game traffic analysis: an mmorpg perspective
DESCRIPTION
Online gaming is one of the most profitable businesses on the Internet. Of all the genres of online games, MMORPGs (Massive Multiplayer Online Role Playing Games) have become the most popular among network gamers, and now attract millions of users who play in an evolving virtual world simultaneously over the Internet. To gain a better understanding of game traffic and contribute to the economic well-being of the Internet, we analyze a 1, 356-million-packet trace from a sizeable MMORPG called ShenZhou Online. This work is, as far as we know, the first formal analysis of MMORPG server traces. We find that MMORPG and FPS (First-Person Shooting) games are similar in that they both generate small packets and require low bandwidths. In practice, the bandwidth requirement of MMORPGs is the lower of the two due to less real-time game playing. More distinctive features are the strong periodicity, temporal locality, irregularity, and self-similarity observed in MMORPG traffic. The periodicity is due to a common practice in game implementation, where game state updates are accumulated within a fixed time window before transmission. The temporal locality in game traffic is largely due to the game’s nature, whereby one action leads to another. The irregularity, which is unique to MMORPG traffic, is due to the diversity of the game’s design so that the behavior of users can vary drastically, depending on the quest at hand. The self-similarity of the aggregate traffic is due to the heavy-tailed active/idle activities of individual players. Moreover, we show that the arrival of game sessions within one hour can be modelled by a Poisson model, while the duration of game sessions is heavy-tailed.TRANSCRIPT
Game Traffic Analysis:Game Traffic Analysis:An MMORPG PerspectiveAn MMORPG Perspective
(MMORPG: Massive Multiplayer Online Role Playing Game)
Apr. 29, 2005
Kuan-Ta Chen, Polly Huang,
Chun-Ying Huang, Chin-Laung Lei
Department of Electrical EngineeringNational Taiwan University
Appeared in ACM NOSSDAV 2005 (15th International Workshop on Network and Operating System Support for Digital Audio and Video)
2Game Traffic Analysis: An MMORPG Perspective
Talk OutlineTalk Outline
Overviewnetwork game research
current status of MMORPG
Trace collection
Traffic characterization & their physical explanations
Conclusion
Future work
3Game Traffic Analysis: An MMORPG Perspective
Network Game ResearchNetwork Game Research
Traffic analysis and modelingpacket size/arrival process
connection time/arrival process
User behavior analysis and modelingsession time, session arrival process
session membership
The effects of network QoSaffect player duration?
affect arrival / departure rate?
affect overall score or running speed?
4Game Traffic Analysis: An MMORPG Perspective
Network Game Research (cont.)Network Game Research (cont.)
Dead reckoning (client-side prediction)
Interest management (preference clustering)
Server discovery mechanism(select a best game server)
Network infrastructuremulticast – group management
multi-tier (proxyed)
p2p / overlay networks
5Game Traffic Analysis: An MMORPG Perspective
MotivationMotivation
Network games contribute 3% -- 4% Internet traffic at backbones in 2000
MMORPGthe dominate network game genre in Asia
exhibit distinct features, such as• no explicit game rounds• no restrict on player number
6Game Traffic Analysis: An MMORPG Perspective
The The Increasing PopularityIncreasing Popularity of MMOGof MMOG
Figure curtsey of http://www.mmogchart.com/
7Game Traffic Analysis: An MMORPG Perspective
Popular GenresPopular Genres inin MMORPGMMORPG
Figure curtsey of http://www.mmogchart.com/
8Game Traffic Analysis: An MMORPG Perspective
Network Games in TaiwanNetwork Games in Taiwan
120 network games in 1st quarter, 2003
market value of 6,800 million NT dollars in 2003 fl 80% share of the whole PC game industry
Gamania (Lineage) owns > 4,000 Mbps links
R. O. has a record of 370,000 players online, equivalent to 1.5% of the population
9Game Traffic Analysis: An MMORPG Perspective
OnlineOnline
a mid-scale, commercial MMORPG in Taiwan
a typical Asia MMORPG ---Microsoft Windows platform (Windows 2000,MS SQL Server)
data communication based on TCP
(By the way) Issues on network protocolsUDP: Ultima Online (EA), Everquest series (SOE)
TCP: World of Warcraft (Blizzard)
no general agreement on the use of base protocol
神州神州 OnlineOnline
It’s me
11Game Traffic Analysis: An MMORPG Perspective
Trace Collection Trace Collection -- SetupSetup
a FreeBSD PCw/ tcpdump
Traffic Monitor
L3 switch
L2 switchL4 switch
Game & Database servers
Monitoringinterface
Managementinterface
Gig
abit
Eth
erne
t
Port forwarding
Game Traffic
12Game Traffic Analysis: An MMORPG Perspective
Trace Collection Trace Collection -- SummarySummary
Trace Sets Date Time
8/29/04 (Sun.) 15:00N2 2 8/30/04 (Mon.) 13:00 12 hr. ?
