+ All Categories
Home > Documents > Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA...

Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA...

Date post: 15-Dec-2015
Category:
Upload: shawn-goldsby
View: 223 times
Download: 0 times
Share this document with a friend
51
NETWORK SUPPORT FOR IP TRACEBACK- SIGCOMM ‘00 Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA resented by Mohammad Hajjat- Purdue Universi Slides courtesy of Teng Fei - Umass April, 2002 1
Transcript
Page 1: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

1

NETWORK SUPPORT FOR IP TRACEBACK-SIGCOMM ‘00

Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson

University of Washington- Seattle, WA

Presented by Mohammad Hajjat- Purdue University

Slides courtesy of Teng Fei - Umass April, 2002

Page 2: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

2

The Problem

Denial of Service (DoS) attack Remotely consume resource of server or network Increase in number and frequency Simple to implement

DoS attacks are difficult to trace: Indirection

Attacking packets sent from slave machines, which under the control of a remote master machine

Spoof of IP source addresses Disguise their location using incorrect IP addresses,

hence the true origin is lost

Page 3: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

3

Packet Marking Traceback

Mark packets with router address deterministically or probabilistically

Trace attack using marked packets Pros

Require no cooperation with ISPs Does not cause heavy network overhead Can trace attack “post mortem”

Page 4: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

4

Multiple Attackers

A1 A2 A3

R5

R3

R6 R7

R4

R2

R1

attackorigin

victim

V

Page 5: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

5

Exact Traceback Problem

A1 A2 A3

R5

R3

R6 R7

R4

R2

R1 V

attackpath

exact tracebackR6, R3, R2, R1

Page 6: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

6

Approximate Traceback Problem

A1 A2 A3

R5

R3

R6 R7

R4

R2

R1 V

approx. tracebackR5, R6, R3, R2, R1

Page 7: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

7

Methodology

I. Marking procedure by routers add information to packets

II. Path reconstruction procedure by victim use information in marked packets convergence time: # of packets to

reconstruct the attack path

Page 8: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

8

Basic Marking Algorithms

I. Node Append II. Node Sampling III. Edge Sampling

Page 9: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

9

I. Node Append

Append address of each node to the end of the packet

Complete, ordered list of routers attack path

original packet router list

Page 10: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

10

I. Node Append

Pros complete, ordered attack path converge quickly (single packet)

Cons infeasibly high router overhead attacks can create false path information

Page 11: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

11

II. Node Sampling

Reserve node file in packet header Router write address in node field with

probability p Reconstruct path using relative # of node

samples Only require additional write, checksum

update

Page 12: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

12

II. Node Sampling

R1

R1

R2 R3

Page 13: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

13

II. Node Sampling

R1

R1

R2 R3

Page 14: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

14

II. Node Sampling

R1

R1

R2 R3

Page 15: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

15

II. Node Sampling

R1

R3

R2 R3

Page 16: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

16

II. Node Sampling

Cons: Slow convergence

need many packets usually order of 10,000 - 100,000

Can not trace multiple attackers ▪

Page 17: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

17

III. Edge Sampling

Edge represent routers at each end of the link

Store edges instead of nodes start and end addresses of edge routers distance from edge to victim

R1 R2

Page 18: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

18

III. Edge Sampling

A router writes its own address in the start field, and 0 into the distance field

Distance field of 0 means the packet is already marked router writes its own address in the end

address field and increase the distance field by 1

Other routers may then reset these fields. Otherwise, the distance field is incremented

Page 19: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

19

III. Edge Sampling

R1 R2 R3

R1 #1 #1

Page 20: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

20

III. Edge Sampling

R1 R2 R3

R1 #1 0

Page 21: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

21

III. Edge Sampling

R1 R2 R3

R1 R2 1

Page 22: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

22

III. Edge Sampling

R1 R2 R3

R1 R2 2

Page 23: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

23

Path Reconstuction

Consider G is a graph with root v Insert tuples (start, end, distance) into G Remove any edge (x, y, d) with d !=

distance from x to v in G Extract path from G

Page 24: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

24

III. Edge Sampling

Pros Converge much faster than node sampling Efficiently discern multiple attacks

Cons Space: requires additional space in the IP

header- 72 bits of space in every IP packet (2 x 32 bit IP address and 8 bit for distance)

