Best Practices for Determining the Traffic Matrix in IP...
Transcript of Best Practices for Determining the Traffic Matrix in IP...
![Page 1: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/1.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 1
Best Practices for Determining the TrafficMatrix in IP Networks
Apricot 2005 - Kyoto, Japan
Tutorial, Monday February 21, 200516:00-17:30
Thomas Telkamp, Cariden Technologies, Inc.Stefan Schnitter, T-Systems
(c) cariden technologies, inc. portions (c) t-systems, adlex inc., cisco systems, juniper networks.
======!"§==Systems=
![Page 2: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/2.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 2
Presenters and Contributors
• Presenters:– Thomas Telkamp, Cariden Technologies, Inc.– Stefan Schnitter, T-Systems
• Contributors:– Benoit Claise, Cisco Systems, Inc.
– Cisco NetFlow
– Tarun Dewan, Juniper Networks, Inc.– Juniper DCU
– Mark Pommrehn, Adlex, Inc.– Adlex NetFlow collector deployment
– Mikael Johansson, KTH– Traffic Matrix Estimation
![Page 3: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/3.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 3
Agenda
• Introduction– Traffic Matrix Properties
• Measurement in IPnetworks– NetFlow– NetFlow Deployment
Case-Study– DCU (Juniper)– BGP Policy Accounting
• MPLS Networks– RSVP based TE– LDP
• Data Collection• LDP deployment in
Deutsche Telekom
• Traffic Matrices in PartialTopologies
• Estimation Techniques– Theory– Example Data– Case-Study
• Summary
![Page 4: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/4.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 4
Traffic Matrix
• Traffic matrix: the amount of data transmittedbetween every pair of network nodes– Demands– “end-to-end” in the core network
• Traffic Matrix can represent peak traffic, ortraffic at a specific time
• Router-level or PoP-level matrices
234 kbit/s
![Page 5: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/5.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 5
Determining the Traffic Matrix
• Why do we need a Traffic Matrix?– Capacity Planning
• Determine free/available capacity• Can also include QoS/CoS
– Resilience Analysis• Simulate the network under failure conditions
– Network Optimization• Topology
– Find bottlenecks
• Routing– IGP (e.g. OSPF/IS-IS) or MPLS
![Page 6: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/6.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 6
Types of Traffic Matrices
• Internal Traffic Matrix– PoP to PoP matrix
• Can be from core (CR) or access (AR) routers
– Class based
• External Traffic Matrix– PoP to External AS
• BGP• Origin-AS or Peer-AS
– Peer-AS sufficient for Capacity Planning and ResilienceAnalysis
• Useful for analyzing the impact of external failures onthe core network (capacity/resilience)
![Page 7: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/7.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 7
Internal Traffic Matrix
CR
CR
CR
CR
PoP
AR
AR
AR
AR
AR
PoP
AR
Customers
Customers
AS1 AS2 AS3 AS4 AS5
Server Farm 1 Server Farm 2
“PoP to PoP”, the PoP being the AR or CR
B. Claise, Cisco
![Page 8: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/8.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 8
External Traffic Matrix
CR
CR
CR
CR
PoP
AR
AR
AR
AR
AR
PoP
AR
Customers
Customers
AS1 AS2 AS3 AS4 AS5
Server Farm 1 Server Farm 2
From “PoP to BGP AS”, the PoP being the AR or CR
The external traffic matrix can influence the internal one
B. Claise, Cisco
![Page 9: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/9.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 9
Traffic Matrix Properties
• Example Data from Tier-1 IP Backbone– Measured Traffic Matrix (MPLS TE based)– European and American subnetworks– 24h data– See [1]
• Properties– Temporal Distribution
• How does the traffic vary over time
– Spatial Distribution• How is traffic distributed in the network?
– Relative Traffic Distribution• “Fanout”
![Page 10: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/10.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 10
Total traffic and busy periods
Total traffic very stable over 3-hour busy period
European subnetwork American subnetwork
![Page 11: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/11.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 11
Spatial demand distributions
Few large nodes contribute to total traffic (20% demands – 80% of totaltraffic)
European subnetwork American subnetwork
![Page 12: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/12.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 12
Fanout factors
Demands for 4 largest nodes, USA Corresponding fanout factors
Fanout factors much more stable than demands themselves!
