Transcript Drafting Behind Akamai - Northwestern University

Drafting Behind Akamai (TravelocityBased Detouring)
Aleksandar Kuzmanovic
Northwestern University
Joint work with:
A. Su, D. Choffnes, and F. Bustamante
To appear in Sigcomm 2006
http://www.cs.northwestern.edu/~akuzma/
Drafting Behind Akamai
Drafting
Detour
2
A. Kuzmanovic
Drafting Behind Akamai
Motivation
Overlay networks
– Can’t change IP, so change the layers above
– E.g., end-point multicast
Common problem
– Build the “view” of the underlying network
– Passive and active measurements
3
A. Kuzmanovic
Drafting Behind Akamai
Problem
Independent measurements
– Redundant
– Non-scalable
– Can cause problems
• E.g., synchronization
Common service needed
– Knowledge plane - MIT
– A routing underlay for overlays - Princeton
– Network Weather Service - Berkeley
4
A. Kuzmanovic
Drafting Behind Akamai
Our Approach
Fact
– CDNs (e.g., Akamai) perform extensive network
and server measurements
• Publish the results through DNS over short time scales
Can overlay networks reuse measurements
collected by production CDNs?
– Significantly reduce the amount of measurements
(a complementary service)
– No new infrastructure need to be deployed
– Inherit the robustness of DNS
– Easy integration with existing systems
5
A. Kuzmanovic
Drafting Behind Akamai
CDN-Driven One-Hop Source Routing
D
A1
E1
A2
E2
An
En
S
DNS Server
6
A. Kuzmanovic
Drafting Behind Akamai
Key Questions
How does Akamai work?
– DNS translation
– How many web replicas does a client see?
– Impact of different sites (e.g., Yahoo vs. NY
Times)?
– DNS redirection dynamics?
– Network or server latency?
An example application: one-hop source
routing
– Mapping CDN servers to overlay nodes
– Low-overhead protocols for exploiting CDN
redirections
7
A. Kuzmanovic
Drafting Behind Akamai
DNS “Black Magic”
Root DNS Server
Akamai High-Level
DNS Server
Top-Level Domain
DNS Server
www.pcworld.com
www.pcworld.com
a1694.g.akmai.net
CNAME:
images.pcworld.com
www.pcworld.com
a1694.g.akamai.net
Akamai Low-Level
PCWorld Authoritative
DNS Server
a1694.g.akmai.net DNS Server
2 ip addresses of
Local DNS Server Akamai Edge Servers
images.pcworld.com
PCWorld
Web Server
2 ip addresses of
www.pcworld.com
Akamai Edge Servers
Akamai
Edge Server
Web
Client
8
A. Kuzmanovic
Drafting Behind Akamai
Measuring Akamai
2-months long measurement
140 PlanetLab (PL) nodes
– 50 US and Canada, 35 Europe, 18 Asia, 8 South
America, the rest randomly scattered
Every 20 sec, each PL node queries an
appropriate CNAME for
– Yahoo, CNN, Fox News, NY Times, etc.
Akamai
Edge Server 1
Akamai Low-Level
DNS Server
.……
Akamai
Edge Server 3
PL Node
A. Kuzmanovic
Akamai
Edge Server 2
Drafting Behind Akamai
9
Initial Results
Berkeley
Purdue
day
night
10
A. Kuzmanovic
Drafting Behind Akamai
Server Diversity for Yahoo
Majority of PL nodes
see between 10 and 50
Akamai edge-servers
Nodes far away
from Akamai
hot-spots
Good overlay-to-CDN
mapping candidates
A. Kuzmanovic
11
Drafting Behind Akamai
Multiple Akamai Customers
Yahoo and NYTimes are
not hosted in U. Oregon
Amazon is not hosted
on UMass and LBNL
FEMA is poor!
12
A. Kuzmanovic
Drafting Behind Akamai
Redirection Dynamics
Brazil
Berkeley
Korea
13
A. Kuzmanovic
Drafting Behind Akamai
Key Questions
How does Akamai work?
– DNS translation
– How many web replicas does a client see?
– Impact of different sites (e.g., Yahoo vs. NY
Times)?
– DNS redirection dynamics?
– Network or server latency?
Potentials for one-hop source routing?
– Mapping CDN servers to overlay nodes
– Low-overhead protocols for exploiting CDN
redirections
14
A. Kuzmanovic
Drafting Behind Akamai
Methodology
10 Best
Akamai
Edge Servers
………
Akamai Low-Level
DNS Server
Planet Lab Node
15
A. Kuzmanovic
Drafting Behind Akamai
Do CDN redirections correlate to network
latencies? (1)
Rank = r1+r2-1
– 16 means perfect correlation
– 0 means poor correlation
MIT and Amsterdam
are excellent
Brazil is poor
16
A. Kuzmanovic
Drafting Behind Akamai
Do CDN redirections correlate to network
latencies? (2)
70% paths are
among best 10%
97% better than
average
CNN is poor!
17
A. Kuzmanovic
Drafting Behind Akamai
Akamai-Driven One-Hop Source Routing
Redirections driven by
network conditions
D
Potential for
CDN-to-overlay mapping
A1
E1
A2
E2
An
S
DNS Server
Redirection dynamics
sufficiently small for
network control
En
18
A. Kuzmanovic
Drafting Behind Akamai
Methodology
Akamai path 1
Akamai path 2
Destination
Akamai path 3
Akamai path 1
Akamai path 10
Akamai path 2
Akamai path 3
Akamai Low-Level
DNS Server
Direct Path
Akamai path 10
Source
19
A. Kuzmanovic
Drafting Behind Akamai
Akamai-driven source routing (1)
Taiwan-UK
UK-Taiwan
80% Taiwan
15% Japan
5 % U.S.
75% U.K.
25% U.S.
20
A. Kuzmanovic
Drafting Behind Akamai
Akamai-driven source routing (2)
Experiment: US (6), Europe (3), S. America (2), Asia (3)
Direct paths better
than Akamai paths
Akamai and direct
paths have similar
performance
25% of Akamai
paths better than
direct paths
21
A. Kuzmanovic
Drafting Behind Akamai
Path pruning
Fact
– Not always is Akamai-driven path better than the
direct one
Practical issues
– How frequently to make a decision whether to use
the direct or the Akamai path?
– Should one use
• the first (of the 2 paths) returned by Akamai (FAS)
• the better (of the 2 paths) returned by Akamai (BTAS)
22
A. Kuzmanovic
Drafting Behind Akamai
Path Pruning Result
BTAS better than FAS
Direct path accounts for
78% of the gain
2 hours update frequency
before the performance declines
23
A. Kuzmanovic
Drafting Behind Akamai
Conclusions
Reuse measurements performed by CDNs
Reverse-engineering Akamai
– DNS redirections sufficiently small
– Strong correlation to network conditions
– All clients see a large number of paths
CDN-driven one-hop source routing
– 25% of Akamai paths outperform direct paths
– 50% of nodes “discovered” by Akamai outperform
direct paths
– Low-overhead pruning algorithms
Global Internet “weather-report” service for
little to no cost
24
A. Kuzmanovic
Drafting Behind Akamai