Allows the administrator to stop or start a node in the cluster.
The -stop subcommand is used to take one or more nodes offline for maintenance. Stopping a node causes the other nodes to consider it as down. A stopped node does not send or receive heartbeat messages. It remains in the stopped state even across reboot operation until a -start subcommand causes it to rejoin the cluster. The -stop subcommand can also be issued while a node is powered off to prevent it from rejoining the cluster when it is rebooted.
Flag name | Description |
---|---|
-a | Causes the -start or -stop operation to be applied to all nodes in the cluster. The invoking node is -started at first and -stopped at the end. |
-m node[,...] | A comma-separated list of nodes to be started or stopped. Each node is attempted independently. That is, there is no functional difference between invoking the command once with a list of N nodes versus N times with one node each. |
-n clustername | The name of the cluster on which the operation is invoked. Since only one cluster is supported, specifying this option has no effect unless the cluster name is incorrect, in which case the command fails. |
-start | Causes the named nodes to start via a START_NODE protocol. Each successfully-started target node is configured to join the cluster automatically on subsequent boots. If the START_NODE protocol fails, the boot behavior of the node is unchanged. |
-stop | Causes the named nodes to stop via a STOP_NODE protocol. Each successfully-stopped node is configured such that it does not join the cluster automatically on subsequent boot. If the STOP_NODE protocol fails, the boot behavior of the node is unchanged. |
clstartstop -stop -n clustername -m nodeA
clstartstop -start -n clustername -m nodeA
clstartstop -stop -n clustername -a
clstartstop -start -n clustername -a