Fanout: relative amount of traffic (as percentage of total)
![Page 13: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/13.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 13
Traffic Matrix Collection
• Data is collected at fixed intervals– E.g. every 5 or 15 minutes
• Measurement of Byte Counters– Need to convert to rates– Based on measurement interval
• Create Traffic Matrix– Peak Hour Matrix
• 5 or 15 min. average at the peak hour
– Peak Matrix• Calculate the peak for every demand• Real peak or 95-percentile
![Page 14: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/14.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 14
Collection Methods
• NetFlow– Routers collect “flow” information– Export of raw or aggregated data
• DCU– Routers collect aggregated destination statistics
• MPLS– LDP
• Measurement of LDP counters
– RSVP• Measurement of Tunnel/LSP counters
• Estimation– Estimate Traffic Matrix based on Link Utilizations
![Page 15: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/15.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 15
NetFlow based Methods
![Page 16: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/16.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 16
NetFlow
• A “Flow” is defined by– Source address– Destination address– Source port– Destination port– Layer 3 Protocol Type– TOS byte– Input Logical Interface (ifIndex)
• Router keeps track of Flows and usage perflow– Packet count– Byte count
![Page 17: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/17.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 17
NetFlow Versions
• !Version 5– the most complete version
• Version 7– on the switches
• Version 8– the Router Based Aggregation
• Version 9– the new flexible and extensible version
• Supported by multiple vendors– Cisco– Juniper– others
![Page 18: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/18.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 18
NetFlow Export
• A Flow is exported when– Flow expires– Cache full– Timer expired
• Expired Flows are grouped together into“NetFlow Export” UDP datagrams for export toa collector– Including timestamps
• UDP is used for speed and simplicity• Exported data can include extra information
– E.g. Source/Destination AS
![Page 19: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/19.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 19
NetFlow ExportB. Claise, Cisco
![Page 20: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/20.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 20
NetFlow Deployment
• How to build a Traffic Matrix from NetFlowdata?– Enable NetFlow on all interfaces that source/sink
traffic into the (sub)network• E.g. Access to Core Router links (AR->CR)
– Export data to central collector(s)– Calculate Traffic Matrix from Source/Destination
information• Static (e.g. list of address space)• BGP AS based
– Easy for peering traffic– Could use “live” BGP feed on the collector
• Inject IGP routes into BGP with community tag
![Page 21: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/21.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial
BGP Passive Peer on the Collector
• Instead of exporting the peer-as ordestination-as for the source and destinationIP addresses for the external traffic matrix:– Don’t export any BGP AS’s– Export version 5 with IP addresses or version 8 with an
prefix aggregation
• A BGP passive peer on the NetFlow collectormachines can return all the BGP attributes:– source/destination AS, second AS, AS Path, BGP
communities, BGP next hop, etc…
• Advantages:– Better router performance – less lookups– Consume less memory on the router– Full BGP attributes flexibility
![Page 22: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/22.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 22
NetFlow: Asymetric BGP traffic
• Origin-as– Source AS1, Destination AS4
• Peer-as– Source AS5, Destination AS4 WRONG!
• Because of the source IP address lookup in BGP
B. Claise, Cisco
![Page 23: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/23.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 23
NetFlow Version 8
• Router Based Aggregation• Enables router to summarize NetFlow Data• Reduces NetFlow export data volume
– Decreases NetFlow export bandwidth requirements– Makes collection easier
• Still needs the main (version 5) cache• When a flow expires, it is added to the
aggregation cache– Several aggregations can be enabled at the same
time
• Aggregations:– Protocol/port, AS, Source/Destination Prefix, etc.
