Defining High-Availability Settings in the Configuration Files

Edit online

After you install Connect:Direct® for UNIX on a shared file system, modify IBM® Connect:Direct parameters to support a clustering environment. Install Connect:Direct for UNIX on a shared cluster file system to use it in a cluster environment. Complete the following procedure to modify the configuration files for a cluster environment:

Procedure

  1. Modify the following parameters:
    • In the initialization parameters file (initparm.cfg), set :comm.info=0.0.0.0;nnnn:\ where nnnn is the number of the listening port you defined during installation.
    • In the api.parms record of the NDMAPI configuration file (ndmapi.cfg), set :tcp.hostname=logical_host_ip_name:\ where logical_host_ip_name is the virtual address of the cluster.
    • In the network map file (netmap.cfg), set :tcp.api=logical_host_ip_name;nnnn:\ where nnnn is the API port you defined during installation.
  2. In the network map file (netmap.cfg), set the outgoing address parameter in the local.node record to specify the local host IP name or address of the floating address to the following value. The remote node will also use this value for network map checking.
    :outgoing.address=(host name |nnnnnn.nnn):\
  3. Modify the following records in the network map file:
    • Set :comm.info=logical_host_ip_name;1364:\ to configure the loopback remote node record.
    • Set tcp.max.time.to.wait to a value other than zero and less than the value set in the resource group manager of the cluster software to allow for clean shutdowns.
  4. In the same volume group as the installation file system, create a user data file system that is shared by all cluster nodes.

Setting Up Additional Configuration Requirements in a SunCluster 3.X Environment

Edit online

The High-Availability cluster commands shown below are not intended to be a complete set of instructions for setting up the High-Availability cluster software. Additional steps may be required to complete the configuration of the High-Availability environment. High-Availability cluster expertise is the responsibility of the customer. White papers detailing specific environments, setup steps, and testing of various High-Availability clusters are available on the Support On Demand web site. In addition to modifying the configuration files, complete the following procedure to set up a SunCluster 3.X cluster:

Procedure

  1. Type the following command to create the cluster resource:
    scdscreate -V SCI -T cd

    Use the V parameter to define the vendor ID and T parameter to define the resource ID.

  2. Type the following command to configure the custom resource scripts:
    scdsconfig
  3. Edit the SCI.cd resource file and change the value of RT_BASEDIR as follows:
    RT_BASEDIR=/opt/SCIcd/bin;

    RT_BASEDIR=/global/vol1/sci/cduserk1/3.5.00/suncluster+scripts/SCIcd/bin;

Configuring Additional Requirements in a SunCluster 2.2 Environment

Edit online

In addition to modifying the configuration files, complete the following steps to complete the SunCluster 2.2 setup:

Procedure

  1. Place the following sample scripts and configuration files in a directory that is available to SunCluster 2.2 software:
    • cd_start.sh
    • cd_stop.sh
  2. Update the scripts as required for your environment.
  3. Copy the sample scripts to all nodes in the cluster.
  4. Issue the hareg command to register the IBM Connect:Direct data service. Refer to the SunCluster documentation for more information. Following is a sample command:
    hareg -r cd \