Transcript Chris Liljenstolpe Presentation

An Approach to IP Network
Traffic Engineering
NANOG
Miami, FL
Chris Liljenstolpe
Cable & Wireless
[email protected]
Scope and Purpose
 Describes C&W’s Traffic Engineering
methodology as well as some of the
reasoning behind it.
 Not “The One True Way,” but a method that
works for us.
Scale of the Previous Design
 Originally a flat network - one layer of routers
interconnected over a complete PVC mesh.
 A Network “event” in 1998 on AS3561 “educated”
the engineering staff on IGP scaling issues.
 This “event” lead to a week of network instability
as it was re-engineered.
 At one time there were 380+ routers in the direct
mesh, accounting for 30k PVCs in the network &
760+ direct IGP associations per router.
Hierarchy
 At one time there were 380+ routers in the direct
mesh, accounting for 30k PVCs & IGP
associations.
 Currently there are no more than 80 routers in any
one mesh due to the addition of hierarchy.
 Due to the shrinking mesh sizes, and code
optimization efforts, calculation times have dropped
from 4 hours to 20 minutes.
Online vs. Offline
 We like to always know where our traffic is
and where it is routed.
 Calculating optimal routing takes time on
dedicated compute platforms…
Layer 2 vs. Layer 3
 Utilizing IGP metrics to adjust traffic flows
on an IP network leads to network-wide (and
sometimes/usually, unplanned) effects in a
large network, due to flooding.
 This can lead to the network equivalent of
the midway game “Hit the groundhog”
IGP Use
 The IGP (in our case 2 level IS-IS) is only
used for link state signaling in normal and
most failure mode conditions.
 In the worst case dual failure mode
condition, the IGP does provide real next-hop
calculations.
IGP Metrics
 Because of the direct router-router adjacencies
provided by the underlying network, a large set of
IGP metrics are not needed.
 The set in use is small, and only used to select
primary vs. secondary path, and discourage
“expensive” link utilization in a multi-point failure
that leads to multi-hop routing.
ATM to MPLS for TE
 ATM w/ PVC’s worked quite nicely


Except for ATM overhead
And lack of high-speed router interfaces
 For our traffic engineering network, we are
treating MPLS as an IP friendly ATM
(actually more like Frame Relay, but never
mind)
Will ’s Replace MPLS?
 Only when the bandwidth required for any routerrouter pair approaches the bandwidth available
from a single  on the DWDM plant AND the cost
of a port on an OXC is significantly cheaper than an
equivalent bandwidth port on an MPLS switch.
 When that occurs, the ’s will be provisioned just
as the MPLS LSP’s are – statically with resilience.
 GMPLS may be the technology used to signal the
path over the OXC, just as MPLS is used for the
LSP’s today.
Tools
 Currently the tools that compute the paths,
and configure the layer 2 and layer 3
equipment with those paths are all developed
and maintained in-house.
 Some have been in continual development
and “tweak” mode for 6 years.
Futures
 Most link failures will be detected and handled at
the layer 2 traffic engineering layer, instead of at
layer 3.
 Path redundancy will grow from 2 to 4 paths per
router-router pair.
 Developments optimization mathematics originally
researched for circuit path layout and analog circuit
design will be utilized in the path layout tools.
 Networks other than the IP backbone will utilize the
traffic engineering core.