![Page 24: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/24.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 24
NetFlow: Version 8 ExportB. Claise, Cisco
![Page 25: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/25.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 25
BGP NextHop TOS Aggregation
• New Aggregation scheme– Only for BGP routes
• Non-BGP routes will have next-hop 0.0.0.0
• Configure on Ingress Interface• Requires the new Version 9 export format• Only for IP packets
– IP to IP, or IP to MPLS
![Page 26: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/26.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 26
BGP NextHop TOS Aggregation
WR
CR
CR
CR
CR
PoP
AR
AR
AR
AR
AR
PoP
AR
Customers
Customers
AS1 AS2 AS3 AS4 AS5
Server Farm 1 Server Farm 2
MPLS coreor IP core with only BGP routes
B. Claise, Cisco
![Page 27: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/27.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 27
MPLS aware NetFlow
• Provides flow statistics per MPLS and IPpackets– MPLS packets:
• Labels information• And the V5 fields of the underlying IP packet
– IP packets:• Regular IP NetFlow records
• Based on the NetFlow version 9 exportNo more aggregations on the router (version8)
• Configure on ingress interface• Supported on sampled/non sampled NetFlow
![Page 28: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/28.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 28
WR
CR
CR
CR
CR
PoP
AR
AR
AR
AR
AR
PoP
AR
Customers
Customers
AS1 AS2 AS3 AS4 AS5
Server Farm 1 Server Farm 2
MPLS
MPLS aware NetFlow: ExampleB. Claise, Cisco
![Page 29: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/29.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 29
NetFlow Summary
• Building a Traffic Matrix from NetFlow data isnot trivial– Need to correlate Source/Destination information
with routers or PoPs
• “origin-as” vs “peer-as”– Asymetric BGP traffic problem
• BGP NextHop aggregation comes close todirectly measuring the Traffic Matrix– NextHops can be easily linked to a Router/PoP– BGP only
• NetFlow processing is CPU intensive on routers– Use Sampling
• E.g. only use every 1 out of 100 packets• Accuracy of sampled data
![Page 30: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/30.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 30
NetFlow Summary
• Various other features are available• Ask vendors (Cisco, Juniper, etc.) for details
on version support and platforms• For Cisco, see Benoit Claise’s webpage:
– http://www.employees.org/~bclaise/
![Page 31: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/31.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 31
NetFlow Case-Study
![Page 32: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/32.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 32
Deployment Scenario
• NetFlow deployment in a large ISP network(“ISP X”) using Adlex FlowTracker– Traffic Engineering Analysis (TEA)
• Goal is to obtain an accurate Traffic Matrix– Router to Router matrix
• Internal Traffic sources/sinks– typically blocks of customer address space in PoPs,
such as such as broadband access devices (DSL orCable Modem termination systems, dedicatedcorporate Internet access routers, dial NASes, etc).
• External traffic sources/sinks– typically public or private peering links (eBGP
connections) in peering centers or transit PoPs
![Page 33: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/33.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 33
Associating Traffic with Routers
• Customer routes in each PoP are advertisedinto iBGP from the IGP– by each of the two backbone routers in each PoP– with the backone router’s loopback address as the
BGP Next Hop IP address for each of the local routesin the PoP
• The Adlex TEA system can pick them up fromthe BGP table via an integrated Zebra softwarerouter component in the Adlex Flow Collector(AFC)
• ISP uses Version 5 Netflow with Adlex FlowCollectors that are BGP Next-Hop aware at thelocal (PoP) and external (Internet) CIDR level
![Page 34: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/34.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 34
ISP X Phase 1: Internet Traffic
• Enable NetFlow on all interfaces on eBGPpeering routers– Flows are captured at the Internet border, from the
peering routers, as they pass through peeringrouters to/from Internet eBGP peers
• Adlex Traffic Engineering Report Server:– Retreives and aggregates summarized flows from
multiple AFCs– Exports daily traffic matrix CSV files to Cariden MATE
• For Modeling, Simulation and Control
• Hourly router-router values actually containthe the highest 15-minute average bandwidthperiod within that whole hour (4 periods/hour)– provides sufficient granularity to get near daily peak
values between routers or PoPs
![Page 35: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/35.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 35
ISP X Phase 1: Internet Traffic
Modeling,Simulation,Control
![Page 36: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/36.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 36
