Problem statement P2P routing Data partitioning Conclusion Exploiting player behavior in distributed architectures for online games Ph.D. defense of Sergey Legtchenko INRIA/LIP6/UPMC/CNRS Supervision: S´ ebastien Monnet Pierre Sens S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 1 / 37
Problem statement P2P routing Data partitioning Conclusion Massively Multiplayer Online Games (MMOGs) Cumulated MMOG population 23 Subscriptions in millions 20 18 16 14 12 10 8 6 4 2 1998 2000 2002 2004 2006 2008 2010 2012 Date Market: $2.7 billions in 2010 S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 2 / 37
Problem statement P2P routing Data partitioning Conclusion MMOGs rely on expensive large-scale infrastructures Datacenter-based: 1000’s of server-blades 100’s of terabytes of DRAM Up to 80% of the financial revenue [Kesselman’05] S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 3 / 37
Problem statement P2P routing Data partitioning Conclusion Problem: architectures are static, workload is dynamic Server 1 Server 2 Server 3 Normal load Normal load Normal load Static game partitioning unadapted to player density evolutions S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 4 / 37
Problem statement P2P routing Data partitioning Conclusion Problem: architectures are static, workload is dynamic Server 1 Server 2 Server 3 Event Normal load Normal load Normal load Static game partitioning unadapted to player density evolutions S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 4 / 37
Problem statement P2P routing Data partitioning Conclusion Problem: architectures are static, workload is dynamic Server 1 Server 2 Server 3 Event Normal load Normal load Normal load Static game partitioning unadapted to player density evolutions S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 4 / 37
Problem statement P2P routing Data partitioning Conclusion Problem: architectures are static, workload is dynamic Server 1 Server 2 Server 3 Event Overloaded Normal load Underloaded Static game partitioning unadapted to player density evolutions S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 4 / 37
Problem statement P2P routing Data partitioning Conclusion Consequence: high cost, low efficiency Current MMOGs: Lots of empty servers [Cheslack-Postava et al, USENIX’12] Lots of overloaded servers [Varvello et al, NetGames’09] Independent game instances limited to few 100’s of players Low quality of service No geo-scale seamless virtual universe No epic battles S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 5 / 37
Problem statement P2P routing Data partitioning Conclusion Academic research on MMOGs State of the art: Well suited Extensive efforts on adaptative for: Peer-to-peer (p2p) mechanisms: [Colyseus, NSDI'06] e l s Load balancing [Donnybrook, SIGCOMM'08] a G c s O [Solipsis, PDPTA'03] M e g M Interest management [Hydra, NetGames'07] r a l [Walkad, IPTPS'09] Server based fast paced Why no impact? [Sirikata, USENIX'12] MMOGs [Najaran et al., NetGames'10] lack of robustness/performance [ALVIC-NG, NetGames'08] Hybrid [Jardine et al., NetGames'08] S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 6 / 37
Problem statement P2P routing Data partitioning Conclusion Contributions of the Thesis Guideline: improving MMOGs by making them aware of player behavior Improving robustness: BlueBanana : increasing resilience of p2p MMOGs to player movement [DSN10] Improving performance: DONUT : improving routing in large-scale p2p MMOGs with heterogeneous peer distributions [SRDS11] QuakeVolt : Efficient data management in server-based MMOGs with strong latency requirements [ongoing work] S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 7 / 37
Problem statement P2P routing Data partitioning Conclusion Outline of the talk Focusing on performance improvement: Part 1: Improving routing in p2p MMOGs with heterogeneous peer distributions with DONUT (approx 20 minutes) Part 2: efficient data management for large-scale virtual battlegrounds with QuakeVolt (approx 15 minutes) S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 8 / 37
Problem statement P2P routing Data partitioning Conclusion Part 1: Improving routing in peer-to-peer MMOGs S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 8 / 37
Problem statement P2P routing Data partitioning Conclusion Context: large-scale p2p MMOGs Nearest-neighbor overlays 2D game partition with associated overlay graph Z o n e o f A Useful properties: Peer A Data locality Greedy routing Cheap Good fault resilience Nearest-neighbors p2p overlays: [Mercury, VON, VoroNet, RayNet] S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 9 / 37
Problem statement P2P routing Data partitioning Conclusion Problem: lack of routing efficiency Efficient greedy routing: O ( log d ( N )) with Small-World shortcuts � Kleinberg , STOC ′ 2000 � Requires estimation of hop distances between peers Game C space ? F Peer B Peer A D Peer A has: Knowledge of direct neighbors E No global topology knowledge S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 10 / 37
Problem statement P2P routing Data partitioning Conclusion Estimating hop distance Uniform distribution: easy Heterogeneous distribution: hard Peer R Peer R d (R,B) euclid d (R,B) hops G d (R,B) hops through O: 2 hops d (R,B) O hops through G: d (R,B): 5 hops hops 4 hops Peer B Peer B d d d proportional to depends on density hops euclid hops S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 11 / 37
Problem statement P2P routing Data partitioning Conclusion Efficiency required despite heterogeneity Routing in MMOGs: Real distributions: non-uniform Joins 70 Density distribution of 60 Second Life islands: Player teleportation 50 "Isle of View" % of cells 40 P joins position c "Dance" bootstrap: peer B 30 20 10 B 0 0 5 10 15 20 25 30 Density (Players per cell) c S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 12 / 37
Problem statement P2P routing Data partitioning Conclusion Efficiency required despite heterogeneity Routing in MMOGs: Real distributions: non-uniform Joins 70 Density distribution of 60 Second Life islands: Player teleportation 50 "Isle of View" % of cells 40 P joins position c "Dance" bootstrap: peer B 30 20 10 B 0 0 5 10 15 20 25 30 Density (Players per cell) c S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 12 / 37
Problem statement P2P routing Data partitioning Conclusion Efficiency required despite heterogeneity Routing in MMOGs: Real distributions: non-uniform Joins 70 Density distribution of 60 Second Life islands: Player teleportation 50 "Isle of View" % of cells 40 P joins position c "Dance" bootstrap: peer B 30 20 10 B 0 0 5 10 15 20 25 30 Density (Players per cell) c S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 12 / 37
Problem statement P2P routing Data partitioning Conclusion Efficiency required despite heterogeneity Routing in MMOGs: Real distributions: non-uniform Joins 70 Density distribution of 60 Second Life islands: Player teleportation 50 "Isle of View" % of cells 40 P joins position c "Dance" bootstrap: peer B 30 20 10 B 0 0 5 10 15 20 25 30 Density (Players per cell) c S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 12 / 37
Problem statement P2P routing Data partitioning Conclusion Efficiency required despite heterogeneity Routing in MMOGs: Real distributions: non-uniform Joins 70 Density distribution of 60 Second Life islands: Player teleportation 50 "Isle of View" % of cells 40 P joins position c "Dance" bootstrap: peer B 30 20 10 B 0 0 5 10 15 20 25 30 Density (Players per cell) P c S.Legtchenko - INRIA/LIP6/UPMC/CNRS • Exploiting player behavior in distributed architectures for online games 12 / 37
Recommend
More recommend