Abstract: The system and method interface and separate formatted content and program code in connection with a content server system. The system includes a document preprocessor, a compiler, and an execution context manager. The document preprocessor parses and analyzes a formatted content document that contains program invocations and translates the program invocations into a procedure having program code suitable for compilation or intepretation. The preprocessor also encodes designated segments of the content document as subprocedures which are invoked within a context managed by the execution context manager. The execution context manager allows values to be mapped to named parameters so that the values are available to the invoked procedures and programs. As a result, the content server system provides separation between the formatted content and program code.

Abstract: A computer system for a web site uses load balancing of new sessions based on measurements and estimates of past, present, and future performance, using the average time each application server takes to respond to a request (the latency) and the number of active sessions on each application server. Generally, once assigned to a session, an application server handles all future requests relating to that session. However, if the load or performance of an application server becomes sufficiently degraded, a failover mechanism causes future requests to be handled by a different application server until the performance improves. Under extreme circumstances, an application server will be restarted. If a web or HTTP server is unable to have requests served by any application server, future requests may be redirected to another web server.

Abstract: A computer system for a web site uses three tiers of servers, web (or HTTP) servers, application servers, and backup servers. The backup servers are responsible for backing up the session data for particular application servers. The system assigns to each web session a session ID, which encodes the IP addresses of the application server and its backup server, and provides an identifier that uniquely identifies the session within that application server. A session is automatically routed to a second application server to process a request if the application server handling the session should fail or not have the requested application. However, the request uses the original session ID. The second application server detects from the session ID that the session may have been handled by the first application server. The second application server decodes the IP address for the backup server for the first application server, and connects to that backup server.