On Win32 platforms DP4 networking components are loaded by modifying the [startup] section of the DP4 configuration file. To load the server you can use the following startup section:
[startup]On Window CE platforms the DP4 configuration file is also used and you can use the following startup section:
[startup]On Unix and Linux machines DP4 Networking components can be loaded from the shell, or in a shell script. The server can be loaded simply by entering tcpmgr at a shell prompt. Alternately, from release 4.525 you can use dbdaemon and the following startup section in the DP4 configuration file:
[startup]When loaded like this TCPMGR will usually display information about connected clients. For more information about this display, and how to suppress it if it is not required, please refer to Network Manager Display.
If the server is accessed from machines running one or other of the Multiple Resilience Configurations, you should add the command tail -server n where n matches the server number used to access the server from client machines. It is also necessary to do this if you are using a network requester with the -server n command tail using any network protocol other than TCP/IP. Please refer to the section on server numbers for further details.
It is possible to replace the local DP4 database manager usually used by the Network Manager with another program, so that the network manager does not actually use a local database. There are three main reasons you might want to do so:
You may wish to replace the DP4 database with another database manager, such as Oracle, but continue to use existing DP4 programs. In this case you would load the server as follows on a Windows server:
[startup]The same general format would be used for loading any other ADC on a server machine.
You may want to feed network requesters using one protocol, into a system using another protocol. For example you might want to allow old MS-DOS machines that only had NetBIOS protocol to access data from a system networked using TCP/IP. You would then load a machine as a network bridge to achieve this:
[startup]Here NTBMGR will receive calls from the MS-DOS machines, and make calls to the regular DP4 interface. These will then be passed to the "real" database server by tcpw.w32.
In another situation, very similar to the last, you may need to need a bridge where both sides use the same network protocol. The two most common reasons for doing this are:
You want to concentrate a number of clients, so that they all appear to be coming from the same location on the network (this may be helpful if there is limit on the available bandwidth or number of connections). This may allow you to run more clients simultaneously than would otherwise be possible.
You want to access data over a WAN link from terminals which only have LAN connectivity.