Triangle-based obstacle-aware load balancing for massively multiplayer games

Massively multiplayer games run on large server farms in order to handle the huge load that is caused by the thousands of players that play these games concurrently. Current systems use very simple load-balancing mechanisms restricting the semantics of the game. In this paper, we present a dynamic load balancing mechanism that considers both the load associated with performing game actions as well as the load incurred through interest management. As clients join, they can take on load due to game actions and update notifications. When players flock to some part of the world, our system is able to dynamically redistribute the imbalanced interest management load among servers. This hybrid mechanism is integrated into the Mammoth multiplayer gaming environment and our real-world experiments show how equal load distribution is maintained under changing workloads.