ISP X Phase 2: PoP-to-PoP Traffic
• Ingress-only Netflow exported from PoP-facinginterfaces on Backbone routers– Enables capturing data flowing between POPs
• Flow assignment accuracy is optimized if eachrouter that exports flows has those flowsanalyzed according to its own BGP table– Thus the traffic collection and analysis system must
process a BGP table per-router
• BGP table per backbone router and perpeering router
![Page 37: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/37.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 37
ISP X Phase 2: PoP-to-PoP Traffic
Modeling,Simulation,Control
![Page 38: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/38.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 38
Example Results
Abbreviated header showing hourly columns:
# Time Router A IP,Time Router BIP,09/01/04 12:00:00 AM Max Bits/s Router A->B(bps),09/01/04 01:00:00 AM Max Bits/sRouter A->B(bps), 09/01/04 02:00:00 AM MaxBits/s Router A->B(bps)
Abbreviated data showing source & destrouter IPs and hourly max-15-minute values:
63.45.173.83,173.27.44.02,2639.64453125,2858.09765625,15155.2001953125,10594.986328125,21189.97265625,8747.2353515625,104866.703125,136815.5,31976.107421875,12642.986328125,8510.578125,6489.88427734375,8192.0
![Page 39: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/39.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 39
Destination Class Usage (DCU)
![Page 40: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/40.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 40
Destination Class Usage (DCU)
• Juniper specific!• Policy based accounting mechanism
– For example based on BGP communities
• Supports up to 16 different traffic destinationclasses
• Maintains per interface packet and bytecounters to keep track of traffic per class
• Data is stored in a file on the router, and canbe pushed to a collector
• But…• 16 destination classes is in most cases too
limited to build a useful full Traffic Matrix
![Page 41: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/41.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 41
DCU Example
• Routing policy– associate routes from provider A with DCU class 1– associate routes from provider B with DCU class 2
• Perform accounting on PE
![Page 42: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/42.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 42
BGP Policy Accounting
![Page 43: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/43.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 43
BGP Policy Accounting
• Accounting traffic according to the route ittraverses
• Account for IP traffic by assigning countersbased on:– BGP community-list– AS number– AS-path– destination IP address
• 64 buckets• Similar to Juniper DCU
![Page 44: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/44.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 44
MPLS Based Methods
![Page 45: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/45.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 45======!"§==Systems=
MPLS Based Methods
Two methods to determine traffic matrices:
Using RSVP-TE tunnels
Using LDP statistics
Some comments on DT’s practical implementation
Traffic Matrices in partial topologies
![Page 46: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/46.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 46======!"§==Systems=
RSVP-TE in MPLS Networks
RSVP-TE (RFC 3209) can be used to establish LSPs
Example (IOS)
interface Tunne99 description RouterA => RouterB tag-switching ip tunnel destination 3.3.3.3 tunnel mode mpls traffic-eng tunnel mpls traffic-eng priority 5 5 tunnel mpls traffic-eng bandwidth 1 tunnel mpls traffic-eng path-option 3 explicit identifier 17 tunnel mpls traffic-eng path-option 5 dynamic!ip explicit-path identifier 17 enable next-address 1.1.1.1 next-address 2.2.2.2 next-address 3.3.3.3!
![Page 47: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/47.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 47======!"§==Systems=
RSVP-TE in MPLS NetworksHow to Obtain the Traffic Matrix (TM)?
Explicitly routed Label Switched Paths (TE-LSP) haveassociated byte counters;
A full mesh of TE-LSPs enables to measure the trafficmatrix in MPLS networks directly;
![Page 48: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/48.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 48======!"§==Systems=
RSVP-TE in MPLS NetworksPro’s and Con’s
Advantage: Method that comes closest a traffic matrixmeasurement.
Disadvantages:
A full mesh of TE-LSPs introduces an additionalrouting layer with significant operational costs;
Emulating ECMP load sharing with TE-LSPs is difficultand complex:
Define load-sharing LSPs explicitly;
End-to-end vs. local load-sharing;
Only provides Internal Traffic Matrix, no Router/PoPto peer traffic
![Page 49: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/49.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 49======!"§==Systems=
Traffic matrices with LDP statistics
MPLS Header IP Packet
1234 . . .
1235 . . .
