You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Pablo J. Rogina edited this page Dec 13, 2017
·
1 revision
In order to provide high availability for your solution based on netty-socketio server, the suggestion is to use several netty-socketio instances and any load-balancing proxy (i.e. NGINX with iphash, HAProxy) on top of your netty-socketio cluster.
This way, as soon as one server goes down, the client will try to reconnect and the load-balancer will redirect it to the other running server. From client point of view, there won't be much delay in that redirection/new connection. A good description of use case scenario and solution is provided here.
Take into account that once you have the cluster of netty-socketio servers setup for high availability, just by adding a Redisson or Hazelcast server, you can easily get the capability of broadcasting messages across multiple servers as described here