ColdFusion 4.0 can be configured in a distributed manner where the ColdFusion engine is running on a separate computer from the Web server. Running ColdFusion in this way might be called distributed or remote ColdFusion.
To run distributed ColdFusion, you must make the following changes to a standard installation:
- On the Web server side, you must notify the ColdFusion Web server plug-in that you want it to talk to a ColdFusion engine on another machine. You do this simply by making appropriate entries in an INI file.
- On the ColdFusion engine side, you must run an additional piece of software, known as the Network Listener Module, that listens for incoming ColdFusion requests and forwards them to the ColdFusion engine running on that machine. The ColdFusion engine itself is a standard release version of the engine with no special modifications to accommodate remoting.
In addition to allowing the ColdFusion engine to be located on a separate machine from the Web server, distributed ColdFusion provides the following unique capabilities:
- It allows the machine hosting the Web server to potentially be of a different architecture from the machine hosting the ColdFusion engine.
- It allows more than one Web server to be served by the same ColdFusion engine.
To provide some degree of security for the data being transferred between the Web server and the ColdFusion engine, that conversation is encrypted using a standard, 56-bit DES encryption algorithm.
Although it's possible for a ColdFusion engine to simultaneously service both local and remote requests, it is not possible for a single Web server to simultaneously dispatch both local and remote ColdFusion requests. When starting up, the ColdFusion Web server plug-in determines if it's to run in local or remote mode and remains in that mode until it's shutdown.
|