Cloudify cluster day 2 operations
Introduction
Cloudify HA solution can contains three different clusters of database nodes, message queue brokers and Cloudify Managers. This require the ability to manage those cluster, which include add/removing/listing/custom-action to the cluster.
Cloudify management service operations
For removing a management service cluster node, please run on a management node (not the one you wish to remove):
cfy cluster remove <A resolvable hostname>
For adding a management service cluster node, please install the new node on the same network settings as in Cloudify management service nodes.
Message queue cluster operations
A new broker node (an installed node) can be added to the cluster by running the following:
Run the following command on the newly installed node to connect a broker to the cluster:
cfy_manager brokers add -j\--join-node <A resolvable hostname or IP of an existing member of the cluster>
Run the following command to register the new broker node at the Cloudify management service:
cfy cluster brokers add <new broker name> <new broker address>
Listing all the current message queue brokers in the cluster is retrieved by running the following command:
cfy_manager brokers list
An existing broker node can be removed from the cluster by running the following command:
- First the broker node should be uninstalled, please run the following:
cfy_manager remove
Run the following command to remove the broker from the cluster, which should be run from one of the remaining cluster nodes:
cfy_manager brokers remove -r\--remove-node <nodename>
Run the following command to unregister the new broker node at the Cloudify management service:
cfy cluster brokers remove <broker name>
- First the broker node should be uninstalled, please run the following:
Database cluster operations
Listing all the current database nodes in the cluster is retrieved by running the following command:
cfy_manager dbs list
A new database node (an installed node) can be added to the cluster by running the following command:
Database’s node id can be obtained by running
cfy_manager node get-id
.cfy_manager dbs add -a\--address <Node address> -i\--node-id <Node's cloudify node id> -n\--hostname <node's hostname>
An existing database node can be removed from the cluster by running the following command:
Database’s node id can be obtained by running
cfy_manager node get-id
.cfy_manager dbs remove -a\--address <node's address> -i\--node-id <Node's cloudify node id>
If a database node becomes unhealthy according to the cluster status, it can be re-initialized by running the following command:
cfy_manager dbs reinit -a\--address <node's address>
For switching the current database master node in the database cluster by running the following command:
cfy_manager dbs set-master -a\--address <node's address>
In case there are changes to the cluster structure (a node has been added or removed) and Cloudify’s Status Reporter is among the installed services, trigger a monitoring configuration update:
cfy cluster db-nodes update