mopdf.com

Capacity planning and .Hints on capacity planning(and other approaches)Andrea BiancoTelecommunication Network tica.polito.it/Andrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 1Capacity (re)planning Design from scratch (or modify) network topology, linkcapacity, routing, to– efficiently use resources– satisfy user requirements In circuit switching networks more standardapproaches In packet switched networks mostly done by trial anderror approaches Steps– Define node position and estimate (or measure) user traffic(during busy hour?) to derive traffic matrix– Define logical topology Includes logical link capacity assignment– (Map to a physical topology)Andrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 2Traffic models as a key to enable capacityassignement Traffic model: how users and aggregation of usersbehave Examples– How long a user uses an ADSL modem– Average size of a file transfer Models change with network usage (and applications)– Guess about the future Models are based on– Measurements– Estimates ( Guesses )Andrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 3Pag. 1

Capacity planning and .Telephone traffic models Call dynamics?– Call arrival model– Shown that call inter-arrival time to be well approximated by an exponentialdistribution– Call arrival process follows a Poisson distribution Memoryless: the fact that a certain amount of time has passed since last callgives no information of time to next call Call duration?– Also modelled as an exponential distribution– Some measurements show it is heavy tailed A non negligible number of calls last for a very long time– Normally neglected Poisson models are easy to manage and well acceptedAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 4Internet traffic models More difficult– Many different applications (although all rely on file transferover UDP and TCP)– Few applications account for most of the traffic But this changes over time Web, P2P– Difficult to model destination distribution Two main features– LAN connections are different from WAN connections Higher bandwidth and longer holding times– Many parameters are heavy tailed #bytes in a callCall durationFew calls are responsible for most of the trafficMeans that even large aggregate of traffic are not smooth (law of largenumbers)Andrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 5Telephone network capacity planning How to size a link to obtain a blocking probability smallerthat a target value Erlang-B formula gives blocking probability as a functionof– Avg number of calls (in erlangs) on a link– Avg call arrival rate r– Avg call holding time h Call load E r h––––Trunk capacity mInfinite number of sourcesm 5, E 3, blocking probability 0.11For a fixed load, as m increases the call blocking probabilitydecreases exponentiallyAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 6Pag. 2

Capacity planning and .Telephone network capacity planning Blocking probability along a path Assume traffic on links is independent– Blocking probability is the product of the probability on eachlink Routing table and traffic matrix determine the load on alink Assign capacity to each link given the load and targetblocking probability– Or add new link– And/or change routing tableAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 7Packet switching (?) capacity planning Very complex– often relies on trial and error procedures Planning problem often divided in two problems– Logical Topology Design– Routing (and Wavelength) Assignment May be formalized as an optimization problem– Joint optimization Unfeasible for complexity and organizational reasons Heuristics– Two step formulation Each step may independently be formalized as an optimizationproblem Often heuristics needed for the two separate problems– See class “Operations research: theory and applications tonetworking” for LTD and RWA examplesAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 8Measure or estimate trafficto create traffic matrix Build for the worst case (?)– Pick the busiest hour (over which time-scale?) Add some safety margin to allow for– Measurement or estimate errors– Future traffic growth– Provisioning against failures? Traffic matrix definition assumes that current pattern predicts futureTime scale critical (traffic over shorter time scale may be heavier)When adding endpoints traffic matrix become obsoleteNot always possible to measure all traffic on all links– Privacy issues Routing policies interact with link load measurements Not always easy to rebuild flow infos from packet infoAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 9Pag. 3

