HA/CDP-Series

	ClusterReplica ENT
	ClusterReplica SQL
	EzOpenBackup Plus!
	A Success Story
	Order/Price
Omni-NFS-Series

 System Failback

Failback is the process to switch the role of Primary station from the current ACTIVE system to the current STANDBY system or a newly turned-on system.

To understand ClusterReplica's Failback process, two concepts related to a cluster system need to be clarified:

1. Default Primary Station - This setting servers as a marker to remind you that the system is originally intended as the Primary station.
2. Data flow direction - Data flow in a cluster system is normally from the Primary station to the Secondary station.

When trying to Failback, the main issue that would concern you is to keep server function with all current data available. A successful Failback requires having all current data copied onto the newly switched on system from the current Primary station.

When dealing with Failback, two situations are to be consider:

1. The newly turned-on system was originally set as the default primary station
2. The newly turned-on system was NOT originally set as the default primary station

In the first situation, when the new system is plugged back in the cluster system, you will be asked if you still want to run it as the Primary station. Keep in mind that the data flow of the cluster system is always from Primary to Secondary system. This means, if you choose to set the newly turned-on system as the Primary station, all data on the current Primary station will be lost.

Therefore, it is important that when you first connect the new system into the cluster system, set it as the Secondary station so that you updated files can be copied to it before you switch it to become the Primary station.

In the second situation, no question will be asked. This is so because it is assumed that the newly turned-on system was inintially intended to be used as the Secondary station. It will stay to be the Secondary station when re-join the cluster system.

This is an example in which the clustered database server is in Failback state. All new files and file updates are copied back to the Primary server to get it ready to resume its ACTIVE duty.

The Secondary server will remain "active" until all files are copied back to the Primary server.

Making sure of the proper server configurations of the fixed server before reconnecting it to the cluster system is important.