Drops
3 0.003%
Period
N1 8 hr.
Trace TCPConn.
Packets (in/out/both)
Bytes (in/out/both)
342M / 353M / 695M
4.7TB / 27.3TB / 32.0TB
N2 54,424 325M / 336M / 661M
4.7TB / 21.7TB / 26.5TB
57,945N1
1,356 million packets in total
13Game Traffic Analysis: An MMORPG Perspective
Term definitionTerm definition
The same rules apply to server traffic, server packets, and server data packets
Client traffic all traffic sent from clients
Client packets all packets sent from clientsClient data packets
all packets excluding “pure TCP ack packets” (40 bytes) sent from clients
14Game Traffic Analysis: An MMORPG Perspective
TrafficTraffic CharacterizationCharacterization
Tiny packets
Low bandwidth requirement
Temporal dependency in bothwithin-connection traffic
aggregate traffic
Periodicity and synchronization
15Game Traffic Analysis: An MMORPG Perspective
Payload size distribution Payload size distribution -- CDFCDF
client packets98% pkts’ payload ≤ 31 bytesthe most two modes occupy 90% fl certaincommands are popular
server packetsavg. payload size 114 bytes
contrast to the mean packet size 400 bytes observed in backbones
16Game Traffic Analysis: An MMORPG Perspective
The consequence of tiny packetsThe consequence of tiny packets
overhead of protocol in client traffic73% bytes are headers
30% bytes (38% pkts) are for pure TCP acks
The challenge on infrastructuresoverall avg. pkt size is 84 bytes
routers generally assume pkt size within 125 -- 250 bytes
when vendors claim 1 Gbps rate, they may indicate 5M pkt/sec for avg. pkt size = 200 bytes
17Game Traffic Analysis: An MMORPG Perspective
Packet load distributionPacket load distribution
< 5 pkt/sec for server data pkt in most of conn.
< 15 pkt/sec for 99% of conn.
18Game Traffic Analysis: An MMORPG Perspective
Bandwidth UsageBandwidth Usage
client traffic< 3 Kbps data
< 8 Kbps w/ ack
server traffic< 13 Kbps
<< 40 Kbps for a FPS (First Person Shooting) game, Counter-Strike
Similar to Warcraft III, aRTS (Realtime Strategy) Game
19Game Traffic Analysis: An MMORPG Perspective
Cannot Overlook the Traffic Cannot Overlook the Traffic ofof MMOGsMMOGs
R. O. in Taiwan announced a record of 370,000 online players
assuming each user use 10 Kbps fl total3.7 Gbps is required in average for just Agame.
20Game Traffic Analysis: An MMORPG Perspective
Packet Interarrivals within a ConnectionPacket Interarrivals within a Connection
User inputs is not Poisson
Temporal dependence exist in both client traffic and server traffic
21Game Traffic Analysis: An MMORPG Perspective
CDF of Packet Interarrival TimesCDF of Packet Interarrival Times
deviation from exponential distribution
22Game Traffic Analysis: An MMORPG Perspective
Diversity in User BehaviorsDiversity in User Behaviors
active player
common player
idle player
active player: close to exponential
idle player: uniform + determinist distr.
common player: in-between the two extremes
神州神州 OnlineOnline
These players are keeping stalls.
