Install a Management Host

After installing on the compute host, set up a Linux host as a management host.

The following steps summarize installation of a management host:

  1. Obtain the necessary files:
    1. Obtain RPM package.
  2. As root, deploy the software on the management host:
    1. Customize cluster properties.
    2. Run RPM.
    3. Set the command-line environment.
    4. Set up automatic system startup.
    5. Grant root privileges to a cluster administrator
  3. As egoadmin, configure and test the management host:
    1. Set the command-line environment.
    2. Join the cluster.
    3. Define the host as a management host.
    4. Set the command-line environment.
    5. Start the host.
    6. Restart the Linux cluster
      Note:

      You need to restart the cluster each time you add a management host to the cluster.

    7. Test that a management host was added to the cluster.

Repeat the steps in this chapter to add more management hosts to your cluster.

To expand your cluster further, you should enable master host failover, as described in Expand the cluster.

Obtain RPM package

Choose the RPM package according to the operating system you are installing on and the host role.

RPM package for required operating system version

Each operating system version has a separate RPM package. Obtain the package to match your host.

For example, for x86 hosts running Linux 2.4 with glibc version 2.3, install the package named ego-linux2.4-glibc2.3-x86-1.2.3.nnnnnn.rpm.

Find your RPM version

To find out which version of RPM you are using, use the rpm --version option. Different versions of RPM require different options to install the packages.

For example:
rpm --version
RPM version 4.2.3

Deploy the software on the management host

Check the following:
  • That you are logged on as root.

  • That the installation directory does not exist, or if it exists, that it is not in use.

    The default installation directory is /opt/ego.

  • That the required connection ports are not in use.

    The default base connection port is 7869. EGO uses five consecutive ports starting from this base port (7869-7873).

    Tip:

    Run netstat -an to see if ports are in use.

  • That the required web server ports are not in use.

    The default web server ports are 8080, 8005, and 8009.

  • That the required web service gateway port is not in use

    The default web service gateway port is 9090.

  • That the required service director port is not in use.

    The service director port is 53.

  • That the cluster administrator account exists.

    The default cluster administrator is egoadmin.

Complete the following steps to deploy the software.

  1. Customize cluster properties.
  2. Run RPM.
  3. Set the command-line environment.
  4. Set up automatic system startup.
  5. Grant root privileges to a cluster administrator

Customize cluster properties

Customize the cluster properties at installation by setting variables to specify the cluster administrator, the cluster name, and the base connection port.

Setting these variables is optional; if the default settings are acceptable, you can install without configuring any of these variables.

What you need to know

  • Cluster name

  • Cluster administrator

  • Connection ports and base port

Set custom variables for installation

Set custom variables before installation if you wish to customize the cluster properties.

You can set environment variables according to your login shell. If you do not wish to use environment variables, create a simple text file /tmp/install.config and enter each variable on a new line. An environment variable is ignored if the same variable is set in the cluster properties configuration file.

  • For sh, ksh, or bash: export VARIABLE_NAME=value

  • For csh or tcsh: setenv VARIABLE_NAME value

  • In install.config: VARIABLE_NAME=value

  • To define the cluster name, set the CLUSTERNAME variable.

    Default cluster name: cluster1

    For example:

    setenv CLUSTERNAME cluster1

  • To define the cluster administrator account, set the CLUSTERADMIN variable.

    Default user account: egoadmin

    setenv CLUSTERADMIN user1

    For example:

  • To define the base connection port, set the BASEPORT variable.

    Default base port: 7869

    For example:

    setenv BASEPORT 7890

Run RPM

You can install two ways:

  • Install to the default installation directory.
  • Install to a custom installation directory.

Install to the default installation directory

Run RPM using the default installation options:

rpm -ivh package_name.rpm

package_name.rpm is the name of the RPM package.

For example:

rpm -ivh ego-linux2.4-glibc2.3-x86-1.2.3-nnnnnn.rpm

Install to a custom installation directory

Run RPM and specify the installation directory:
  • For RPM version 4.2.x or later:

    rpm -ivh --prefix install_dir package_name.rpm
    • install_dir is the installation directory

    • package_name.rpm is the name of the RPM package

    For example:
    rpm -ivh --prefix /opt/test/ ego-linux2.4-glibc2.3-x86-1.2.3-nnnnnn.rpm
  • For RPM version 4.1.x or earlier, if the --prefix option is not supported, set one more environment variable before you run the package:

    setenv RPM_INSTALL_PREFIX install_dir
    rpm -ivh package_name.rpm
    • install_dir is the installation directory

    • package_name.rpm is the name of the RPM package

    For example:
    setenv RPM_INSTALL_PREFIX /opt/test
    rpm -ivh ego-linux2.4-glibc2.3-x86-1.2.3-nnnnnn.rpm

The installer will create the installation directory if it does not already exist.

Set the command-line environment

On Linux hosts, set the environment before you run any EGO commands. You need to do this once for each session you open. Both root and egoadmin accounts use EGO commands to configure and start the cluster.

