The QlikView Server documentation implies that clustering is done for load balancing and that all documents would be available on all nodes of the cluster, as if the cluster acts like one server.
However, we're interested in having individual QV servers for different departments of an organization, but they would like to share their named user CALs across all the servers rather than have each server have its own user CALs.
Effectively, what we want is separate servers that share a common pool of CALs.
Is this possible using clustering? What effects does this have on manageability, security, etc?