Capacity planning and .Define topology and assign capacity Logical topology definition– Define network connectivity (may include alternative paths forrobustness to failures)– Geographical/cost considerations Some links may be easier to be obtained– Available capacity may impose some constraints (see later) Capacity assignments– Enough capacity to carry traffic defined in the traffic matrix– Actual paths depend on routing May define optimal paths dynamically (look at link load?) Risk of reaction (higher capacity, become more attractive for routing,requires more capacity, etc etc) Easier to assign capacity for static routingAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 10Logical Topology Design Input– Traffic matrix– Costs (e.g. number and speed of links I can pay for, price, performance) Output– Logical topology that “best” suits the traffic matrix with the cost constraints “Extreme” solutions if costs is the number of links– No cost constraint Full meshSingle-hop approachEach node process only generated/received trafficCircuit switching like solution– Minimize costs Tree/ring/star topologiesMulti-hop approachNodes must process also in-transit trafficPermits grooming to “match” traffic load to channel capacityPacket switching like solutionAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 11Routing (and wavelenght) assignment Need to map to a physical topology the defined logical topology– Create physical links, or virtual circuits, or lightpaths– Any difference? Lightpath, if transparent, imply no processing of local traffic Often done by the organization/provider that owns the physicalinfrastructure– Often different from the one that has defined the logical topology Physical topologies may impose constraints on logical topologiesfeasibility due to lack of resources Once established, the logical topology, owners “do not see” thephysical topology– Performance depend on the logical topology layout only Besides selecting routes on the physical topology, in WRnetworks, also wavelengths should be assignedAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 12Pag. 4

Capacity planning and .Trivial examplePhysical topology3 links, VCs, ls are requiredLogical topologyAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 13Examples Start with a traditional first-generation telephone network(which has only point-to-point optical links and electronicswitching)– Typical current solutions consider a SONET/SDH ringtopology, in which OC48-STM16 (2.5 Gb/s) links are used– Add-drop multiplexers (ADM) and digital cross-connects (DCX)are used Then we extend it to a ring that exploit WDM as atransmission technology Finally, we examine the possibility of moving to a WRsolutionAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 14First example The following (normalized) traffic matrix must be 0.51D0.50.50.51.5TOT1111.5 If traffic grows, a capacity increase must be .542.5 a WDM upgrade may be an alternativeAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 15Pag. 5

Capacity planning and .Traditional SONET ringADM Add-Drop MultiplexerDCS Digital Cross-ConnectAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 16WDM ring: WDM on linksAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 17WDM ring Traffic routing # OC-481112111 3 wavelengths are neededAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 18Pag. 6

Capacity planning and .WR ring By using WR, we build a logical topology, in which links areoptical lightpaths, on top of the physical topology, in which pointto-point connections on a ring are available No need to process in-transit trafficAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 19WR ring Optical switches may be used to add flexibility: in caseof traffic changes, or to manage faults, the logicaltopology may changeAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 20Second exampleABCPhysical topology router Without OADM router routerrouter routerThree IP routers, A, B e C, with 10 Gb/s interfaces50 Gb/s of traffic for each router pairWith OADM OADMrouterAndrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 21Pag. 7

Capacity planning and .Second example 10 wavelengths are required in both fibers Without OADMs, the physical topology and thelogical topology are the same, being a bustopology in both cases; router B has 20 interfaces With OADMs, the logical topology is a ring; routerB has only 10 interfaces The cheapest solution depends on the relativecost of components (OADMs vs. router ports)B5AB105CA10C5Andrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 22Third example Three different logical topologies overlaid on the same bidirectional ringphysical topology. Compare them considering:– the number of interfaces at routers– the number of wavelengths– the distance (number of hops) in the logical topology Traffic is uniform, so that each router transmits t% of the capacity of a WDMchannel to all other routers (t/N-1 from router to router)WDM point-to-pointhubfull-mesh(PWDM)(single hub)(fully optical)hub(an additional hub is added)Andrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 23Third exampleNumber ofrouter ports(Q)Number ofwavelengths(W)Distance(number ofhops)PWDMQ 2WHubQ t Full-meshQ (N-1) t/(N1) W t/8(N 1 1/(N-1)) W t N/2W t / N-1 (N2/8 N/4)N/221Andrea Bianco – TNG group - Politecnico di TorinoComputer Network Design and Management - 24Pag. 8