we are using QlikView in a clustered architecture. There´s a loadbalancer in front of our two Webservers, just one Distribution Server and two QlikView Servers.
Today one of our QlikView Server had an error and died. The Server was down.
Meanwhile the Tasks tried to distribute applications and got an error too. Although the other QlikView-Server was Online.
Why? Because the Tasks try to distribute on both QlikView-Servers. If one of them is not reachable, the tasks stop imediately with an error. The application has even not been distributed on the running Server.
I think, this is not what a Clustering should do. Off course the distributed applications where still available on the running Server, but they have not been distributed with actual data.