Instalation considerations in an Adabas to Adabas environment

In order to use the new Adabas to Adabas replication feature, the following requirements must be met:

 

  1. Both the source and target Adabas nucleus must be version 6.3.1 or above.

  2. The source and target systems must be the same platform.  (i.e. both systems must be Solaris, etc.)

  3. An Adabas to Adabas license must be installed for the source and target nucleus as well as the Open Systems Event Replicator

  4. A CDD is required and it must contain connections for at least two Adabas 6.3.1 databases.  It is not necessary to import files, but the connections must be defined.  If two 6.3.1 connections are not present in the CDD, Adabas to Adabas replication will not be available in the Replication Administrator.

  5. Before doing Adabas the Adabas replication, the source database must be prepared for replication.

    1. ADADBM has a new parameter called REPLICATION_FILES that prepares the source nucleus for replication
    2. Syntax is ADADBM DBID=xx REPLICATION_FILES=(file1, file2, file3, file4) where xx is the database ID of the source and file1 - file4 are the file numbers reserved for replication.  An example is:
      ADADBM DBID=1 REPLICATION_FILES=(15,16,17,18)
  6. ADADBM can also be called with the REMOVE_REPLICATION parameter to remove the replication files.  After making this call, the nucleus will no longer be able to replicate files until ADADBM is called again with REPLICATION_FILES.
  7. The environment variable ADABINPATH must be set in the environment(s) where the Replication Server (controller) is running as well as where the CONNX SQL Gateway Server's listener is running.  On Unix systems, this environment can be exported in the start up scripts.  The startup scripts are in the connx directory and are named eventserver and connxserver respectively.  On Windows, this needs to be set in the Environment Settings section of the Computer properties.  
    Note: In Windows, the installer will set ADABINPATH automatically.
    Note2: In Unix/Linux the CONNX SQL Gateway Server (data server) must be running in a user account that has permission to execute the binaries in the ADABIN location.  It is recommended that both the data server as well as the Event Server (controller) be run from the same user account that was used to start the Adabas nucleus.

 

The Open Systems Event Replicator is designed to be able to be distributed across multiple machines.  Certain components must be installed on the source and target machines, however.

 

Components and their locations are:

  • Source Adabas server:

    1. Adabas SQL Gateway client

    2. Adabas SQL Gateway Adabas data server

    3. Event Producer (EP)
       

  • Target Adabas server
    1. Adabas SQL Gateway Adabas data server
  • Controller (Event Server):
    1. Adabas SQL Gateway client
    2. Open Systems Event Replicator Controller (Event Server)
  • Replication Administrator machine (must be Windows XP SP2 or above):
    1. Adabas SQL Gateway with Open Systems Event Replicator selected

While the Open Systems Event Replicator is designed to be distributed across multiple machines, in an Adabas to Adabas environment, the following configuration is most likely optimal:

  • Source Adabas server:
    1. Adabas SQL Gateway Client
    2. Adabas SQL Gateway Adabas data server
    3. Open Systems Event Replicator with both the controller and EP installed
  • Target Adabas server:
    1. Adabas SQL Gateway Adabas data server
  • Replication Administrator machine (must be Windows XP SP2 or above)
    1. Adabas SQL Gateway with Open Systems Event Replicator selected

 

Note: For information about installing the Adabas SQL Gateway, please see the CONNX Installation Guide.  For information installing Event Replicator, please see the previous topics in this chapter for your specific platform.