Introduction 1-1 Chapter 1 Part 3 Delay, loss and throughput These slides derived from Computer Networking: A Top Down Approach, 6 th edition. Jim Kurose, Keith Ross Addison-Wesley, March Comp 365 Computer Networks Fall 2014

Introduction 1-2 Chapter 1: roadmap 1.1 What is the Internet? 1.2 Network edge  end systems, access networks, links 1.3 Network core  circuit switching, packet switching, network structure 1.4 Delay, loss and throughput in packet-switched networks 1.5 Protocol layers, service models 1.6 Networks under attack: security 1.7 History

Delay, Loss, and Throughput  Three concepts important at all layers of networks  Want to minimize delay and loss, maximize throughput Introduction 1-3

Introduction 1-4 How do loss and delay occur? packets queue in router buffers  packet arrival rate to link exceeds output link capacity  packets queue, wait for turn A B packet being transmitted (delay) packets queueing (delay) free (available) buffers: arriving packets dropped (loss) if no free buffers

Introduction 1-5 Four sources of packet delay  nodal processing  queueing  Transmission delay  Propagation delay A B propagation transmission nodal processing queueing We’ll study routers in detail in chapter 4

Introduction 1-6 Four sources of packet delay  1. d proc (nodal processing):  check bit errors  determine output link  Typical delay: < msec A B propagation transmission nodal processing queueing We’ll study routers in detail in chapter 4

Introduction 1-7 Four sources of packet delay A B propagation transmission nodal processing queueing  2. d queue (queueing)  time waiting at output link for transmission  depends on congestion level of router  Typical delay: micro to milliseconds We’ll study routers in detail in chapter 4

Introduction 1-8 Delay in packet-switched networks 3. d trans (Transmission delay):  R=link bandwidth (bps)  10Mbps ethernet, R=10Mbps  L=packet length (bits)  time to send bits into link d trans = L/R  Typical: micro to millisec A B propagation transmission nodal processing queueing

Introduction 1-9 Delay in packet-switched networks 4. d prop (Propagation delay):  Depends on the medium (fiber optics, twisted-pair, copper wire, etc.)  d = length of physical link  s = propagation speed in medium (~2x10 8 m/sec)  propagation delay d prop = d/s  Typical values of s: 2x10 8 meters/sec to 3x10 8 m/s  Typical delay: milliseconds A B propagation transmission nodal processing queueing Note: s and R are very different quantities!

Introduction 1-10 Caravan analogy  cars “propagate” at 100 km/hr (propagation speed)  toll booth takes 12 sec to service car (transmission time)  car~bit; caravan ~ packet toll booth toll booth ten-car caravan 100 km

Introduction 1-11 Caravan analogy  Q: How long until caravan is lined up before 2nd toll booth?  Time to “push” entire caravan through toll booth onto highway =  12*10 = 120 sec  Time for last car to propagate from 1st to 2nd toll both:  100km/(100km/hr)= 1 hr  A: 62 minutes toll booth toll booth ten-car caravan 100 km

Introduction 1-12 Caravan analogy (more)  Cars now “propagate” at 1000 km/hr  Toll booth now takes 1 min to service a car  Q: Will cars arrive to 2nd booth before all cars serviced at 1st booth? toll booth toll booth ten-car caravan 100 km

Introduction 1-13 Caravan analogy (more)  Cars now “propagate” at 1000 km/hr  Toll booth now takes 1 min to service a car  Q: Will cars arrive to 2nd booth before all cars serviced at 1st booth?  Yes! After 7 min, 1st car at 2nd booth and 3 cars still at 1st booth.  1st bit of packet can arrive at 2nd router before packet is fully transmitted at 1st router!  See Ethernet applet at AWL Web site toll booth toll booth ten-car caravan 100 km

Introduction 1-14 Nodal delay  d proc = processing delay  typically a few microsecs or less  d queue = queuing delay  depends on congestion  d trans = transmission delay  = L/R, significant for low-speed links  d prop = propagation delay  a few microsecs to hundreds of msecs