You need to reset the environment if the environment changes during your session, for example, if you run egoconfig mghost, which changes the location of some configuration files.

These examples assume the default installation directory /opt/ego.

  • For csh or tcsh, use cshrc.ego:

    source /share/ego/kernel/conf/cshrc.ego

  • For sh, ksh, or bash, use profile.ego:

    . /share/ego/kernel/conf/profile.ego

Set up automatic system startup

Enable EGO to start automatically when the host restarts.

Automatic startup

By default, you must start EGO manually if a host restarts.
Note:

For ease of administration, you should use egosetrc.sh to enable automatic startup. This feature starts EGO automatically when the host restarts.

Enabling automatic system startup creates an ego link under: /etc/rc.d/init.d

Set automatic startup on your host

Run the command egosetrc.sh.

Grant root privileges to a cluster administrator

Optional. A root user within a Linux environment can choose to give root privileges within the cluster to the cluster administrator.

Check the following:

  • That you are logged on as

    root
    .

  • That

    /etc/ego.sudoers
    does not already exist.

  • That the cluster administrator account exists.

    The default cluster administrator is

    egoadmin
    .

By default, only root can start, stop, or restart the cluster.

Give root privileges to egoadmin so that egoadmin can start a local host in the cluster, or shut down or restart any hosts in the cluster from the local host. For egoadmin or root to start the cluster, or start any hosts specified by name, you need to be able to run rsh across all hosts in the cluster without having to enter a password; see your operating system documentation for information about configuring rsh.

Do the following to give root privileges to egoadmin for one host. Run the command on each host in the cluster.

Run the egosetsudoers.sh command.

When you run egosetsudoers.sh, it does the following:

It creates the /etc/ego.sudoers file. The file owner is

root
and the permissions are set to
600
because you ran this command as
root
. Only the root user can edit this file.

It will setuid the egosh command and change the owner of egosh to root.

Whenever you see instructions to log on as root to start, stop, or restart a host in the cluster, you may log on as egoadmin instead.

Configure and test the management host

You have installed EGO on the host using the full package. You are logged on as egoadmin.

Complete the following steps to configure the host as a management host.

  1. Set the command-line environment.
  2. Join the cluster.
  3. Define the host as a management host.
  4. Set the command-line environment.
  5. Start the host.
  6. Restart the Linux cluster
    Note:

    You need to restart the cluster each time you add a management host to the cluster.

  7. Test that a management host was added to the cluster.

Set the command-line environment

On Linux hosts, set the environment before you run any EGO commands. You need to do this once for each session you open. Both root and egoadmin accounts use EGO commands to configure and start the cluster.

You need to reset the environment if the environment changes during your session, for example, if you run egoconfig mghost, which changes the location of some configuration files.

These examples assume the default installation directory /opt/ego.

  • For csh or tcsh, use cshrc.ego:

    source /share/ego/kernel/conf/cshrc.ego

  • For sh, ksh, or bash, use profile.ego:

    . /share/ego/kernel/conf/profile.ego

Join the cluster

Run egoconfig to join the cluster.

Always specify the name of the master host:

egoconfig join master_host_name

For example:

egoconfig join HostM

Define the host as a management host

Make the host a management host.

What you need to know

  • Host roles

  • Shared directory

Define a management host

Take this step on every management host, including all master candidates.

Run the egoconfig command to configure the host:

egoconfig mghost shared_dir

where shared_dir is the shared directory that contains important files such as configuration files to support master host failover.

For example:

egoconfig mghost /share/ego

After you run egoconfig mghost, the host:

  • Has access to important system files on the shared directory

  • Belongs to the ManagementHosts resource group.

Remember:

The shared directory is the same for all management hosts.

Set the environment to make changes take effect.

Set the command-line environment

On Linux hosts, set the environment before you run any EGO commands. You need to do this once for each session you open. Both root and egoadmin accounts use EGO commands to configure and start the cluster.

You need to reset the environment if the environment changes during your session, for example, if you run egoconfig mghost, which changes the location of some configuration files.

These examples assume the default installation directory /opt/ego.

  • For csh or tcsh, use cshrc.ego:

    source /share/ego/kernel/conf/cshrc.ego

  • For sh, ksh, or bash, use profile.ego:

    . /share/ego/kernel/conf/profile.ego

Start the host

Run egosh to start EGO on your host.
egosh ego start

Restart the Linux cluster

Log on with root permissions on the local host. If the master host is up, log onto any host in the cluster; if the master is down, log onto a management host.

Restart EGO on all hosts in the cluster:

egosh ego restart all

EGO restarts on all Windows and Linux hosts in the cluster. Services are not restarted with this command.

Test that a management host was added to the cluster

  1. Start the command console.
  2. Look for the host in the ManagementHosts group in your cluster:
    egosh resource group ManagementHosts
    If you can see the host name in the Resource List, the host is successfully added to the cluster and configured as a management host.
    Note:

    This test only detects hosts that are running.