Data Replication and Failover Protection

When you install an MCS server cluster, all configuration changes take place through the cluster's primary server. Configuration data, event log files, and trace log files are dynamically replicated across all servers in the cluster, either every ten minutes or when data is changed through the primary server, ensuring that all active servers have access to the latest data.

When a change is made through the primary server, MCS notifies all other servers to start replication. This replication occurs after a one minute delay so that multiple changes can be replicated at the same time. This optimizes replication, but it also means that the servers can be out of sync for 60 seconds while the replication process completes.

Failover Protection

The status of the primary server is monitored by a managed service in MCS. If the primary server ever goes off-line, the server with next highest priority in the cluster takes over primary server responsibilities. Several things happen to minimize data loss and ensure the availability of MCS:

• The server that detected the failure of the primary server determines which server has next highest priority and should take over primary server responsibilities.

• It also determines which server has the most up-to-date data (in case a recent change has not been replicated to all servers in the cluster).

• The server that takes over interim primary server responsibilities updates its data from the most up-to-date data, and replicates that data to all other servers in the cluster.

• It also notifies the other servers that it is now acting as primary server.

Unless the priority of the primary server is changed through the MCS console, when the original primary server comes back on-line, it resumes primary server responsibilities.

	Overview of Clustering
	Managed Services in MCS
	Configuring Failure Detection
	Changing the Priority of a Clustered Server