High availability deployment and best practices

A fault-tolerant (high-availability (HA)) deployment safeguards you against any unplanned downtime of a member node. If one member node of the cluster goes down, the connection is automatically diverted to the other available nodes. This topic walks you through some best practices for such fault-tolerant BMC Client Management server cluster deployment.

For information about installing a BMC Client Management cluster for high availability, see Installing a BMC Client Management cluster for high availability.

BMC Client Management high availability architecture

The basic concept of a BMC Client Management fault-tolerant cluster deployment is to have the service run on individual member nodes and use a common master file directory from a shared location. The following diagram provides a high-level view of the architecture:

High availability  best practices

When deploying this cluster architecture, ensure the following:

• Minimum two servers as member nodes for the cluster with individual static IP address.
• Only one master directory on a shared location, which must be accessible to each member node in the cluster.
• A cluster management system to ensure that only one member node has the service running at any given time.
  • For Windows: See https://docs.microsoft.com/en-us/windows-server/failover-clustering/create-failover-cluster.
  • For Linux: Different distributions have different high-availability (HA) suites in their repositories. See the corresponding documentation or consult technical support for the distribution.

  Note

  You can manage the cluster by starting and stopping the service on each member node manually. However, it is not recommended for production environment.

• All the member nodes have Java installed as per software requirement. It is recommended that all the member nodes have the same Java version.