When upgrading a clustered database deployment, all nodes in the cluster must be unclustered., We call it "disband the Cisco Meeting Server Cluster". before version 3.7, there is no warning message if the admin launches the upgrade process when the node is still clustered.
From version 3.7, Meeting Server will display a message to warn the admin about this requirement. The node in the cluster identifies if the database cluster is enabled in the server configuration and notifies the admin to disband the nodes before upgrading. Admin can choose to abort the upgrade process by pressing CTRL+C within 20 seconds.
Let's take an example:
First upgrade the Cisco Meeting management.
1-Take a Backup of all nodes Configuration. From Command Line CLI. Use the backup snapshot <filename> command.
2-Use WinSCP SFTP to Upload the upgrade.img file to all nodes.
3-Verify that the file is copied using the upgrade list command.
4-Disconnect the slaves CMS2 and CMS3 and the callbridge node CMS4 from the Primary node CMS1 using the database cluster remove command.
5-Disconnect the Primary database server CMS1 using the database cluster remove command.
6-Check again the database status to confirm using the database cluster status command.
7-Start with non-database servers (callbridge nodes) first, in this case CMS4. Run the command upgrade <filename>.
8-Go to the database slaves CMS2 and CMS3 and run the command upgrade <filename>.
9-Finaly go to the Master (Primary) database CMS1 and run the command upgrade <filename>.
10-Cluster the database, run the database cluster initialize command on the Primary server CMS1.
11-Join all slaves CMS2 and CMS3 to the primary CMS1 using the database cluster join <Database_primary_IP address> command.
12-Connect CallBridge node (non-database server) CMS4 to the primary using the database cluster connect <Database_primary_IP_address> command.
13-Verify the database status on all node using the database cluster status command.
14-Once all database servers are connected and in sync, on the primary database node CMS1, run the command database cluster upgrade_schema.
