Specifications
Chapter 15: Advanced Administration
255
For Time Interval, enter how often CC-SG should check its
connection with the other nodes. Valid range is 5-20 seconds.
Note: Setting a low Time Interval will increase the network traffic
generated by heartbeat checks. You may want to set higher intervals
for clusters with nodes located far apart from each other.
For Failure Threshold, enter the number of consecutive
heartbeats that must pass without a response before a CC-SG
node is considered failed. Valid range is 2-10 heartbeats.
3. Click Update to save your changes.
Switch the Primary and Secondary Node Status
You can exchange the roles of Primary and Secondary nodes when the
Secondary, or Backup, node is in the "Joined" state. When the
Secondary node is in the "Waiting" state, switching is disabled.
After the roles are switched, the former Primary node is in the "Waiting"
state. To recover the cluster configuration, join the "Waiting" node as the
Backup.
See Recover a Cluster (on page 255).
To switch the Primary and Secondary nodes
1. Choose Administration > Cluster Configuration.
2. In the Configuration tab, click Switch Primary And Backup.
3. Join the new Secondary node as the Backup node. See Recover a
Cluster (on page 255).
Recover a Cluster
When a cluster is broken because of a node failure, or the failed
Secondary node is in Waiting status, you can rebuild the cluster to
recover the Primary and Secondary node status.
If the Primary and Secondary nodes lose communication with one
another, the Secondary node will assume the role of the Primary node.
When connectivity resumes, you may have two Primary nodes. It is
impossible to recover the cluster with two Primary nodes. Recover works
only when there is one Primary and one Waiting.
To recover a cluster with two Primary nodes, you have two options. Log
into each Primary node, delete the cluster in each, and then create the
cluster again. Or, log into one of the Primary nodes and restart it so that
it changes to the Waiting status, then follow the instructions to recover a
cluster.
To recover a cluster:
1. Choose Administration > Cluster Configuration.