Remove nodes from a cluster setup
As an RPA platform administrator, you can remove nodes from a Control Room cluster setup to replace them or upgrade them for enhanced performance.
Prerequisites
Ensure the following:
- Verify that all the primary nodes and databases have backups.
- Ensure that you have administrator or root privileges on the Control Room servers.
Procedure
-
Run the following command in the Linux shell to identify the master node.
curl -k --user es_client:Automation123 https://172.31.46.2:47599/_cat/nodes
-
Remove the non-master nodes N2 and N3 from the cluster.
-
Log in to the server as an administrator and run the following command
to stop all the Control Room services.
sudo systemctl stop controlroom*
- Log in to any of the non-master nodes.
- Edit the cluster.properties file located at /opt/automationanywhere/automation360/config.
-
Remove the IP addresses for the original three nodes.
Note: Perform this action for all the nodes in the cluster. When you remove the IP addresses of the original three nodes, do not change their sequence.
- Edit the elasticsearch.yml file at /opt/automationanywhere/automation360/elasticsearch/config.
- Remove the old IP addresses from the list in the discovery.zen.ping.unicast.hosts attribute.
- Restart the non-master nodes, one at a time.
-
Log in to the server as an administrator and run the following command
to stop all the Control Room services.
-
Remove the master node N1 from the cluster.
-
Log in to the server as an administrator and run the following command
to stop all the Control Room services.
sudo systemctl stop controlroom*
- Log in to the any of the non-master nodes.
- Edit the cluster.properties file and remove the IP address of the old master node.
-
Run the following command in any of the other nodes to identify the new
master node.
curl -k --user es_client:<es password>https://172.31.18.37/_cat/nodes
- Edit the elasticsearch.yml file at /opt/automationanywhere/automation360/elasticsearch/config.
- Update the IP address of the new master node in the cluster.initial_master_nodes attribute.
- Remove the IP address of the old master node from the list in the discovery.zen.ping.unicast.hosts attribute.
-
Run the following command to start the services on each node:
sudo systemctl start controlroom*
-
Log in to the server as an administrator and run the following command
to stop all the Control Room services.
-
To verify the health of the cluster, run the following command in the command
line:
curl -k --user es_client:<es password>https://172.31.18.37/_cat/nodesYou must wait until the replication status turns green. The replication status turns green when the cluster is fully synchronized.