The DME provides a powerful way to redistribute media by allowing you to reach multiple/remote locations and multiple users with minimal use of streaming bandwidth. Streams can be converted from unicast to multicast or delivered as Flash HDS and Apple HLS streams from an RTP source. Since the DME accepts multiple types of input streams and provides multiple ways to output streams, it may not be entirely clear which use cases apply to you and what is the simplest way to deploy your solution using the DME. The best way to determine how to use the DME effectively is to understand three basic factors:

Once you have a better understanding of these issues you are ready to start considering what type of input streams you will have (RTP or RTMP) and how will they be distributed. For example they can be pushed to the DME, pulled from the DME, or by unannounced unicast from the source or an announced auto-unicast to the DME. You will also know how your clients will be viewing the content, for example as RTP, RTMP, or both, using a standalone player, an embedded web page, or through VBrick's VEMS Portal Server. You will also know whether or not the content needs to be relayed to another remote DME or to a CDN for Internet Distribution. Finally, knowing how many users you have and the bandwidth consumed by each will help to clarify how many DMEs and which models you will need to distribute the streams. By gathering this information in advance, and reading this manual carefully, you can help to ensure a successful deployment of the DME in your own unique environment.

To help you understand the various options available, the topic Configure a DME Stream defines a number of typical use cases—not all of which will apply to you. The use cases can help to simplify the configuration. They can help, for example to avoid deploying a simple solution in an overly complex way. In other cases you may also choose one method for one requirement, and have to choose a different method for a second requirement, meaning you will have two input streams when one could just as easily be used for both.

Firewalls can also play an important role in determining which use cases are appropriate. When no firewalls apply, a push or an auto unicast solution can be easily deployed. However if the DME is behind a firewall, you probably cannot reach it with a push without having to reconfigure the firewall. Similarly, you can probably pull a stream from a source into the DME. However if the source is also behind a firewall, more network planning, such as placing the DME in a "DMZ" (which the source can push to and the destination can pull from) may be a better solution. If virtual IP addresses are used, you will need to know more about the configuration of the network; and if deploying RTP streams that will travel over UDP, your firewall may need to be configured to allow UDP data in and out.