Reimage a Node and Add the Node Back with the Same IP Address
Perform the following procedure to reimage a node and add the node back with the same IP address when the primary database node in the Disaster Recovery (DR) cluster goes down:
- Reset the DR when the primary database node goes down on the active site.
- Delete the node that is down by selecting to transfer its role to other JBoss node in the cluster.
- Reimage the new Virtual Machine (VM) with the IP address of the deleted node and add it back to the cluster.
-
When the setup goes into maintenance mode, cluster formation stops. To create the
cluster again, perform the following:
Note:
You must create the cluster formation between the JBoss VIP node and the node that took the primary database role in Step 2.
Run the following command on the initial Virtual IP address (VIP) node before addition of nodes:
pcs cluster stop --force pcs cluster destroy
Run the following command on the VIP node:
/usr/bin/systemctl start corosync pcs cluster setup jmp-CLUSTER <node-hostname>
Run the following command to authorize the host on both VIP node and the node that took the primary database role in step 2:
pcs host auth <other node's hostname>. Username: hacluster Password: jnpr123! pcs cluster start --all pcs cluster enable
Run the following command to check pcs cluster status:
pcs cluster status Now On other node: pcs cluster start
Check the
pcs cluster status
on both VIP node and the node that took the primary database role in step 2. Both the nodes must be in active status.To add a Virtual IP address, run the following command:
pcs resource create virtual_ip ocf:heartbeat:IPaddr2 ip=<ip address> cidr_netmask=<subnet> nic=eth0:0 op monitor interval=30s pcs resource create before_vip lsb:beforeSwitchingVIP
To start the pacemaker, run the following command:
/usr/bin/systemctl start pacemaker
Run the following command when the pcs resource status for the VIP stops:
pcs property set stonith-enabled=false Chk ifconfig for VIP association on all nodes
To remove a node from the cluster, run the following command:
pcs cluster node remove <node-hostname>
You may need to restart the services. Use the following command to restart services on all the nodes:
systemctl stop jmp-watchdog systemctl stop jboss systemctl stop jboss-dc
Check for the https services on the VIP node. If the service is down, bring it up.
The cluster is now up and running.
Configure and start the DR.