Example: End-to-end delay  Do Interactive Exercise “one-hop delay” from the online student resources: cw/index.html Introduction 1-15

Example: End-to-end delay  Do Interactive Exercise “End-to-end delay” from the online student resources: cw/index.html Introduction 1-16

Introduction 1-17 Queueing delay (revisited)  R=link bandwidth (bps)  this is transmission time  L=packet length (bits)  a=average packet arrival rate traffic intensity = La/R  La/R ~ 0: average queueing delay small  La/R -> 1: delays become large  La/R > 1: more “work” arriving than can be serviced, average delay infinite!

Introduction 1-18 “Real” Internet delays and routes  What do “real” Internet delay & loss look like?  Traceroute program: provides delay measurement from source to router along end-end Internet path towards destination. For all i:  sends three packets that will reach router i on path towards destination  router i will return packets to sender  sender times interval between transmission and reply. 3 probes Traceroute is available at (already installed in OS X) See also the graphical interface to Traceroute called PingPlotterhttp:// Traceroute is available at (already installed in OS X) See also the graphical interface to Traceroute called PingPlotterhttp://

Introduction 1-19 “Real” Internet delays and routes 1 cs-gw ( ) 1 ms 1 ms 2 ms 2 border1-rt-fa5-1-0.gw.umass.edu ( ) 1 ms 1 ms 2 ms 3 cht-vbns.gw.umass.edu ( ) 6 ms 5 ms 5 ms 4 jn1-at wor.vbns.net ( ) 16 ms 11 ms 13 ms 5 jn1-so wae.vbns.net ( ) 21 ms 18 ms 18 ms 6 abilene-vbns.abilene.ucaid.edu ( ) 22 ms 18 ms 22 ms 7 nycm-wash.abilene.ucaid.edu ( ) 22 ms 22 ms 22 ms ( ) 104 ms 109 ms 106 ms 9 de2-1.de1.de.geant.net ( ) 109 ms 102 ms 104 ms 10 de.fr1.fr.geant.net ( ) 113 ms 121 ms 114 ms 11 renater-gw.fr1.fr.geant.net ( ) 112 ms 114 ms 112 ms 12 nio-n2.cssi.renater.fr ( ) 111 ms 114 ms 116 ms 13 nice.cssi.renater.fr ( ) 123 ms 125 ms 124 ms 14 r3t2-nice.cssi.renater.fr ( ) 126 ms 126 ms 124 ms 15 eurecom-valbonne.r3t2.ft.net ( ) 135 ms 128 ms 133 ms ( ) 126 ms 128 ms 126 ms 17 * * * 18 * * * 19 fantasia.eurecom.fr ( ) 132 ms 128 ms 136 ms traceroute: gaia.cs.umass.edu to Three delay measurements from gaia.cs.umass.edu to cs-gw.cs.umass.edu * means no response (probe lost, router not replying) trans-oceanic link

traceroute  How do you use it? A230247$ traceroute gaia.cs.umass.edu Introduction 1-20

traceroute  How does it work? (from wikipedia): Traceroute works by increasing the “time-to-live ” value of each successive batch of packets sent. The first three packets sent have a time-to-live (TTL) value of one (implying that they are not forwarded by the next router and make only a single hop). The next three packets have a TTL value of 2, and so on. When a packet passes through a host, normally the host decrements the TTL value by one, and forwards the packet to the next host. When a packet with a TTL of one reaches a host, the host discards the packet and sends an ICMP time exceeded (type 11) packet to the sender, or an echo reply (type 0) if its IP address matches the IP address that the packet was originally sent to. The traceroute utility uses these returning packets to produce a list of hosts that the packets have traversed in transit to the destination. The three timestamp values returned for each host along the path are the delay (aka latency) values typically in milliseconds (ms) for each packet in the batch. Introduction 1-21

