On the Death of BGP

1
On the Death of BGP
Timothy G. Griffin Intel Research,
Cambridge UK tim.griffin_at_intel.com http//www.cam
bridge.intel-research.net/tgriffin/

• MSN
• July 8, 2004

2
Hacking our way to disaster
Provider A (Tier 1)
Provider B (Tier 1)
peering
provider/customer
Provider C (Tier 2)
primary
backup
customer
customer sends community that lowers local
preference below a providers
3
BANG!
Provider A (Tier 1)
Provider B (Tier 1)
peering
provider/customer
Provider C (Tier 2)
primary
backup
customer
customer is happy that backup was installed
4
OUCH!
Provider A (Tier 1)
Provider B (Tier 1)
peering
provider/customer
This is a stable BGP routing! Its NOT ciscos
fault!
1/2 Wedgie
Provider C (Tier 2)
primary
backup
customer
One solution --- reset BGP session on backup
link!
Better --- C should translate its customer depref
communities to those of Provider A when
re-exporting routes to A (not DT!)
5
REALLY, OUCH!
BELL NET
CIRCUIT NET
¾ Wedgie
HappyPackets (Tier 2)
NetNet (Tier 2)
primary
primary
backup
backup
LoadBalancer
P1
P2
6
What the heck is going on?

• There is no guarantee that a BGP configuration
  has a unique routing solution.
• When multiple solutions exist, the
  (unpredictable) order of updates will determine
  which one is wins.
• There is no guarantee that a BGP configuration
  has any solution!
• And checking configurations NP-Complete
• Complex policies (weights, communities setting
  preferences, and so on) increase chances of
  routing anomalies.
• yet this is the current trend!
• And you have to code your policies in an assembly
  language that was never designed, but has EVOLVED
  to allow the required HACKS

7
Deconstruction
Specification of routing message semantics
BGP
Generic Path Vector
or
MSP
SCTP
TCP
TCP
MSP Managed Session Protocol (Veiera, Griffin)
8
Routing Algebras to the Rescue!
Network Routing with Path Vector Protocols
Theory and Applications. Joao Luis Sobrinho.
SIGCOMM 2003
Signature
5
6
r
e c r p f
c c c f f f
r r r f f f
p p p p p f
c
4
2
3
Label
c
c
p
1
0
9
Really Hard Bits

• Tools and techniques for design and development
  of new Algebras (ongoing work with João Sobrinho)
• Auto-generation and checking of global
  constraints
• Generalize distrbuted cycle detection algorithms.
  
• Error reporting?
• Generalize interface between routing protocols
  and forwarding engine

10
Just Imagine .

• Inventing, implementing, and deploying new
  routing Algebras without ever going to the IETF
  or a vendor.
• Selecting a routing Algebra from a certified
  library.
• Working with another Autonomous Systems to
  develop a routing algebra that is best suited to
  solving your problems.
• Having a guarantee that the network is free of
  anomalies caused by policy interactions