Trademarks. Third-Party Information. Jabber is a registered trademark of the Jabber Software Foundation. Acknowledgments, страница 6

function escapeURI(uri) {             index = uri.indexOf(‘?’);             if (index == -1) return uri;             prefix = uri.substring(0, index);             uri = uri.substring(index);             return prefix += escape(uri); }

You call this function on the URI you are about to pass into the shared object, then use the result of the function instead of the original URI. Flash Player 8 resolves this problem for Flash Media Server.

A second problem surfaces when RTMPT is specified in the URI. You may specify this protocol only for the first prefix. The following are valid URIs:

rtmpt://foo/?rtmp://bar/app/inst rtmpt://foo:443/?rtmp://bar/app/inst The following are invalid URIs:

rtmpt://foo/?rtmpt://bar/app/inst rtmp://foo/?rtmpt://bar/app/inst

Configuring edge servers

All Flash Media Server functionalities—live streaming, on-demand streaming, messaging, shared object handling, scripting—take place as application instances. Application developers write scripts to create Flash Media Server applications, instances of which are then distributed by the Flash Media Server. A minimal application runs on the edge to perform connection aggregation, stream splitting, stream caching, and smart state management. This separation of tasks off loads to the edge server a large portion of the processing from the main application that runs on the origin server.

Flash Media Server and network administrators can configure edge servers to function in different ways. The XML configuration files define how the edge servers connect clients to the origin server. These settings determine how the incoming URLs are routed between the edges and the origin server. You can have a computer run Flash Media Server in a mixed mode, in which certain virtual hosts on this system run applications locally, while others run applications remotely.

To configure Flash Media Server, you edit or modify the XML tags in its configuration files. Following are some representative examples:

■ The Anonymous tag in the Vhost.xml file configures the edge server as either an implicit (transparent) or explicit proxy.

For more information, see “Anonymous” in Managing Flash Media Server.

■ The Mode tag in the Vhost.xml file configures Flash Media Server to run as an origin or edge (proxy) server.

For more information, see “Mode” in Managing Flash Media Server.

■ The LocalAddress tag in the Vhost.xml file allows the administrator to control the network traffic by isolating incoming and outgoing traffic to separate network interfaces.

For more information, see “LocalAddress” in Managing Flash Media Server.

■ The Scope tag in the Application.xml file determines the process scope in which to run the application. Set this tag to run the applications and proxies in inst mode. Each application and virtual host has its own process.

For more information, see “Scope” in Managing Flash Media Server.

You can find examples of the configuration files, their tag structures, detailed information about the tags, including cross references to associated tags, syntax, and examples in Chapter 3, “Configuration Files” in Managing Flash Media Server.

Deploying a cluster of edge servers

You can also deploy a group of connected edge servers as a cluster. Each edge server in a cluster can access the origin server. Clustering lets you distribute across the many edges in the cluster all connections to applications that are running on the origin server. The origin server always remains secure behind the edge servers.

For example, Organization A deploys a cluster of edge servers on its internal network. The edge servers are defined as explicit proxies and dedicated to service specific users. When the clients or users want to connect to the origin server, the client first connects to one of the edge servers in the cluster. It is at this point where the connection requests are authorized or rejected. If the request is valid, the edge server makes the connection to the origin server, where the application is running. The origin server returns the results to the edge server, which passes them on to the requesting client. The edge server also stores the results in a cache, making it available to other servers and clients.