Cluster Configuration

Upon agent installation, the node is considered part of its own 1-node cluster. The following configuration is merged in <OSVCETC>/cluster.conf:

cluster.secret is auto-generated, if not already defined.

cluster.name is set to default, if not already defined.

Bootstrap a new cluster

If the node joins an existing cluster, skip this section.

Add Heartbeats

On a new cluster, the heartbeats configuration need to be applied only on the first node. The joining nodes will fetch this configuration from this joined node.

For example, the simplest heartbeat configuration would be

om cluster set --kw hb#1.type=unicast

Display the result

om mon

Voir aussi

Agent Heartbeats

Add Stonith methods

Stonith is optional. Skip to the next section if not concerned.

On a new cluster, the stonith configuration can be applied on the first node. The joining nodes will fetch this configuration from this joined node.

For example, a dummy stonith configuration would be

om cluster set --kw stonith#node2.cmd=/bin/true

This configuration will execute /bin/true on the node taking over a service which was previously running on the now stalled node2.

Good, isolated fencing packages are freely available. For one, https://github.com/ClusterLabs/fence-agents

Add Arbitrators

Arbitrators are optional. Skip to the next section if not concerned.

The arbitrator configuration can be applied on any node of the cluster.

om cluster set --kw arbitrator#1.name=relay1 \
               --kw arbitrator#1.secret=10231023102310231023102310231023

This configuration will ask for the agent on node relay1 for its vote in a quorum race, if needed to get a majority.

The arbitrator#1.secret value comes from the cluster.secret value on the arbitrator relay1.

Voir aussi

Agent Daemon Quorum

Join a Cluster

The joining node can choose to join any of the cluster node already joined.

On the joined node node1

om cluster get --kw cluster.secret

On the joining node node2

om daemon join --secret <secret> --node node1

Note

  • If the node was frozen before the join, it is left frozen after the join.

  • If the node was not frozen before the join, the join process freezes it. If the join is successful, the node is thawed. If not, the node is left frozen.