3.7 TCP congestion. reliable data transfer. sliding window. Lecture 4: Transport layer III: flow control and congestion control & Network layer I: IP

Transcription

1 TDTS06 Computer s Lecture 4: Transport layer III: flow control and congestion control & Network layer I: IP Juha Takkinen, juha.takkinen@liu.se IDA/ADIT/IISLAB, Linköpings universitet Slides are modified from J.F Kurose and K.W. Ross TDTS06 Lecture 4: flow control and congestion control; ip 3-1 TCP Flow control (Suppose TCP receiver discards out-of-order segments) spare room in buffer = RcvWindow = RcvBuffer-[LastByteRcvd - LastByteRead] Rcvr advertises spare room by including value of RcvWindow in segments Sender limits unacked data to RcvWindow guarantees receive buffer doesn t overflow flow control sender won t overflow receiver s buffer by transmitting too much, too fast TDTS06 Lecture 4: flow control and congestion control; ip 3-3 Chapter 3 outline 3.1 Transport-layer services (done!) 3.2 Multiplexing and demultiplexing (done!) 3.3 Connectionless transport: UDP (done!) 3.4 Principles of reliable data transfer (done!) sliding window 3.5 Connection-oriented transport: TCP segment structure reliable data transfer flow control connection management 3.6 Principles of congestion control 3.7 TCP congestion control TDTS06 Lecture 4: flow control and congestion control; ip 3-2 TCP Connection Management: opening a connection client: connection initiator Socket clientsocket = new Socket("hostname","port number"); server: contacted by client Socket connectionsocket = welcomesocket.accept(); TDTS06 Lecture 4: flow control and congestion control; ip 3-4 1

2 TCP Connection Management: closing a connection Step 1: client end system sends TCP FIN control segment to server close client server Step 2: server receives FIN, replies with ACK. Closes connection, sends FIN. close client closes socket: clientsocket.close(); closed TDTS06 Lecture 4: flow control and congestion control; ip 3-5 TCP Connection Management (cont.) Step 3: client receives FIN, replies with ACK. closing client server Enters timed wait - will respond with ACK to received FINs Step 4: server, receives ACK. Connection closed. Note: with small modification, can handle simultaneous FINs. closed closing closed TDTS06 Lecture 4: flow control and congestion control; ip FSM: TCP Connection Management timed wait timed wait TDTS06 Lecture 4: flow control and congestion control; ip 3-7 Principles of Congestion Control Congestion: informally: too many sources sending too much data too fast for to handle different from flow control! manifestations: lost packets (buffer overflow at routers) long delays (queueing in router buffers) a top-10 problem! TDTS06 Lecture 4: flow control and congestion control; ip 3-8

3 Causes/costs of congestion Host B Host A λ in : original data unlimited shared output link buffers 1. Queueing delay becomes infinite close to max rate of link 2. With finite queues a) Packets may be dropped/lost and must therefore be retransmitted λ ou b) Packets may be unnecessarily retransmitted too, because of long delay and subsequent timeout 3. With competing packet flows, resource may be wasted when a packet that was forwarded later is dropped/lost by a downstream router TDTS06 Lecture 4: flow control and congestion control; ip 3-9 TCP congestion control: additive increase, multiplicative decrease (AIMD) Approach: increase transmission rate (window size), probing for usable bandwidth, until loss occurs additive increase: increase CongWin by 1 MSS every RTT until loss detected multiplicative decrease: cut CongWin in half after loss 24 Kbytes 16 Kbytes 8 Kbytes congestion window time time TDTS06 Lecture 4: flow control and congestion control; ip 3-11 Approaches towards congestion control Two broad approaches towards congestion control: Host-centric (end-end) congestion control: no explicit feedback from congestion inferred from endsystem observed loss, delay approach taken by TCP Network-centric congestion control: routers provide feedback to end systems single bit indicating congestion (SNA, DECbit, TCP/IP ECN, ATM) explicit rate sender should send at TDTS06 Lecture 4: flow control and congestion control; ip 3-10 t 3 congestion window size Saw tooth behavior: probing for bandwidth TCP Congestion Control: details sender limits transmission: LastByteSent-LastByteAcked CongWin Roughly, rate = CongWin RTT Bytes/sec CongWin is dynamic, function of perceived congestion How does sender perceive congestion? loss event = timeout or 3 duplicate acks TCP sender reduces rate (CongWin) after loss event three mechanisms: AIMD (see previous slide) slow start conservative after timeout events TDTS06 Lecture 4: flow control and congestion control; ip 3-12

