InfoScale™ 9.0 Cluster Server Generic Application Agent Configuration Guide - AIX, Linux, Solaris

Last Published:
Product(s): InfoScale & Storage Foundation (9.0)
Platform: AIX,Linux,Solaris

Sample configuration for an init process and generic application component

This section describes a sample procedure for using the Arctera High Availability Configuration wizard.

To configure monitoring for the following across two virtual machines - Machine1 and Machine2:

  • an init process, such as CUPS

  • a generic application, MyApplication

As part of the configuration process, the wizard configures a 2-node cluster between the machines Machine1 and Machine2 running on hosts - Host1 and Host2, respectively.

Let us assume that the generic application (MyApplication) can be started, stopped, monitored, and forcibly stopped by using the following scripts, respectively:

  • start_MyComponent

  • stop_MyComponent

  • monitor_MyComponent

    The monitor_MyComponent script is written to comply with the MonitorProgram attribute of generic applications. For more information, see the description of the Application agent attributes in the Cluster Server Bundled Agents Reference Guide.

  • forcestop_MyComponent

To configure application monitoring using the Arctera High Availability Configuration wizard

  1. Launch the Arctera High Availability Configuration wizard.
  2. Review the information on the Welcome panel and click Next.
  3. Select Generic Application, and then click Next.
  4. On the Component Selection panel, enter a name for the CUPS process, for example, cups_Program and click Add Component.

    The component you added (cups_Program) appears in the Component box.

  5. Specify the following details to configure the cups_Program for monitoring:

    Start program

    /etc/init.d/cups start

    Stop program

    /etc/init.d/cups stop

    Force-stop program

    Not specified

    Monitor program

    /etc/init.d/cups status

    Note:

    init processes such as CUPS, do not require special monitor scripts. VCS uses the status option of the init script for monitoring. However you can also use your own program scripts to monitor such processes.

    Application-related processes to monitor

    cupsd -C /etc/cups/cupsd.conf

    Application-generated PID files

    /var/run/cupsd.pid

    Enable intelligent monitoring for this application

    Selected by default to enable intelligent monitoring

    User

    username. For example, root.

    Note:

    You must specify at least one or more of the following: Monitor program, application-related processes to monitor, application-generated PID files.

  6. To configure MyApplication for monitoring, add a name for the MyApplication component, for example, MyComponent and click Add Component.

    The component you added (MyComponent) appears in the Component box.

  7. Specify the following details to configure MyComponent for monitoring:

    Start program

    /myapplication/bin/start_MyComponent

    Stop program

    /myapplication/bin/stop_MyComponent

    Force-stop program

    /myapplication/bin/forcestop_MyComponent

    Monitor program

    /myapplication/bin/monitor_MyComponent

    Application-related processes to monitor

    Not specified

    Application-generated PID files

    Not specified

    Enable intelligent monitoring for this application

    Selected by default to enable intelligent monitoring

    User

    username. This user name is used to run all the agent scripts. For example, root.

  8. Click Next.
  9. If the MyApplication application requires storage, on the Storage Selection panel, select the appropriate mount points and click Next.
  10. On the Define Start-Stop Order panel, you can define the relationship between the CUPS process and MyApplication.

    To bring the CUPS process online first and then MyApplication, in the Parent Component list, select MyComponent and in the Depends on box, select cups_Program.

  11. Click Next.
  12. On the Configuration Inputs panel, the wizard lists Machine1 - the machine from which you launched the wizard. The wizard also lists Machine1 in the Application failover targets list. To add Machine2 to the cluster, click Add System, and in the Add System dialog box, specify the following details for Machine2:

    System Name or IP address

    Machine2

    User name

    username

    Typically, this is the root user.

    Password

    password

    Use the specified user account on all systems

    Select to use the specified user account on all the cluster systems.

  13. Click Next.
  14. On the Virtual Network Details panel, select MyComponent, and then select IPv4 and specify the following details for each failover system:

    Virtual IP address

    IP address

    Subnet Mask

    Subnet mask

    Network Adapter

    For Machine1: eth0

    For Machine2: eth1

  15. Click Next.
  16. On the Summary panel, review the VCS cluster configuration summary and then click Next to proceed with the configuration.
  17. On the Implementation panel, the wizard creates the cluster, configures application monitoring, and creates cluster communication links. The wizard displays the status of each task. After all the tasks are complete, click Next.
  18. Click Finish to complete the wizard workflow.

    This completes the application monitoring configuration.

    If the application status shows as not running, click Start to start the configured components on the system.