Since 8.2 the solution for scaling of servers is done by the using of an AMQP server like RabbitMQ by the use of Broadcasting databroadcast and the Load balancer section.

In order to serve large number of users, multiple Servoy Application Servers can be clustered to works like one massive server.

To have an even user load among the servers in cluster, a load balancer is needed that is configured to be using sticky sessions.

Multiple server in a cluster also provide failover if one of the servers fails. But the sessions on a server in the cluster that does fail will not be recovered, they need to restart the solution, relogin. Servoy clustering is based on scaling not recover.

Quick Links: