Transaction Models for Massively Multiplayer Online Games

Massively Multiplayer Online Games are considered large distributed systems where the game state is partially replicated across the server and thousands of clients. Given the scale, game engines typically offer only relaxed consistency without well-defined guarantees. In this paper, we leverage the concept of transactions to define consistency models that are suitable for gaming environments. We define game specific levels of consistency that differ in the degree of isolation and atomicity they provide, and demonstrate the costs associated with their execution. Each action type within a game can then be assigned the appropriate consistency level, choosing the right trade-off between consistency and performance. The issue of durability and fault-tolerance of game actions is also discussed.

[1]  Johannes Gehrke,et al.  Scalability for Virtual Worlds , 2009, 2009 IEEE 25th International Conference on Data Engineering.

[2]  Kaiwen Zhang,et al.  Persistence in massively multiplayer online games , 2008, NETGAMES.

[3]  Tristan Henderson,et al.  Latency and User Behaviour on a Multiplayer Game Server , 2001, Networked Group Communication.

[4]  Alan Fekete Weak Consistency Models for Replicated Data , 2009, Encyclopedia of Database Systems.

[5]  Clark Verbrugge,et al.  A peer auditing scheme for cheat elimination in MMOGs , 2008, NETGAMES.

[6]  Philip A. Bernstein,et al.  Relaxed-currency serializability for middle-tier caching and replication , 2006, SIGMOD Conference.

[7]  Jim Gray,et al.  A critique of ANSI SQL isolation levels , 1995, SIGMOD '95.

[8]  Jörg Kienzle,et al.  Comparing interest management algorithms for massively multiplayer games , 2006, NetGames '06.

[9]  Srinivasan Seshan,et al.  Donnybrook: enabling large-scale, high-speed, peer-to-peer games , 2008, SIGCOMM '08.

[10]  Srinivasan Seshan,et al.  Colyseus: A Distributed Architecture for Online Multiplayer Games , 2006, NSDI.

[11]  Sergei Gorlatch,et al.  High-Level Development of Multiserver Online Games , 2008, Int. J. Comput. Games Technol..

[12]  Adi Botea,et al.  Near Optimal Hierarchical Path-Finding , 2004, J. Game Dev..

[13]  Jin Chen,et al.  Locality aware dynamic load management for massively multiplayer games , 2005, PPoPP.

[14]  Mihai Burcea,et al.  Towards scalable and transparent parallelization of multiplayer games using transactional memory support , 2010, PPoPP '10.

[15]  Parag Agrawal,et al.  Asynchronous view maintenance for VLSD databases , 2009, SIGMOD Conference.

[16]  Johannes Gehrke,et al.  SEMMO: a scalable engine for massively multiplayer online games , 2008, SIGMOD Conference.

[17]  Jörg Kienzle,et al.  Mammoth: a massively multiplayer game research framework , 2009, FDG.