Installing additional runtime cluster nodes on Linux

After installing the initial runtime cluster node on your Linux operating system, install additional runtime cluster nodes to give your application more processing power.

Before you begin

Before completing these steps, install the initial node.

About this task

You must run these steps on each machine that you are adding to the .

Procedure

1. On the machine on which you want to run the node, mount the NFS directory containing the initial installation (<molecule_installation_directory>) where <molecule_installation_directory> is the name of the directory.

2. Start the node manually by running this command:

   <molecule_installation_directory>/bin/atom start

3. (Optional) For a SysV-style initialization where the node starts automatically at startup, do the following:

   1. Create a link in the /etc/init.d scripts directory to the <molecule_installation_directory>/bin/atom script.

   2. Configure the system to start this script by using existing SysV setup tools.

   note

   The specifics of this step might vary depending on your distribution.

4. After starting the new node, verify that it joined the by looking in the <molecule_installation_directory/logs<date>.container.<machine_IP_address>.log file for a message like this:

   INFO: Started <machine_IP_address> as CloudletAddress <machine_IP_address>:7800, initial cluster view: CloudletAddress <other_machine_IP_address:7800, CloudletAddress <machine_IP_address>:7800, ...