The game discovery process of many network games involves game clients querying one or more master servers who hold a list of all currently active game servers. In this report we investigate the manner in which a Wolfenstein Enemy Territory (ET) master server chooses to maintain and report the list of currently active ET servers around the Internet. We note how the ordering of game servers in the master server's reply influences game server selection by clients, and show that the master server periodically re-orders the list of active game servers. Using data gathered between late 2005 and early 2006 we observe patterns of periodic re-ordering that ensure every active game server has an equal chance of being polled by game clients, regardless of the game server's location on the Internet. To reduce the probe traffic on game servers we believe some form of filtering is desirable to bias game client probing towards closer/local game servers.
[1]
J. Meigs,et al.
WHO Technical Report
,
1954,
The Yale Journal of Biology and Medicine.
[2]
Sebastian Zander,et al.
Dissecting server-discovery traffic patterns generated by multiplayer first person shooter games
,
2005,
NetGames '05.
[3]
Sebastian Zander,et al.
Client RTT and hop Count Distributions viewed from an Australian ‘Enemy Territory’ Server
,
2006
.
[4]
David A. Kennedy,et al.
Configuring a "Wolfenstein: Enemy Territory" Server for Online Play
,
2004
.
[5]
Grenville J. Armitage,et al.
An experimental estimation of latency sensitivity in multiplayer Quake 3
,
2003,
The 11th IEEE International Conference on Networks, 2003. ICON2003..