10.10.10.1/32FEC
…………PO1/2400012351234OutIntBytesOutLabelInLabel
4124
•In a MPLS network, LDP can be used to distribute labelinformation;
•Label-switching can be used without changing the routingscheme (e.g. IGP metrics);
•Many router operating systems provide statistical dataabout bytes switched in each forwarding equivalence class(FEC):
![Page 50: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/50.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 50======!"§==Systems=
Traffic matrices with LDP statisticsUse of ECMP load-sharing
5
4
3
21
Router 1:999 10 ..... PO1/0
Router 2:10 11 ..... PO1/010 12 ..... P03/0
Router 3:11 20 ..... PO1/0
Router 4:12 21 ..... PO1/0
Router 5:20 999 .... PO1/021 999 ..... P03/0
1011
12
21
20
1 2
3
4
5
![Page 51: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/51.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 51======!"§==Systems=
Traffic matrices with LDP statistics
•The given information allows for a forward chaining;
•For each router and FEC a set of residual paths can becalculated (given the topology and LDP information)
•From the LDP statistics we gather the bytes switched oneach residual path;
•Problem: It is difficult to decide whether the router underconsideration is the beginning or transit for a certain FEC;
•Idea: For the traffic matrix TM, add the paths traffic toTM(A,Z) and subtract from TM(B,Z) . [4]
A B Z
![Page 52: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/52.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 52======!"§==Systems=
Traffic matrices with LDP statisticsExample
5
4
3
2120
10
10
10
10
1 2 3 4 5
1 0 0 0 0 0
2 0 0 0 0 0
3 0 0 0 0 0
4 0 0 0 0 0
5 0 0 0 0 0
1 2 3 4 5
1 0 0 0 0 0
2 0 0 0 0 0
3 0 0 0 0 0
4 0 0 0 0 0
5 0 0 10 0 0
Procedure for a demand from router 1 torouter 2 and 20 units of traffic.
3
4
521
1 2 3 4 5
1 0 0 0 0 0
2 0 0 0 0 0
3 0 0 0 0 0
4 0 0 0 0 0
5 0 0 10 10 0
1 2 3 4 5
1 0 0 0 0 0
2 0 0 0 0 0
3 0 0 0 0 0
4 0 0 0 0 0
5 0 0 10 10 0
1 2 3 4 5
1 0 0 0 0 0
2 0 0 0 0 0
3 0 0 0 0 0
4 0 0 0 0 0
5 20 -20 10 10 0
1 2 3 4 5
1 0 0 0 0 0
2 0 0 0 0 0
3 0 0 0 0 0
4 0 0 0 0 0
5 20 0 0 0 0
Router1 Router2
![Page 53: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/53.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 53======!"§==Systems=
Practical ImplementationCisco’s IOS
•LDP statistical data available through “show mplsforwarding” command;
•Problem: Statistic contains no ingress traffic (only transit);
•If separate routers exist for LER- and LSR- functionality, atraffic matrix on the LSR level can be calculated
•A scaling process can be established to compensate amoderate number of combined LERs/LSRs.
LSR-A LSR-B
LER-A1 LER-B1
![Page 54: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/54.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 54======!"§==Systems=
Practical ImplementationJuniper’s JunOS
•LDP statistical data available through “show ldp traffic-statistics” command;
•Problem: Statistic is given only per FECs and not peroutgoing interface;
•As a result one cannot observe the branching ratios for aFEC that is split due to load-sharing (ECMP);
•Assume that traffic is split equally;
•Especially for backbone networks with highly aggregatedtraffic this assumption is met quite accurately.
![Page 55: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/55.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 55======!"§==Systems=
Practical ImplementationResults
•The method has been successfully implemented inDeutsche Telekom’s global MPLS Backbone;
•A continuous calculation of traffic matrices (15minaverages) is accomplished in real-time for a network of 180routers;
•The computation requires only one commodity PC;
•No performance degradation through LDP queries;
•Calculated traffic matrices are used in traffic engineeringand network planning.
![Page 56: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/56.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 56======!"§==Systems=
Practical ImplementationDeployment Process
TM trans- formation (to virtual
topology)
GenerateTopology
RouterConfigs
TM
calculation
LDPData
TM validation/
scaling
LINKUtilizations
Make -TM Process
TM for planning and
traffic engineering
![Page 57: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/57.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 57======!"§==Systems=
Conclusions for LDP method
•Our method can be implemented in a multi-vendornetwork;
•It does not require the definition of explicitly routed LSPs;
•It allows for a continuous calculation;
•There are some restrictions concerning
•vendor equipment;
•network topology.
•See Ref. [4]
![Page 58: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/58.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 58
Traffic Matrices inPartial Topologies
![Page 59: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/59.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 59======!"§==Systems=
Traffic Matrices in Partial Topologies
•In larger networks, it is often important to have a TM for apartial topology (not based on every router)
•Example: TM for core network (planning and TE)
•Problem: TM changes in failure simulations
•Demand moves to another router since actual demand startsoutside the considered topology (red):
C-B C-A C-B C-A
![Page 60: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/60.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 60======!"§==Systems=
Traffic Matrices in Partial Topologies
•The same problem arises with link failures
•Results in inaccurate failure simulations on the reducedtopology
•Metric changes can introduce demand shifts in partialtopologies, too.
