Good Network Updates for Bad Packets Arne Ludwig, Matthias Rost, - - PowerPoint PPT Presentation

good network updates for bad packets
SMART_READER_LITE
LIVE PREVIEW

Good Network Updates for Bad Packets Arne Ludwig, Matthias Rost, - - PowerPoint PPT Presentation

Good Network Updates for Bad Packets Arne Ludwig, Matthias Rost, Damien Foucard, Stefan Schmid 1 Updates happen Network updates happen Changing security policies Network updates are challenging Even with global view Potential


slide-1
SLIDE 1

1

Good Network Updates for Bad Packets

Arne Ludwig, Matthias Rost, Damien Foucard, Stefan Schmid

slide-2
SLIDE 2

2

Updates happen

  • Network updates happen

– Changing security policies

  • Network updates are challenging

– Even with global view

  • Potential high damage if fail

– Security policy violation

slide-3
SLIDE 3

3

Example

slide-4
SLIDE 4

4

Example

slide-5
SLIDE 5

5

Example

Waypoint Enforcement (WPE)

slide-6
SLIDE 6

6

Example

  • Eventual consistency
slide-7
SLIDE 7

7

Example

✔ Eventual consistency ➢ Transient consistency?

Bad packet

slide-8
SLIDE 8

8

Example

✔ Eventual consistency ➢ Transient consistency?

slide-9
SLIDE 9

9

Example

✔ Eventual consistency ➢ Transient consistency?

slide-10
SLIDE 10

10

Example

✔ Eventual consistency ➢ Transient consistency?

slide-11
SLIDE 11

11

Example

✔ Eventual consistency ➢ Transient consistency?

slide-12
SLIDE 12

12

Example

✔ Eventual consistency ✗ Transient consistency

slide-13
SLIDE 13

13

Outline

  • What could possibly go wrong?
  • It's not a trivial thing!
  • But we present an optimal solution.
slide-14
SLIDE 14

14

Model and a Trivial Compression

Solid lines = current path

slide-15
SLIDE 15

15

Model and a Trivial Compression

Solid lines = current path Dashed lines = new path Flow-specific path

slide-16
SLIDE 16

16

Model and a Trivial Compression

Solid lines = current path Dashed lines = new path Flow-specific path

slide-17
SLIDE 17

17

Model and a Trivial Compression

Solid lines = current path Dashed lines = new path Flow-specific path Safe to be updated Safe to be left untouched

slide-18
SLIDE 18

18

Consistency Properties

  • WPE = every packet traverses the waypoint at

least once

  • LF = loop freedom
slide-19
SLIDE 19

19

Update all “simultaneously“?

slide-20
SLIDE 20

20

Update all “simultaneously“?

Not possible in practice! What could possibly go wrong?

slide-21
SLIDE 21

21

Update all “simultaneously“?

Not possible in practice! What could possibly go wrong? Update times can vary significantly (up to 10x higher than median [Dionysus – SIGCOMM'14])

slide-22
SLIDE 22

22

Update all “simultaneously“?

slide-23
SLIDE 23

23

Update all “simultaneously“?

  • Not waypoint enforced!
slide-24
SLIDE 24

24

Delay ?

slide-25
SLIDE 25

25

Delay ?

  • Not loop free!
slide-26
SLIDE 26

26

Update possible?

slide-27
SLIDE 27

27

Update possible?

slide-28
SLIDE 28

28

Update possible?

slide-29
SLIDE 29

29

Update possible?

  • Consistent transient states!
slide-30
SLIDE 30

30

Rounds

  • Round = set of parallel updates
  • ➔ Minimize number of rounds / communication
  • verhead
slide-31
SLIDE 31

31

Greedy Update Fails

  • Greedy approach may:

– take up to times more rounds – fail to find solution

See paper!

slide-32
SLIDE 32

32

Greedy Update Fails

  • Greedy approach may:

– take up to times more rounds – fail to find solution

See paper!

slide-33
SLIDE 33

33

WPE - Update Algorithm

1.Switches < WP (new), > WP (old)

slide-34
SLIDE 34

34

WPE - Update Algorithm

1.Switches < WP (new), > WP (old)

slide-35
SLIDE 35

35

WPE - Update Algorithm

1.Switches < WP (new), > WP (old)

slide-36
SLIDE 36

36

WPE - Update Algorithm

1.Switches < WP (new), > WP (old) 2.Switches < WP (new), < WP (old)

slide-37
SLIDE 37

37

WPE - Update Algorithm

1.Switches < WP (new), > WP (old) 2.Switches < WP (new), < WP (old)

slide-38
SLIDE 38

38

WPE - Update Algorithm

1.Switches < WP (new), > WP (old) 2.Switches < WP (new), < WP (old) 3.Remaining switches

slide-39
SLIDE 39

39

WPE - Update Algorithm

1.Switches < WP (new), > WP (old) 2.Switches < WP (new), < WP (old) 3.Remaining switches Constant in 3 rounds, but not LF!

slide-40
SLIDE 40

40

LF and WPE Conflict

slide-41
SLIDE 41

41

LF and WPE Conflict

  • violate WPE; violate LF
slide-42
SLIDE 42

42

Mixed Integer Program

WPE LF Minimize Rounds

slide-43
SLIDE 43

43

Mixed Integer Program

Optimal solution Unclassified (stopped 600sec) Not solvable (provably) Mixed Integer Program

slide-44
SLIDE 44

44

Solvability Analysis

Greedy MIP Unclear No solution

  • % of solvable instances?
  • % of failed greedy?
  • 1k random permutations per size
  • Max duration 600 seconds
slide-45
SLIDE 45

45

Solvability Analysis

Greedy MIP Unclear No solution

slide-46
SLIDE 46

46

Solvability Analysis

Greedy MIP Unclear No solution

slide-47
SLIDE 47

47

Solvability Analysis

Greedy MIP Unclear No solution

slide-48
SLIDE 48

48

Conclusion

  • Transient consistency is not easy to guarantee
  • LF and WPE might even conflict
  • Greedy can fail to find consistent updates

Dynamic WPE + LF updates are hard to find!