36.0 System Requirements

When allocating cluster resources to support a high availability (HA) installation, consider the following requirements:

  • (Conditional) For HA appliance installations, ensure that the Sentinel HA appliance with a valid license is available. The Sentinel HA appliance is an ISO appliance that includes the following packages:

    • Operating system: SLES 12 SP3

    • SLES High Availability Extension (SLES HAE) package

    • Sentinel software (including HA rpm)

  • (Conditional) For traditional HA installations, ensure that the following are available:

    • Operating system: SLES 11 SP4 or SLES 12 SP1 or later

    • SLES HAE ISO image with valid licenses

    • Sentinel installer (TAR file)

  • (Conditional) If you are using the SLES operating system with kernel version 3.0.101 or later, you must manually load the watchdog driver on the computer. To find the appropriate watchdog driver for your computer hardware, contact your hardware vendor. To load the watchdog driver, perform the following:

    1. In the command prompt, run the following command to load the watchdog driver in the current session:

      /sbin/modprobe -v --ignore-install <watchdog driver name>
    2. In the /etc/init.d/boot.local file, add the following line to ensure that the computer automatically loads the watchdog driver at every boot time:

      /sbin/modprobe -v --ignore-install <watchdog driver name>
  • Ensure that each cluster node that hosts the Sentinel services meet the requirements specified in Section 5.0, Meeting System Requirements.

  • Ensure that sufficient shared storage is available for the Sentinel data and application.

  • Ensure that you use a virtual IP address for the services that can be migrated from node to node on failover.

  • Ensure that your shared storage device meets the performance and size characteristics requirements specified in Section 5.0, Meeting System Requirements. Use a standard SLES virtual machine configured with iSCSI Targets as shared storage.

    For iSCSI, you should use the largest Message Transfer Unit (MTU) supported by your hardware. Larger MTUs benefits the storage performance. Sentinel might experience issues if latency and bandwidth to storage is slower than recommended.

  • Ensure that you have a minimum of two cluster nodes that meet the resource requirements for running Sentinel in the customer environment. Two SLES virtual machines is recommended.

  • Ensure you create a method for the cluster nodes to communicate with the shared storage, such as FibreChannel for a SAN. Use a dedicated IP address to connect to the iSCSI Target.

  • Ensure you have a virtual IP address that can be migrated from one node to another node in a cluster to serve as the external IP address for Sentinel.

  • Ensure you have at least one IP address per cluster node for internal cluster communications. You can use a simple unicast IP address, but multicast is preferred for production environments.