•But accurate (failure) simulations are essential forplanning and traffic engineering tasks
![Page 61: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/61.jpg)
T-Systems, BU Technologiezentrum,Dr. St. Schnitter14.06.2004, Seite 61======!"§==Systems=
Traffic Matrices in Partial TopologiesUse of Virtual Edge Router in Simulations
•Introduce virtual edge devices as new start-/endpoints fordemands•Map real demands to virtual edge devices;•Model depends on real topology;•Tradeoff between simulation accuracy and problem size.
V-E
R-A R-B
V-E1 V-E2
C1 C2
![Page 62: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/62.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 62
Estimation Techniques
![Page 63: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/63.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 63
Demand Estimation
• Problem:– Estimate point-to-point demands from measured link
loads
• Network Tomography– Y. Vardi, 1996– Similar to: Seismology, MRI scan, etc.
• Underdetermined system:– N nodes in the network– O(N) links utilizations (known)– O(N2) demands (unknown)
• Must add additional assumptions (information)
![Page 64: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/64.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 64
Example
6 Mbps
B
C
A
y: link utilizationsA: routing matrixx: point-to-point demands
Solve: y = Ax -> In this example: 6 = AB + AC
D
![Page 65: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/65.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 65
Example
Solve: y = Ax -> In this example: 6 = AB + AC
00
6 Mbps
6 Mbps
AC
AB
Additional information
E.g. Gravity Model (everysource sends the samepercentage as all other sourcesof it's total traffic to a certaindestination)
Example: Total traffic sourced atSite A is 50Mbps.Site B sinks 2% of total networktraffic, C sinks 8%.AB =1 Mbps and AC =4 Mbps
Final Estimate: AB = 1.5 Mbps and AC = 4.5 Mbps
![Page 66: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/66.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 66
Real Network: Estimated Demands
International Tier-1IP Backbone
![Page 67: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/67.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 67
Estimated Link Utilizations!
International Tier-1IP Backbone
![Page 68: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/68.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 68
Demand Estimation Results
•Individual demands– Inaccurate estimates…
•Estimated worst-case link utilizations– Accurate!
•Explanation– Multiple demands on the same path indistinguishable,
but their sum is known– If these demands fail-over to the same alternative path,
the resulting link utilizations will be correct
![Page 69: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/69.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 69
Estimation with Measurements
• Estimation techniquescan be used incombination withdemand meadsurements– E.g. NetFlow or partial
MPLS mesh
• This example: Greedysearch to find demandswhich decreases MRE(Mean Relative Error)most.– A small number of
measured demandsaccount for a large dropin MRE
Data from [1]
![Page 70: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/70.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 70
Estimation Summary
• Algorithms have been published– Commercial tools are available– Implement yourself?
• Can be used in multiple scenarios:– Fully estimate Traffic Matrix– Estimate Peering traffic when Core Traffic Matrix is
know– Estimate unknown demands in a network with partial
MPLS mesh (LDP or RSVP)– Combine with NetFlow
• Measure large demands, estimate small ones
• Also see AT&T work– E.g. Nanog29: How to Compute Accurate Traffic
Matrices for Your Network in Seconds [2]
![Page 71: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/71.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 71
Traffic Matrix EstimationCase-Study
![Page 72: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/72.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 72
TM Estimation Case-Study
• Large ISP network– 77 Routers– 166 Circuits
• Known Traffic Matrix– Direct MPLS measurement
• Case-study will evaluate:– How does estimated TM compare to known TM?– How well do tools that require a TM work when given
the estimated TM?
• TM estimation using Cariden MATE Software– Demand Deduction tool
![Page 73: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/73.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 73
Procedure
• Start with current network and known TM– save as “PlanA” (with TM “Known”)
• IGP Simulation for non-failure• Save Link Utilizations and Node In/Out traffic• Estimate Traffic Matrix
– New TM: “Estimated”– Save as “PlanB”
• Do an IGP Metric Optimization on both networks– Using known TM in planA– Using estimated TM in PlanB
• Simulate IGP routing on both optimized networks– using known Traffic matrix for both
• Compare Results!
