Replacing Multiple Nodes in a Cluster

If the cluster itself is not healthy anymore or if multiple nodes need to be replaced, you must restore backup files taken around the same time to all of the nodes in the cluster.

Follow these steps to restore backups to multiple nodes in a cluster.

CAUTION

Backup files must be taken around the same time. If the backup file of one node is out of sync from the others, the restore process will be unsuccessful.

When restoring to multiple nodes, it is critical that you perform the restore process on all nodes at the same time.

Use the restore local command to restore the cluster from the backup file as you cannot restore the cluster by using a backup file from another cluster.

Log on to the CLI as a system administrator.
Run the enable command to enable privileged mode on the CLI.
```
ruckus> enable
Password: ********
ruckus#
```
Run the remote restore command to transfer the backup file from the FTP server to the controller.
```
ruckus# remote restore <ftp-username> <ftp-password> <ftp-server-address> <ftp-server-port> <directory>
idx version date
------------------------------------------
1 1.1.0.0.207 2012-10-16 06:46:07 GMT
2 1.1.0.0.209 2012-10-17 05:20:51 GMT
Please choose a backup version to get from remote FTP: 2
Remote restore process starts
Remote restore process completed
```
The ftp-server-port is optional.
NOTE
If there is only one backup file on the FTP server, the system will automatically transfer this file to the controller. If there are multiple files, it will show the list of all available files and you will be prompted to select the file that you want to transfer. If the backup files are in the root directory, use "/" in {directory}. If the backup files are in a subdirectory, use "/{subdir}/{subdir}" to indicate the subdirectory in which the system should check.

After all backup files for all nodes have been transferred from the FTP server to the controller, run the restore local command to restore the backup file to the controller.

ruckus# restore local
This action will REBOOT the system. Do you want to only restore this SCG node (yes/no)? yes
idx version date
------------------------------------------
1 1.1.1.0.93 2013-02-01 03:09:27 GMT 
2 1.1.1.0.93 2013-02-03 07:21:24 GMT
Please choose a backup version to restore (ex: 1, 2, ...): 2

Verify that the following message appears on each node:
Remote restore process completed

This indicates that the node is ready for the restore process.
Once all nodes are ready for the restore process, run the restore command for all nodes at the same time.

Contents

Replacing Multiple Nodes in a Cluster