24Game Traffic Analysis: An MMORPG Perspective
CDF of Packet Interarrival TimesCDF of Packet Interarrival Times
50% are around 200 msfl signature of periodic message dispatch
25Game Traffic Analysis: An MMORPG Perspective
Temporal Locality in Client TrafficTemporal Locality in Client Traffic
ACF: auto-correlation functions
clustering nature in player actions
active actions: walking, fighting
inactive actions: talking, viewing equipments, trading, idle
26Game Traffic Analysis: An MMORPG Perspective
100 s 120 s 120 s 150 s30 s
400 s140 s
Clustered nature in user actionsClustered nature in user actions
Client packet inter-arrival times from a player
27Game Traffic Analysis: An MMORPG Perspective
Temporal Locality in Server TrafficTemporal Locality in Server Traffic
Server pkt primarily convey position updates
the rate of position updates depends on the number of nearby characters
28Game Traffic Analysis: An MMORPG Perspective
Spatial Locality in # of NeighborsSpatial Locality in # of Neighbors
Pos Neighbor #
(3,3) 1
(2,3) 2
…(1,1) 4
(1,2) 5
(1,3) 4
29Game Traffic Analysis: An MMORPG Perspective
Transformation: SpatialTransformation: Spatial Locality Locality toto Temporal LocalityTemporal Locality
t Neighbor #
1 1
2 2
3 1
…10 4
11 5
12 4
The spatial locality shows up in terms of temporal locality in the traffic as the characters move continuously on the map.
30Game Traffic Analysis: An MMORPG Perspective
Temporal locality in server trafficTemporal locality in server traffic
Server packet rates for a player
31Game Traffic Analysis: An MMORPG Perspective
Aggregate Packet ArrivalsAggregate Packet Arrivals
the aggregate packet arrival time seriescount the number of incoming/outgoing packets regardless of connections
take samples every 10 ms
Patterns identifiedtemporal dependence
periodicity
32Game Traffic Analysis: An MMORPG Perspective
ACF of aggregate server pkt arrivalsACF of aggregate server pkt arrivals
periodity: 200 ms
position updates are synchronous to all clients
incurring packet bursts
33Game Traffic Analysis: An MMORPG Perspective
ACF of aggregate client pkt arrivalsACF of aggregate client pkt arrivals
Positive temporal dependence up tothree minutes.
34Game Traffic Analysis: An MMORPG Perspective
The The ““flash crowdsflash crowds”” effectseffects
ACF for successive 3 minutes
Players are active simultaneously at times
“global events”lead to flash crowds effects
35Game Traffic Analysis: An MMORPG Perspective
Frequency Components in Server TrafficFrequency Components in Server Traffic
5 HzPSD: power spectral density
another proof of the 200 ms round
Servers seem to adapt frequency by certain metrics, such as number of nearby characters
Multiples of 5 Hz
36Game Traffic Analysis: An MMORPG Perspective
Frequency Components in Client TrafficFrequency Components in Client Traffic
by auto-walk and auto-attack timers
Adjust frequency by level/skill and weapons held
client timers are synchronized
6 Hz
Multiples of 6 Hz
Does the sync. lead to performance problems?
37Game Traffic Analysis: An MMORPG Perspective
ConclusionConclusion
Traffic characteristics & physical explanationsTiny packetsLow bandwidth requirement for individual clientNon-poisson user inputs (diversity of user behaviors)Temporal dependency in within-connection traffic (clustering nature in user inputs)Temporal dependency in aggregate traffic (flash crowds effects)Periodicity and synchronization(common implementation practice)
38Game Traffic Analysis: An MMORPG Perspective
Conclusion (cont.)Conclusion (cont.)
TCP maybe an overkill for MMORPG, e.g., in oveall client traffic
73% bytes used by TCP/IP headers
30% bytes used by pure TCP acks
39Game Traffic Analysis: An MMORPG Perspective
Future WorkFuture Work
assessment for impacts of network QoS
more traffic characterizationIdentify abnormal sessions, e.g., bots in Lineage & Ragnarok Online
Identify game traffic
performance evaluation
To provide a better network To provide a better network infrastructure for netgamesinfrastructure for netgames
Questions?Questions?
Thank You!Thank You!
[email protected]://kuan.ilife.cx/