1. Network Game Research – Introduction and Counter- strike Analysis Mark Claypool

2. Outline ● Overview of Network Games ● Net Games Conference ● Research Issues ● Analysis Counter-strike ● Our Results

3. Why Study Games? ● Rapidly increasing in popularity – Forrester Research: 18 million on-line in 2001 – Consoles on-line ● Playstation 2 on-line (9/2002) ● Xbox Live (12/2002) – Cell phones ● Doom port (Nokia) ● Design networks to better accommodate traffic

4. Game Types ● First Person Shooters – Doom, Quake, Counter-strike, … ● Massive Multi-Player Online Role Playing – Everquest, Earth and Beyond, … ● Real-Time Strategy – Warcraft, Starcraft … ● Other – Misc – not any genre above ● Example: Diablo2, Racing, … – Non-networked ● Example: Thief – Multiplayer, but relaxed real time ● Example: Chess, Bridge

5. Net-Games - Workshop on Network and System Support for Games ● First was 2002 in Braunschweig, Germany – http://wwwmc.tm.uni-karlsruhe.de/netgames2002/ ● Second is in Redwood City, California – http://confman.eecs.umich.edu/netgames2003/ ● Sponsored by: – ACM – Electronic Arts – Microsoft

6. Net-Games Program Committee – Sugih Jamin, University of Michigan (Chair) – Mostafa Ammar, Georgia Institute of Technology – Grenville Armitage, Swinburne University of Technology – John Buchanan, Electronic Arts – Jon Crowcroft, University of Cambridge – Christophe Diot, Sprintlabs – Wu-chang Feng, Oregon Health and Science University – Carsten Griwodz, University of Oslo – Jim Kurose, UMass at Amherst – John Laird, University of Michigan – Brian Neil Levine, UMass at Amherst – Martin Mauve, University of Mannheim – Hiroyuki Morikawa, University of Tokyo – Dan Rubenstein, Columbia University – Srinivasan Seshan, Carnegie Mellon University – Wilson Yuen, City University of Hong Kong – Lars Wolf, TU Braunschweig

7. Net-Games Topics ● Multi-player game architectures and platforms ● Prevention and detection of cheating ● Games on mobile and resource-scarce devices ● AI and techniques for latency hiding ● Modeling, usage studies, and characterization ● Enabling protocols for networked games ● Systems support for authentication and accounting ● Put research issues here

8. Provisioning On-line Games: A Traffic Analysis of a Busy Counter-Strike Server Wu-chang Feng, Francis Chang, Wu-chi Feng, Jonathan Walpole

9. Goal ● Understand the resource requirements of a popular on-line FPS (first-person shooter) game

10. Why FPS? ● While there are other game types … ● Gaming traffic dominated by first-person shooter genre (FPS) [McCreary00]

11. Networked FPS lineage Doom Doom II Quake + QuakeWorld variants + Team Fortress + Capture the Flag Quake II + Soldier of Fortune + Heretic II Quake III Arena + Medal of Honor Allied Assault + Return to Castle Wolfenstein + Soldier of Fortune 2 + Jedi Knight II Doom III Half-Life + Counter-Strike + Day of Defeat + Urban Terror + Team Fortress Classic + Team Fortress 2 Unreal Unreal Tournament Unreal Tournament 2003 + America's Army: Operations 8 of top 10 games derived from one of two lineages

12. About the game... ● Half-Life modification ● Two squads of players competing in rounds lasting several minutes ● Rounds played on maps that are rotated over time ● Each server supports up to 32 players

13. What is Counter-strike (CS)?

16. About the game... ● Centralized server implementation – Clients update server with actions from players – Server maintains global information and determines game state – Server broadcasts results to each client ● Sources of network traffic – Real-time action and coordinate information – Broadcast in-game text messaging – Broadcast in-game voice messaging – Customized spray images from players – Customized sounds and entire maps from server

17. Why CS?

19. The Trace ● cs.mshmro.com (129.95.50.147) – Dedicated 1.8GHz Pentium 4 Linux server – OC-3 – 70,000+ unique players (WonIDs) over last 4 months ● One week in duration 4/11 – 4/18 ● 500 million packets ● 16,000+ sessions from 5800+ different players

20. A week in the life...

21. Variance time plot (Normalized to base interval of 10ms) Less variance < 50ms Remains to 30 min Decreases above 30 min

22. Digging deeper Interval size=10msInterval size=50ms ● Periodic server bursts every 50ms – Game must support high interactivity – Game logic requires predictable updates to perform lag compensation

23. Digging deeper ● Low utilization every 30 minutes – Server configured to change maps every 30 minutes – Traffic evenly pegged otherwise.... Interval size=1sec Interval size=30min

24. Finding the source of predictability ● Games must be fair across all mediums (i.e. 56kers) – Aggregate predictability due to “saturation of the narrowest last-mile link”

25. Packet sizes ● Supporting narrow last-mile links with a high degree of interactivity requires small packets – Clients send small single updates – Servers aggregate and broadcast larger global updates

26. Implications ● Routers, firewalls, etc. must be designed to handle large bursts at millisecond levels – Game requirements do not allow for loss or delay (lag) – Should not be provisioned assuming a large average packet size [Partridge98] – If there are buffers anywhere, they must... ● Use ECN (no, doesn’t use TCP) ● Be short (i.e. not have a bandwidth-delay product of buffering) ● Employ an AQM that works with short queues – Rate-based not Queue-based

27. Implications ● ISPs, game services – Must examine “lookup” utilization in addition to link utilization – Concentrated deployments of game servers may be problematic ● Large server farms in a single co-lo ● America's Army, UT2K3, Xbox

28. On-going work ● Other pieces in the provisioning puzzle – Aggregate player populations – Geographic distributions of players over time (IP2Geo) ● Impact on route and packet classification caching ● Other FPS games – HL-based: Day of Defeat – UT-based: Unreal Tournament 2003, America's Army – Quake-based: Medal of Honor: Allied Assault – Results apply across other FPS games and corroborated by other studies

29. Future work ● Games as passive measurement infrastructure – Only widespread application with continuous in-band ping information being delivered (measurement for free) – “Ping times” of all clients broadcast to all other clients every 2-3 seconds – 20,000+ servers, millions of clients ● Games as active measurement infrastructure – Thriving FPS mod community and tools – Server modifications [Armitage01] ● Other game types

30. Some Net-Games Research at WPI ● Mark Claypool, David LaPoint, and Josh Winslow. “Network Analysis of Counter-strike and Starcraft”, In Proceedings of the 22nd IEEE International Performance, Computing, and Communications Conference (IPCCC), Phoenix, Arizona, USA, April 2003. Online at: http://www.cs.wpi.edu/~claypool/papers/net-game/ ● Compared FPS (Counter-strike) with RTS (Starcraft) ● Smaller number of traces ● Controlled set of users

31. Our Results ● Confirms packet sizes for Counter-strike ● Starcraft results: – Packets smaller – Linear increase with number of players ● Starcraft traffic much smoother (bandwidth usage) than Counter-strike – Counter-strike bursty during ‘fire-fights’ ● (PEDS talk later)

32. Network Game Research – Introduction and Counter- strike Analaysis Mark Claypool