![Page 74: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/74.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 74
Estimated Demands
![Page 75: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/75.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 75
Worst-Case Link Util. (No. Opt)
• No Metric Optimization• PlanA Traffic Matrix:
– Known
• PlanB Traffic Matrix:– Estimated
• IGP Simulation– Circuit + SRLG failures
• Compare Worst-CaseLink Utilizations (in %)
![Page 76: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/76.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 76
Normal Link Utilizations (Opt.)
• IGP Metric Optimization– PlanA Traffic Matrix:
• Known
– PlanB bandwidth level:• Estimated
• IGP Simulation– PlanA Traffic Matrix:
• Known
– PlanB bandwidth level:• Original
• Compare Base LinkUtilizations (in %)– non-failure
![Page 77: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/77.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 77
Normal Link Utilizations (Opt.)
• Scenario: same asprevious slide
• Compare Sorted LinkUtilizations– non-failure
• Colors:– based on measured
demands: BLUE– based on estimated
demands: RED
![Page 78: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/78.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 78
Worst-Case Link Utilizations (Opt)
• Scenario: same• Compare Worst-Case
Link Utilizations (in %)– Circuits + SRLG failures
![Page 79: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/79.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 79
Worst-Case Link Utilizations (Opt)
• Scenario: same• Compare Sorted Worst-
Case Link Utilizations(in %)– Circuits + SRLG failures
• Colors:– based on measured
demands: BLUE– based on estimated
demands: RED
![Page 80: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/80.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 80
TM Estimation Case-Study
• Works very well on this ISP topology/traffic!– Also on AT&T, and all other networks we tried
• Even more accurate if used in combinationwith demand measurements– E.g. from NetFlow, DCU or MPLS
![Page 81: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/81.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 81
Summary &Conclusions
![Page 82: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/82.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 82
Overview
• “Traditional” NetFlow (Version 5)– Requires a lot of resources for collection and
processing– Not trivial to convert to Traffic Matrix
• BGP NextHop Aggregation NetFlow providesalmost direct measurement of the TrafficMatrix– Verion 9 export format– Only supported by Cisco in newer IOS versions
• Juniper DCU is too limited (only 16 classes) tobuild a full Traffic Matrix– But could be used as adjunct to TM Estimation
![Page 83: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/83.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 83
Overview
• MPLS networks provide easy access to theTraffic Matrix– Directly measure in RSVP TE networks– Derive from switching counters in LDP network
• Very convenient if you already have an MPLSnetwork, but no reason to deploy MPLS justfor the TM
• Estimation techniques can provide reliableTraffic Matrix data– Very useful in combination with partially know Traffic
Matrix (e.g. NetFlow, DCU or MPLS)
![Page 84: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/84.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 84
Contact
Thomas TelkampCariden Technologies, [email protected]
Stefan [email protected]
![Page 85: Best Practices for Determining the Traffic Matrix in IP ...apricot.net/apricot2005/slides/T4-4.pdf · APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial](https://reader033.fdocuments.us/reader033/viewer/2022042108/5e87975644a3d6504065836a/html5/thumbnails/85.jpg)
APRICOT 2005: Best Practices for Determining the Traffic Matrix ... Tutorial 85
References
1. A. Gunnar, M. Johansson, and T. Telkamp, “Traffic Matrix Estimation on aLarge IP Backbone - A Comparison on Real Data”, Internet MeasurementConference 2004. Taormina, Italy, October 2004.
2. Yin Zhang, Matthew Roughan, Albert Greenberg, David Donoho, Nick Duffield,Carsten Lund, Quynh Nguyen, and David Donoho, “How to Compute AccurateTraffic Matrices for Your Network in Seconds”, NANOG29, Chicago, October2004.
3. AT&T Tomogravity page: http://www.research.att.com/projects/tomo-gravity/
4. S. Schnitter, T-Systems; M. Horneffer, T-Com. “Traffic Matrices for MPLSNetworks with LDP Traffic Statistics.” Proc. Networks 2004, VDE-Verlag 2004.
5. Y. Vardi. “Network Tomography: Estimating Source-Destination TrafficIntensities from Link Data.” J.of the American Statistical Association, pages365–377, 1996.