A method and system for inserting and examining Cookies in the data streams of HTTP connections for the purpose of persistently directing HTTP connections to the same destination. The present invention enables a network transmission device, e.g., a router or controller, to reliably and efficiently direct subsequent HTTP connections from the same client to the same server (destination) for accessing the requested resources. There are four modes for employing the Cookie to persistently direct HTTP connections. The associative mode inserts a Cookie into an HTTP response that uniquely identifies the client so that when a client's subsequent HTTP request is compared to a table, the HTTP request will be routed to a previously selected destination associated with the client. The passive mode inserts Cookie information into an HTTP response that uniquely identifies a previously selected destination such as a node server so that when a client's subsequent HTTP request is examined, it will be provided to the previously selected destination. In the rewrite mode, a network transmission device manages the destination information that is rewritten over blank Cookie information generated by the destination producing the HTTP response. The insert mode enables a network transmission device to insert and remove Cookie information in the data packets for HTTP requests and responses prior to processing by the destination. The Cookie information can include a time stamp. Other types of Cookies include priority, path and hops.

Method and system for storing load balancing information with an HTTP cookie

A system and method for balancing the load on virtual servers managed by server array controllers at separate data centers that are geographically distributed on a wide area network such as the Internet. The virtual servers provide access to resources associated with a domain name request by a client program. When a Primary Domain Name System (DNS) determined the requested domain name is delegated to a EDNS, the EDNS employs metric information and statistics to resolve an ip address for a virtual server that is selected by the EDNS to optimally balance the load and provide access to resources associated with the domain name. The EDNS may employ a static or a dynamic load balancing method to select the virtual server most suited to balance the load across all of the virtual servers. The EDNS may include a Primary DNS or a Secondary DNS. The EDNS employs an agent program located at geographically distributed data centers to collect metric information related to a host machine, server array controller, virtual servers and the path for providing resources associated with the domain name to a client making the request. The EDNS collects the metric information from the agent program out of band of the domain name resolution process. The server array controller may include the agent program and the agent program may be provided as a stand alone machine that is coupled to the server array controller or a host machine.

Method and system for balancing load distribution on a wide area network

A system and method for balancing the load on virtual servers managed by server array controllers at separate data centers that are geographically distributed on a wide area network such as the Internet is described. The virtual servers provide access to resources associated with a domain name request by a client program. When a Primary Domain Name System (DNS) determined the requested domain name is delegated to a EDNS, the EDNS employs metric information and statistics to resolve an IP address for a virtual server that is selected by the EDNS to optimally balance the load and provide access to resources associated with the domain name. The EDNS may load balance name servers. Additionally, the name server load balancing system may bridge disparate content delivery networks. Internet addresses are divided into geographical information that is used to delegate traffic. Also, metric information is collected and analyzed to help distribute the traffic.

Method and system for balancing load distrubution on a wide area network

A method and system for managing the replication and version synchronization of updates to a set of source files on geographically distributed heterogeneous content servers with minimal impact on a network's bandwidth. The configuration of each content server is either manually entered or automatically determined. The current version of the source files are created on at least one source server. A Primary global server stores a copy of the current version of the set of the source files along with the configuration of each content server. The Primary global server generates and distributes a particular version change container and version distribution list to each remotely located Secondary global server. Each Secondary global server employs the version distribution list and the contents of the version change container to identify the current version of each source file necessary to upgrade the set of source files on each local content server. Each identified source file is copied to a sub-directory on each local content server associated with the Secondary global server. At each local content server, the renaming of each copied source file is employed to update to the current version of the set of source files on the content server. A versioned file tree repository for the set of source files includes archived objects. When the version distribution list identifies a previous version, the current version of source files on the local content servers can be rolled back to the previous version.

Method and system for automatically updating the version of a set of files stored on content servers

An apparatus is related to connection management for a communications network. A control component receives a data flow requesting a resource from a client, identifies the client, and determines when the data flow is unassociated with a connection to a requested resource. The control component selects a new content server for an unassociated resource request when either the identified client was previously unknown or the identified client has exceeded a maximum number of connections with a previously selected content server. The control component selects the previously selected content server when the identified client has not exceeded the maximum number of connections. A switch component is employed to maintain a connection between the client and the selected content server such that the client receives the requested resource. Utilizing cached connection information for up to “N” connections enhances the speed of connections between the client and the selected content server.

Simplified method for processing multiple connections from the same client

A pattern writing method for X-ray mask fabrication including forming a uniform membrane layer on an X-ray absorbing layer and forming an etch mask on the layer of X-ray absorbing material including the steps of providing a layer of material sensitive to radiation. The layer of material has internal stresses which are altered by exposure to the radiation. The material is exposed in associated areas (e.g. a spiral) such that the internal stresses within the layer of material and altered internal stresses in the associated areas are substantially offset to reduce distortion in the X-ray mask.

Daniel L. Laird

Papers

Pattern writing method during X-ray mask fabrication