the naming service.

Abstract: A method and system for providing naming operations. Contexts are created and organized hierarchically under an initial context. The hierarchy is retained in a non-persistent storage medium. The objects and contexts created as a result of various naming operations are removed from the naming system responsive to a reboot of a server. In one embodiment, the entire naming hierarchy is represented in nested hash tables with a nesting level of only two.

Abstract: Embodiments include a system for updating software components in a computing system. The update system may deploy software components into a centralized database. The update system may provide an interface for deploying software components into the database. The update system may map the software components into a relational database or similar database system. In an additional embodiment, the deployed software components may be downloaded from the centralized database during a start up process for a computing system.

Abstract: Embodiments include a system for managing files or software components by use of a descriptor file. The descriptor file may track references to the file or software component by other files and software components. A software component that has no references to it may be removed from a system.

Abstract: A method and system to provide easy access to factories in a naming service. The system receives an indication that a factory is needed by a client. A plurality of resolvers are checked in turn to find the factory indicated. If a resolver that can provide the factory is found, the factory is returned to the client. In one embodiment, resolvers register with a resolver manager that performs the checking in response to a naming operation. Because the resolver manager can access all registered resolvers, access to the factories is independent of a class path of the user.

Abstract: A cluster includes a plurality of application server instances, a central services instance that includes a message server, and a database. The application server instances each include a dispatcher, a plurality of redundant server nodes, and a socket connection between the dispatcher and each of the server nodes for handling communications relating to processing a client request. A separate socket connection between the message server and each of the server nodes is provided for handling internal communications between the server nodes. Additionally, a third socket connection may be established directly between server nodes.

Abstract: Embodiments include a timeout event management system that registers timeout events and checks for and corrects inaccuracies in timing caused by hibernation or system time changes. The timeout event management system may trigger an event after an intended delay time or at an intended expiration time. A handler program may be called in response to the triggered timeout. In an additional embodiment, the timeout system may track timeout events in a priority queue data structure.

Abstract: Embodiments include a system for loading components with complex intra-dependencies. Components in the system may be assigned at start up to a common loader module. The system detects reference cycles amongst the set of components in the system. All components in a reference cycle may be assigned for loading to the same common loader. This system avoids deadlock scenarios by identifying reference cycles at start up and assigning each cycle to a single common loader. The embodiments of the system also analyze components to be loaded that are identified after start up to determine if they cause a new reference cycle. Components that cause a new reference cycle may not be allowed to be loaded to prevent deadlock loading scenarios.