Internet Routing Instability Craig Labovitz G. Robert Malan Farnam Jahanian Presenters: Supranamaya Ranjan Mohammed Ahamed Appeared: SIGCOMM ‘97.

Internet Routing Instability

Craig LabovitzG. Robert MalanFarnam Jahanian

Presenters: Supranamaya Ranjan Mohammed Ahamed

Appeared: SIGCOMM ‘97

Internet Structure

• Many small ISP’s at lowest level

• Small number of big ISP’s at core

The Core of the Internet

Verio

UUNet

Sprint

rice.edu

• Routing done using BGP at core

• Inter-domain routing could be RIP/OSPF etc

BGP Overview

Verio

UUNet

Sprint 128.42.x.x196.29.x.x

128.42.x.x

100.100.x.x196.29.x.x

92.92.x.x

100.100.x.x196.29.x.x

92.92.x.x

BGP Overview (contd.)

• Similar to Distance Vector routing

• Path Vector protocol

• Loop detection done using AS_PATH field

Peering session (TCP)

R1 R2

• Exchange full routing table at start

• Updates sent incrementally

Key Point

The volume of BGP messages exchanged is abnormally high

• Most messages are redundant / unnecessary and do not correspond to and topology or policy changes

Consequence: Instability

• Normal data packets handled by dedicated hardware

• BGP packet processing consumes CPU time

• Severe CPU processing overhead takes the router offline

Route Flap Storm:

C

B

A

• Router A temporarily fails

• When A becomes alive B & C send full routing tables

• B & C fail…cascading effect

How do we avoid /lessen the impact of these problems?

Route Dampening

• Router does not accept frequent route updates to a destination

• Might signal that network has erratic connectivity

• Increment counter for destination when route changes

• Counter exceeds threshold stop accepting updates

Problem:

• Future legitimate announcements are accepted only after a delay

• Decrement counter with time

Prefix Aggregation/Super-netting

• Core router advertises a less specific network prefix

• Reduces size of routing tables exchanged

Problems:

- Internet addresses largely non-hierarchically assigned

Prefix aggregation is not effective because:

- 25% of prefixes multi-homed

- Multi-homed prefixes should be exposed at the core

- Domain renumbering not done when changing ISP’s

Route Servers

• O(N) peering sessions per Router

• 1 peering session per router

Route Server

In-spite of all these measures the BGP message overheadis unexpectedly high

Evaluation Methodology

• Data from Route Server at M.A.E west (D.C) peering point

• Peering point for more than 60 major ISP’s

• Time series analysis of message exchange events

• Nine month log

Observation: Lot’s of redundant updates

• Duplicate route with-drawls

Number of With-drawls Unique ISP

A 23276 4344

F 86417 12435

I 2479023 14112

Ratio

5

7

175

One Reason:

- Stateless BGP

- No state of previous with-drawls maintained

Observation: Instability Proportional to ActivityAfter removing duplicate messages:

Tim

e o

f d

ay

Lesser messages

Lesser messages

10:00 AM

ISP infrastructure up-grade

Instability density with time

6:00

12:00

18:00

24:00

Evidence from Fine Grained Structure

Conjecture: BGP packets are competing with data packets duringhigh bandwidth activity.

Nu

mb

er

of

insta

bilit

y e

ven

ts

Frequency (1/hour)

7 days

24 hours

Pow

er

sp

ectr

al d

en

sit

y

Observation: Instability & size uncorrelated

• ISP’s serving more network prefixesmay not contribute more to instability

AADiff

Proportion of routing table

Pro

port

ion

of

an

nou

ncem

en

ts

WADup

Proportion of routing table

Pro

port

ion

of

an

nou

ncem

en

ts

WADiff

Proportion of routing tablePro

port

ion

of

an

nou

ncem

en

ts

Observation: Instability distributed over routes

75% median

• 20% to 90% of routes change 10 times or less

• No single route contributes significantly to instability

10

# of announcements per prefix+AS

Cu

mu

lati

ve p

rop

ort

ion

Observation: Synchronized updates

• Inter-arrival times of updates shows periodicity

• 30 s and 1 minute patterns

• Some routers collect and send Updates once every 30 s

• Routers get synchronized

Possible reasons:

• Border router- Internal router: interaction misconfigured??

AADiff

Inter Arrival Time distribution for AADiff’s

Pro

port

ion

30s

1min

End-to-end Perspective

Chinoy: “Dynamics of Internet routing information” (SIGCOMM 93)

Measurements on NSFNET showed: - Processing and forwarding latency of BDP update is 3 orders of magnitude more than the latency incurred in forwarding data packets - Will lead to packet drops during the intervening period

Paxson: “End-to-End routing behavior in the internet” (SIGCOMM 96)

- Routing loops introduce loops into other router’s routing tables

- An end-to-end route changes every 1.5 hours on an average

End-to-End perspective (Paxson)

Pathology type

Probability in 1995

Probability in 1996

Long-livedRouting loops

Short-livedRouting loops

Outage>30s

Total

0.065%~

0.14%~same

same

0.96% 2.2%

3.4%1.5%

Summary and Conclusions

• Redundant routing information flows in core

• Instability distributed across autonomous systems

Possible reasons for instability:

- Stateless BGP updates

- Misconfigured routers

- Synchronization

- Clocks driving the links not synchronized (link “flaps”)

Follow-up work & impact

• Migration from stateless to stateful BGP decreased duplicate withdrawals by an order of magnitude

“Origins of Internet Routing Instability”-1999

• But Duplicate Announcements (AADup) doubled

• Reason: Non-transitive attribute filtering not implemented

- BGP specification: “never propagate non-transitive attributes”..

- ASPATH is transitive attribute

- MED (Multi Exit Discriminator) is NOT transitive

Propagating MED’s Causes Oscillations