Using the machinename parameter

Under Client/Server Binding for Enterprise Server, the "machinename" parameter has somewhat different behavior and syntax than with previous versions. When using Client/Server Binding with CCI, specifying a machinename simply told CCI to look for the named service on the given machine. CCI still had to query the service directory on the named machine to find the service.

Client/Server Binding for Enterprise Server uses Micro Focus Common Client (MFCC), the configuration file of which specifies the directories to query. The machinename parameter is now used to tell the client exactly where to connect to the service's listener - just like a URL for a web page. When machinename is used, there is no directory lookup.

Using machinename has some advantages. First, the client (and the user) knows exactly which service it will connect to. Second, it is faster than querying a directory. Third, it is useful for penetrating a firewall, since the firewall does not have to be configured to allow Micro Focus Directory Server (MFDS) queries.

However, using directory lookup (and not using machinename) also has advantages. It frees the client (and the user) from having to know where the service is located. It lets the system administrator move the service from one machine to another, or change the port of its listener, or let the operating system select the port. It permits simple failover (if the service is supplied on multiple machines, directory search will find a running instance) and load balancing (MFCC will pick an available instance of the service at random, so clients will be spread roughly equally among servers).

Broadly speaking, directory lookup is more resilient and direct addressing using machinename gives you more control. Your application requirements and execution environment will determine which is better for you.

When you use the machinename parameter with Client/Server Binding for Enterprise Server, the value can be in several forms: