High Availability

This section discusses High Availability (HA). High Availability is a system characteristic that allows it to maintain high levels of uptime. This can be valuable in file transfer systems that serve thousands of users, support supply chains, or perform mission-critical data transfers. A highly available file transfer system is perfect for businesses who have very low tolerance for downtime.

 

In MFT Server, high availability is typically implemented by building either active-active or active-passive high availability cluster configurations. Active-active MFT server HA clusters typically consist of two (2) or more simultaneously active MFT Server nodes and are mainly used for achieving load balancing. On the other hand, active-passive HA clusters typically consist of one active MFT Server node and one (1) or more passive nodes. The purpose of an active-passive configuration is to have a readily available node that can take over once the active node fails.

 

A key requirement for setting up high availability clusters of MFT Server instances is a centralized global datastore. This datastore can be any robust RDBMS such as MySQL, Oracle, or MS SQL Server.

 

Note: The use of a shared/centralized global datastore is not suitable when nodes must bind services to specific IP addresses. The shared/centralized global datastore setup will only work when all services are bound to IP address 0.0.0.0, which means services will be listening on all network interfaces available on the node.

 

See Configuring a centralized global datastore.