Compatibility ▪

Page 25: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

25

Encoding Issue

Overload the IP identification field used for fragmentation

Decreases the space requirement store the XOR of the edge addresses (edge-

id)- B XOR A XOR B = A

Pros: Reduced space

Cons: Increases reconstruction time

Page 26: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

26

Marking With XOR

a b c d v

attack path

resulting XOR edges

a XOR b b XOR c c XOR d d

Page 27: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

Reconstructing With XOR

a XOR b

b XOR c

c XOR d d

creconstructed path

b

a27

Page 28: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

28

Subdividing Edge-id

Reduce per packet space more by dividing the edge-id (XORed address) into k non-overlapping packets, and store only 1 of them

Need offset of fragment

Page 29: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

29

Creating Unique Edge-ids

Problem: Edge-id fragments are not unique with multiple attackers, multiple edge

fragments with the same offset and distance Solutoin: Bit-interleave hash code with IP

address

Page 30: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

30

Creating Unique Edge-ids

0000...1111Address Hash(Address)

0011…1100

00000101...11111010

Bit-interleave

send k fragments into network

0 k-1

Page 31: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

31

Candidate Edge-ids

Combine all permutations of fragments at each distance with disjoint offset values

Check that the hash matches hash of the address

Page 32: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

32

Construction Candidate Edges

0000...1111Address? Hash(Address)?

0011…1100

00000101...11111010

0

k-1

Hash(Address?)0011…1100

=?

No, reject

Yes, correct address

Page 33: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

33

Encoding Edge Fragments

Overload the 16-bit identification field used to differentiate IP fragments

Page 34: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

34

Testing the Algorithm

Simulator Create random paths Originate attacks

Marking probability is 1/25 1,000 random test runs vary path lengths

Page 35: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

35

Experimental Results

number of packets to reconstruct paths

Page 36: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

36

Thanks for listening

Questions?

Page 37: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

37

Backup slidesFuture Work

Suffix validation spoof end edges include a router “secret”

Attack origin (host) Find attacker (person)

Page 38: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

38

Related Research

Steven M. Bellovin ICMP Traceback Message AT&Thttp://www.research.att.com/~smb/papers/draft-bellovin-itrace-00.txt

Alex Snoeren Hash-Based IP Traceback BBN SigCOMMhttp://www.acm.org/sigcomm/sigcomm2001/p1-snoeren.pdf

Page 39: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

39

References

Stefan Savage Practical Network Support For IP Traceback http://www.cs.washington.edu/homes/savage/papers/UW-CSE-00-02-01.pdf

Sara Sprenkle Practical Network Support

Duke Universityhttp://www.duke.edu/~ses12/presentations/nerdSavage.ppt

Hal Burch IP Traceback Carnegie Mellon Universityhttp://axp.missouri.edu/~cecs481/Talks/rrp83a.ppt

Page 40: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

40

DoS Counter Measures

Ingress filtering Link testing

input debugging controlled flooding

Logging

Page 41: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

41

Ingress Filtering

Block packets with invalid source addresses

Pros Moderate management/network overhead

Cons require widespread deployment hard to do in backbone/transit network

Page 42: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

42

Link Testing

Start from victim and test upstream links Recursively repeat until source is located Assume attack remains active until trace

complete

Page 43: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

43

Input Debugging

Victim recognize attack signature Install filter on upstream router Pros

May use software to help coordinate Cons

Require cooperation between ISPs Considerable management overhead

Page 44: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

44

Controlled Flooding

Flooding link with large bursts of traffic during attack

Observe attacking packet rate change to determine the source

Pros Ingenious

Cons Itself a denial of service - possible worse

Page 45: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

45

Logging

Key routers logging packets Data mining to analysis Pros

Post mortem Cons

High resource demand

Page 46: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

46

ICMP Traceback

Sample packets with low probability Copy data and path information in a

new ICMP packet Pros

reconstruct path information with large amount of packet

Cons ICMP may be filtered

Page 47: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

47

DoS Attack Assumptions

Attacker may generate any packet Multiple attackers may conspire Attackers may be aware they are being

traced packets may be lost or reordered

Page 48: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

48

Design Assumptions

Attackers send numerous packets Route between attacker and victim is

fairly stable Routers have limited CPU and memory Routers are not widely compromised

Page 49: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

49

IP Header Encoding

Backwards compatibility Two problems

Writing same values into id fields of frags from different datagrams

Writing different values into id fields of frags of same datagrams

Page 50: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

50

Fragmentation Issues

Copy data into ICMP packet Check the checksum at higher level etc

Page 51: Stefan Savage, David Wetherall, Anna Karlin and Tom Anderson University of Washington- Seattle, WA Presented by Mohammad Hajjat- Purdue University Slides.

51

Evaluation

Longer convergence time divide edge-id into 8 fragments attacker’s distance is 10 hops 2150 packets to converge with 95% certanty

few seconds Robust with multiple attackers


Recommended