Deleting a Node from the Junos Space Fabric
You can delete a node from the Junos Space fabric directly by selecting the node and selecting Delete Fabric Node from the Actions menu. You must remove the deleted node from the network and reimage it. Then, you can add it to the fabric by selecting Administration > Fabric and the Add Fabric Node icon.
You cannot delete a primary Fault Monitoring and Performance Monitoring (FMPM) node if a secondary FMPM node exists. Junos Space Network Management Platform displays the following error message:
Primary FMPM node cannot be deleted if secondary FMPM node exist.
The workaround to delete the primary FMPM node is to perform one of the following actions:
Shut down the primary FMPM node and then delete the node.
Reboot the primary FMPM node and then delete the node. When you reboot this node, automatic failover takes place and the secondary FMPM node takes over as the primary FMPM node.
When you delete dedicated database nodes, you cannot delete both the primary and secondary database nodes from the fabric. You can delete either the primary database node or the secondary database node, but not both nodes.
You can delete a node from the fabric under the following conditions:
In a fabric with two or more nodes, if that node does not disrupt activities of other nodes.
If a node is configured for high availability—with load balancing and as a database server capability—and another node has the capacity to assume that role. You are prompted to enable that role on another candidate node before deleting that node. If you delete a high-availability node, but no node exists to which you can transfer that role, high availability does not occur.
When you delete a fabric node, Junos Space Platform performs the following tasks:
Removes references to the host name and IP address of that node from the remaining nodes
Stops database replication on both the deleted node and the backup database node
Makes the database backup copy in that node unavailable for the remaining nodes to restore the database from the backup copy
Copies the database to the new database node
Shuts down all services that interact with other nodes
When an FMPM node is deleted, the FMPM data from the FMPM node is first backed up and restored on the Junos Space node, and then the FMPM node is deleted from the Junos Space fabric. Thereafter, the network monitoring service is enabled on the Junos Space node.
You can delete only one node at a time. You must have Super Administrator or System Administrative role access privileges to delete a node.
To delete a node:
- Select Administration > Fabric.
- Select the node that you want to delete, and click the Delete Fabric Node icon.
- In the Warning dialog box, confirm that you want to delete
the node by clicking Continue.
If a node you want to delete is not configured for high availability or a node is configured for high availability but there is no other node available to assume that role, the Delete Node dialog box appears displaying the node name and management IP address of only the node that you want to delete.
If a node is configured for high availability, the Delete Node dialog box notifies you of that fact and lists all candidate nodes that have the capacity to assume that role.
When you delete a database node, only non-load-balancer nodes with the same configuration as the node you are deleting are listed as candidate nodes.
If a node hosts one or more virtual machines, then the warning message also indicates the IP addresses of the virtual machines that will be deleted.
- In the Delete dialog box, select the node that you want to delete.
- Click Delete.
Node deletion is scheduled as a job immediately after you click Delete. Deleting a node generates an audit log entry. The Delete Fabric Node Job Information dialog box appears.
- In the Delete Fabric Node Job Information dialog
box, click the Job ID link.
The Job Management inventory landing page appears displaying this job. From this page, you can verify and monitor information about the node you are deleting, such as the job type, job ID, percentage of task completion, job state, scheduled start and end times, username, a brief job summary, and so on.
When you delete a node, a UDP communication exception occurs. This behavior is normal.
When you delete a load balancer node, a VIP switch may occur and cause the Junos Space Platform progress indicator to appear. This behavior is normal.