Cluster monitoring

Clusters simply mean the collection of nodes that communicate with each other. Cluster monitoring is important to validate the efficient working of the individual nodes in the clusters or the clusters collectively.

Veritas Linux cluster monitoring

The Veritas Linux cluster monitoring includes monitoring cluster parameters such as cluster node state, service group state, resource state, and service group failover status.

Prerequisites

  1. Veritas cluster setup on Linux.
  2. Since the templates are based on “Agent-based G2 Custom Monitors”, Root agent needs to be installed on all the cluster nodes.

Supported Metrics

Click here to view the supported metrics

Template NameMonitor NameMetric Name
Agent G2 - Linux Veritas Cluster MonitoringG2 - Linux Veritas Cluster Group Failover Monitorsystem_linux_veritas_cluster_group_failover_status
G2 - Linux Veritas Cluster Monitorsystem_linux_veritas_cluster_group_online_status
system_linux_veritas_cluster_group_state
system_linux_veritas_cluster_node_state
system_linux_veritas_cluster_resource_state
system_linux_veritas_cluster_resource_online_status

Supported versions

Veritas Cluster software (Veritas Infoscale 7.3.1 version) running on CentOS 7.

Veritas Linux cluster parameters

  • Veritas cluster group online status: Monitors the cluster group running on the nodes. The metric graphs display one of the following values:
    • 0 - Service group online on a cluster node.
    • 1 - Service group is not online on a cluster node.

  • Veritas cluster group failover status: Validates whether the cluster groups are running on the preferred owner nodes. The system generates critical alerts otherwise and the metric graphs display one of the following values:
    • 0 - No change.
    • 1 - Cluster group change from one node to another due to failover.
    • 2 - The specific cluster group is not online on a cluster node.

  • Veritas cluster group state: Monitors the Veritas cluster group current state. The different states are:
    • OFFLINE
    • ONLINE
    • FAULTED
    • PARTIAL
    • STARTING
    • STOPPING
    • MIGRATING
    • OFFLINE|FAULTED
    • OFFLINE|STARTING
    • PARTIAL|FAULTED
    • PARTIAL|STARTING
    • PARTIAL|STOPPING
    • ONLINE|STOPPING

  • Veritas cluster node state: Validates whether the cluster node is in RUNNING state and raises critical alerts if the state varies. Different cluster node states are:
    • RUNNING
    • ADMIN_WAIT
    • CURRENT_DISCOVER_WAIT
    • CURRENT_PEER_WAIT
    • EXITING
    • EXITED
    • EXITING_FORCIBLY
    • FAULTED
    • INITING
    • LEAVING
    • LOCAL_BUILD
    • REMOTE_BUILD
    • STALE_ADMIN_WAIT
    • STALE_DISCOVER_WAIT
    • STALE_PEER_WAIT
    • UNKNOWN

  • Veritas cluster resource online status: Monitors the cluster resource status and generates critical alerts if the cluster is not in the ONLINE state. The metric graphs display one of the following values:
    • 0 - Resource state is online on a cluster node.
    • 1 - Resource state is not online on a cluster node.

  • Veritas cluster resource state: Monitors the current state of the cluster resource. Different cluster states are:
    • OFFLINE
    • ONLINE
    • FAULTED
    • PARTIAL
    • STARTING
    • STOPPING
    • MIGRATING
    • OFFLINE|FAULTED
    • OFFLINE|STARTING
    • PARTIAL|FAULTED
    • PARTIAL|STARTING
    • PARTIAL|STOPPING
    • ONLINE|STOPPING

Constraint

Generating duplicate alerts for the same issues after applying templates on all Veritas Linux cluster nodes.