“Real” Internet delays and routes  Exercise  Each person pick a different time  Each person pick a different destination (continental or outside continent)  Use traceroute or PingPlotter  Report back results tomorrow Introduction 1-22

Introduction 1-23 Packet loss  queue (aka buffer) preceding link in buffer has finite capacity  packet arriving to full queue dropped (aka lost)  lost packet may be retransmitted by previous node, by source end system, or not at all A B packet being transmitted packet arriving to full buffer is lost buffer (waiting area)

Other Delays  Dial-up modems have large encoding/decoding delays (other technologies don’t)  Some protocols purposely delay transmission (to share medium). Chap 5.  Media packetization delay (as in VOIP)  Must digitize speech & encode it Introduction 1-24

1-25 Throughput  throughput: rate (bits/time unit) at which bits transferred between sender/receiver  instantaneous: rate at given point in time at which the destination host is receiving the file  average: rate over longer period of time File = F bits transfer takes T seconds for host to receive average throughput = F/T bits/sec Some downloading apps display the instantaneous rate as you download

1-26 Throughput  throughput: rate (bits/time unit) at which bits transferred between sender/receiver  throughput is determined by the transmission and propagation rates of all the switches and links and by the delays encountered  With throughput, however, we don’t look for individual rates/delays but just measure how fast bits arrive.  throughput is a coarse-grained measure

Introduction 1-27 Throughput  throughput: rate (bits/time unit) at which bits transferred between sender/receiver server, with file of F bits to send to client link capacity R s bits/sec link capacity R c bits/sec pipe that can carry fluid at rate R s bits/sec) pipe that can carry fluid at rate R c bits/sec) server sends bits (fluid) into pipe the server can pump R s bits through the first pipe and the router can pump R c bits through the second pipe Think of throughput as the width of the pipe not the length of the pipe.

Introduction 1-28 Throughput (more)  R s < R c What is average end-end throughput? R s bits/sec R c bits/sec link on end-end path that constrains end-end throughput bottleneck link The server can only pump R s bits through its pipe. The router could pump more, but it’s only receiving R s bits so it can only pump R s bits.

Introduction 1-29 Throughput (more)  R s > R c What is average end-end throughput? R s bits/sec R c bits/sec link on end-end path that constrains end-end throughput bottleneck link The router receives R s bits, but can only pump R c bits so the rate at which the client receives bits is R c

Throughput  For F bits and rates R s and R c what is the (approx) time it takes to transfer a file?  F/{min R s, R c }  For F bits and rates R 0, R 1, … R n how long does it take?  F/{min R 0, R 1,…,R n } Introduction 1-30

Introduction 1-31 Throughput: Internet scenario 10 connections (fairly) share backbone bottleneck link R bits/sec RsRs RsRs RsRs RcRc RcRc RcRc R  per-connection end-end throughput: min(R c,R s,R/10)  in practice: R c or R s is often bottleneck Backbone is, in general, over provisioned; seldom causes delay. Tier 2, 3, etc. cause delay.

Introduction 1-32 Throughput: Internet scenario 10 connections (fairly) share backbone bottleneck link R bits/sec RsRs RsRs RsRs RcRc RcRc RcRc R  Assume all connections except R c and R s and R are very large  if R >> R c and R s the bottleneck is R c or R s Backbone is, in general, over provisioned; seldom causes delay. Tier 2, 3, etc. cause delay.

Introduction 1-33 Throughput: Internet scenario 10 connections (fairly) share backbone bottleneck link R bits/sec RsRs RsRs RsRs RcRc RcRc RcRc R  Now assume R c = 1 Mbps, R s = 2 Mbps, and R = 5 Mbps  The bottleneck is now the shared link, R  If each download gets about the same amount of packets through R, the rate for each is 5 Mbps/10 = 500 kbps

Throughput: Example  Do Interactive Exercise “End to End Throughput” from cw/index.html cw/index.html Introduction 1-34