Appendix : Administration : MashZone NextGen Server Administration : Deploying MashZone NextGen Instances, Clusters or Artifacts : MashZone NextGen Dashboard in a clustered scenario : Configuration : Load balancer
Load balancer
"sticky sessions" mode
The load balancer must run in "sticky sessions" mode. That means that all subsequent requests belonging to the same session will be send to the same server that handled the first request. The load balancer must be configured accordingly.
WebSocket communication
MashZone NextGen uses sockJS for the client - server communication. Requests that end with "/websocket" have to be routed to MashZone NextGen nodes, using ws/wss as protocol. If the load balancer can not handle the websocket communication (e.g., HTTP header connection: upgrade, upgrade: websocket), sockJS uses different HTTP fallback protocols.
Fail-over mode
In fail-over mode, only one server handles client requests and the second server just runs in background. If the first server fails, the second one takes over all traffic and handles the requests until the main server is available again. In this mode, the two servers have to be more or less identical installations with both, RTBS servers running and both connected to the same database. RTBS configuration is as it would be in a single server scenario.
Note:  
The RTBS traffic must also be mounted to the balancer worker in this case, not to the worker node that is the master instance.
Copyright © 2013-2017 Software AG, Darmstadt, Germany.

Product LogoContact Support   |   Community   |   Feedback