Installing a master archive of an IBM® Process Federation Server configuration on another federated system

You can use a master archive of a Process Federation Server configuration to create another Process Federation Server in the federated environment.

Before you begin

Make sure that you have a master archive of a Process Federation Server configuration.
IBM Business Automation Workflow V 18.0.0.2
Note: From IBM Business Automation Workflow 18.0.0.2 onwards, you can use Process Federation Server with its default embedded Elasticsearch 1.7.2 node or you can enable connection to a remote Elasticsearch 6.x cluster.

If you choose to work with Process Federation Server connected to a remote Elasticsearch cluster, 4 of the following procedure does not apply.

Procedure

  1. Transfer the master archive file to the new system.
  2. Extract the contents of the archive file and run the following command. 
    java -jar package_file_name [Options] install_location
    Where:
    package_file_name.zip
    The name of the file that you are extracting.
    Options
    --acceptLicense
    Automatically indicate acceptance of license terms and conditions.
    For a list of all the options, see Archive file extraction options.
    install_location
    The path to the directory where you want to install the products. If you do not specify an installation location, the directory that contains the archive file is used.
    Restriction: Ensure that the directory names in the path contain only the following characters:

    0-9 A-Z a-z (space) & ( ) + - . _

  3. Remove duplicated code.

    If the installation of the master Process Federation Server configuration was not done in the pfs_install_root directory, you will have two copies of the bpmFederatedProcessServer directory on the new system after you install the master archive. During installation of the archive, the ibmProcessFederationServer directory is copied to the pfs_install_root directory on the new system. However, the archive still contains a copy of the pfs_install_root/ibmProcessFederationServer directory. You can delete the pfs_install_root/ibmProcessFederationServer from the archive on the new system.

  4. Optional: IBM Business Automation Workflow V 18.0.0.2 If you use Process Federation Server with its default embedded Elasticsearch 1.7.2 node, clean up the Process Federation Server configuration on the new system.
    • If the original Process Federation Server configuration and the new configuration are not clustered together, perform the following tasks.
      1. Check that the configured properties in the server.xml file are suitable for your environment.
      2. Delete the /elasicsearch directory on the new system, and start Process Federation Server on the new system.
    • If the original Process Federation Server configuration and the new configuration are clustered together, make the following adjustments to the server.xml file for the new Process Federation Server configuration.
      • Change the Elasticsearch node name.
      • Check for conflicting ports.
      • Ensure that the server.xml file is pointing to the correct database drivers, for example for DB2, <fileset dir="pfs_install_root/ibmProcessFederationServer/wlp-ext/jdbcdrivers/DB2" includes="db2jcc4.jar db2jcc_license_cisuz.jar"/>.
      • If both servers are to run on the same host machine, change the value of the Elasticsearch transport.tcp.port property to a value that is different from other nodes on the cluster.
      • On all servers in the cluster, change the value of the Elasticsearch discovery.zen.ping.unicast.hosts property in the server.xml file to include the host names of all the cluster members and the Elasticsearch transport TCP port, for example, discovery.zen.ping.unicast.hosts="<hostname>:9300,<hostname>:9301".
      • On all servers in the clusters, adjust the value for the discovery.zen.minimum_master_nodes property to n/2 +1, where n is the number of master eligible nodes in the Elasticsearch cluster.
      • Ensure that host-specific SSL certificates and signer certificates for any Process Federation Server clients are updated, for example, any HTTP Server in front of the process federation servers.
      • Before you start the new Process Federation Server for the first time, delete the /elasticsearch directory on the server.