cluster

Generate clusters that optimize modularity and print the result to stdout

$ cluster [input.gv]

$ cluster -C [5] [input.gv]

$ cluster -c [0|1] [input.gv]

$ cluster -o [output.gv] [input.gv]

$ cluster -v [input.gv]

cluster <command> [<options>] [<arguments>]
cluster {version|status|ccs|node|rgmanager|config|fence|...}

Note: This command typically operated with specific subcommands, each having its own set of options and arguments.

status
    Displays the current status of the cluster, including node health and resource group states.

version
    Shows the version information for the cluster tools.

ccs
    Provides an interface to the Cluster Configuration System, allowing for configuration retrieval and updates.

node
    Manages cluster nodes, including displaying node status and performing node-specific actions.

rgmanager
    Controls the resource group manager, used for starting, stopping, and managing cluster services (resource groups).

config
    Allows for manipulation of cluster configuration files.

fence
    Manages fence devices and fencing operations within the cluster.

The cluster command was a central command-line interface for managing components of the Red Hat Cluster Suite (RHCS), including CMAN (Cluster Manager) and rgmanager (resource group manager). It provided subcommands to check cluster status, manage cluster nodes, configure cluster services (resources), and control the cluster daemon.

While crucial in its time for setting up and maintaining high-availability and load-balancing clusters, it has largely been superseded by modern cluster management tools like pcs (Pacemaker/Corosync Command-Line Interface) and is considered deprecated on contemporary Linux distributions.

Its primary role was to interact with the underlying cluster infrastructure, allowing administrators to query cluster state, start/stop services, and perform administrative tasks on cluster members.

The cluster command is largely deprecated and is not present on modern Linux distributions by default. It was primarily used with the Red Hat Cluster Suite (RHCS) and its underlying CMAN/rgmanager components. Modern high-availability clusters on Linux typically use Pacemaker and Corosync, managed via the pcs command. Attempting to use cluster on a system not configured with RHCS will result in a "command not found" error or unexpected behavior. Functionality described applies to its historical usage.

The cluster command did not itself run the cluster; rather, it interacted with daemon processes like cman (for cluster membership and messaging) and rgmanager (for resource and service management). It provided a user-friendly abstraction layer over these complex background services.

Cluster configuration was traditionally stored in /etc/cluster/cluster.conf, often managed by the ccs (Cluster Configuration System) utility, which the cluster command could interface with.

The cluster command gained prominence as part of the Red Hat Cluster Suite (RHCS) in the early 2000s, serving as the primary administrative interface for managing high-availability and load-balancing clusters on Red Hat Enterprise Linux and its derivatives. It was instrumental for administrators managing mission-critical services by orchestrating resources and ensuring failover capabilities through CMAN and rgmanager.

Its usage began to decline with the shift towards the Pacemaker and Corosync stack as the industry standard for open-source high-availability, which offered a more flexible and robust architecture. By the mid-2010s, pcs emerged as the preferred command-line tool for these new cluster technologies, effectively rendering the original cluster command obsolete for new deployments.

pcs(8): The modern command-line interface for Pacemaker/Corosync clusters., cman(5): Cluster Manager (the daemon often managed by this command)., rgmanager(5): Resource Group Manager (the daemon for resource groups)., corosync(8): A core component of modern cluster infrastructure., pacemaker(8): A key component of modern cluster infrastructure.