摘要:
A method of tracing data on a network for a Web request that detects events of the Web request as the Web request is serviced. The method provides a unique identifier (ID) for the Web request. The ID is associated with all events for the Web request in a trace log. When such events occur in the servicing of the Web request, each event is recorded with the ID of the Web request in a trace log. The trace log can be accessed via the ID to debug a Web request service failure.
摘要:
Use of configuration information to specify particular conditions under which trace events are to be logged. When accessing trace events generated by various modules, configuration data is referred to specifying condition(s) under which the trace events should be logged. If the log condition(s) are satisfied, the trace events are logged. Otherwise, they may be discarded. Thus, the number of trace events in the log may be significantly reduced to those trace events that satisfy conditions of interest to an evaluator.
摘要:
An improved web server design includes a kernel-based listener service that is configured by a user-mode based administrative process. When a web request is received by the listener service it is placed in a queue associated with an application pool having one or more user-mode web applications configured to process received web requests. The web applications are selectively started and stopped by the administrative process.
摘要:
A web server system is provided which includes a web garden made up of a plurality of instances of a user-mode web application. The web server includes a kernel-mode listener service that is configured to receive web requests and selectively provide the TCP/IP connections to selected instances of the user-mode web application. It does this by establishing a plurality of request queues corresponding to the number of instances of the web application logic. The web server also includes a user-mode administrative process that is configured to selectively start each of the instances of the web application and maintain the size of the resulting web garden, for example, based on a defined maximum web garden size parameter.
摘要:
A web server arrangement is provided having a kernel-mode listener service capable of receiving web requests and providing the web requests to at least one user mode web application. The listener service and the web application are operatively configured by a user-mode administrative process. The administrative process is configured to determine when a web application needs to be recycled and to selectively recycle web applications.
摘要:
Improved methods and arrangements are provided in which a web server or like process includes a kernel-mode universal listener (UL) capability that utilizes the URL associated with an HTTP request to determine which user-mode worker process/application is to handle the request and route the request to the targeted worker process/application. Thus, for example, one implementation includes a kernel-mode HTTP listener that is selectively configured via a Web admin service (WAS) based a configuration store. The resulting configuration file in the kernel-mode UL process defines a plurality of configuration groups and application pools, etc. The Configuration groups include certain URL's, which have configurable properties applied to them. The configurable properties define which user-mode worker process is to handle HTTP requests. Fore example, application pools can be provided, which identify a worker process or processes that the HTTP requests are routed to. In certain implementations, an HTTP request queuing capability is also provided. The request queue provides time for worker processes/applications to start or get ready.