4 TCP Slow Start When connection begins, increase rate exponentially until first loss event: Host A Host B double CongWin every RTT done by incrementing CongWin for every ACK received Summary: initial rate is slow but ramps up exponentially fast TDTS06 Lecture 4: flow control and congestion control; ip RTT time Refinement: inferring loss After 3 dup ACKs: CongWin is cut in half window then grows linearly But after timeout event: CongWin instead set to 1 MSS; window then grows exponentially to a threshold, then grows linearly Philosophy: 3 dup ACKs indicates capable of delivering some segments timeout indicates a more alarming congestion scenario TDTS06 Lecture 4: flow control and congestion control; ip 3-15 Refinement Q: When should the exponential increase switch to linear? A: When CongWin gets to 1/2 of its value before timeout. Implementation: Variable Threshold At loss event, Threshold is set to 1/2 of CongWin just before loss event TDTS06 Lecture 4: flow control and congestion control; ip 3-14 Summary: TCP Congestion Control When CongWin is below Threshold, sender in slow-start phase, window grows exponentially. When CongWin is above Threshold, sender is in congestion-avoidance phase, window grows linearly. When a triple duplicate ACK occurs, Threshold set to CongWin/2 and CongWin set to Threshold. When timeout occurs, Threshold set to CongWin/2 and CongWin is set to 1 MSS TDTS06 Lecture 4: flow control and congestion control; ip 3-16

5 Chapter 3: Summary principles behind transport layer services: multiplexing, demultiplexing reliable data transfer flow control congestion control instantiation and implementation in the Internet UDP TCP Next: leaving the edge (application, transport layers) into the core TDTS06 Lecture 4: flow control and congestion control; ip 3-17 Källa: Deering/IETF, 2001 Chapter 4: Network Layer 4. 1 Introduction 4.2 Virtual circuit and datagram s 4.4 IP: Internet Protocol Datagram format IPv4 addressing ICMP IPv6 (lecture 10) 4.5 Routing algorithms (lecture 5) Link state Distance vector Hierachical routing 4.6 Routing in the Internet (lecture 6) RIP OSPF BGP TDTS06 Lecture 4: flow control and congestion control; ip 3-18 IP Addressing: introduction IP address: 32-bit identifier for host, router interface interface: connection between host/router and link router s typically have multiple interfaces host typically has one interface IP addresses associated with each interface = TDTS06 Lecture 4: flow control and congestion control; ip

6 IP datagram format IP protocol version number header length (bytes) type of data max number remaining hops (decremented at each router) upper layer protocol to deliver payload to how much overhead with TCP? 20 bytes of TCP 20 bytes of IP = 40 bytes + app layer overhead head. len 32 bits ver length 16-bit identifier time to live type of service upper layer flgs data (variable length, typically a TCP or UDP segment) header checksum 32 bit source IP address fragment offset 32 bit destination IP address total datagram length (bytes) for fragmentation/ reassembly Options (if any) E.g. timestamp, record route taken, specify list of routers to visit. Network layer transport segment from sending to receiving host on sending side encapsulates segments into datagrams on rcving side, delivers segments to transport layer layer protocols in every host, router router examines header fields in all IP datagrams passing through it application transport application transport TDTS06 Lecture 4: flow control and congestion control; ip TDTS06 Lecture 4: flow control and congestion control; ip 3-23 Interplay between routing and forwarding The service model of the layer value in arriving packet s header routing algorithm local forwarding table header value output link Q: Whatservice model to use for the channel that transports a datagram from A to B? Example, individual datagram: Guaranteed delivery Guaranteed delivery with less thatn 40 ms delay Example, flow of several datagrams Delivery of datagrams in order Guaranteed minimum bandwidth to flow Req. on max interval between datagarams TDTS06 Lecture 4: flow control and congestion control; ip 3-24 What does Internet have? 6

7 What does Internet have? Packet-switching, Best effort IP Fragmentation & Reassembly Why? Network links have MTU (max.transfer size) - largest possible link-level frame. different link types, different MTUs How? Large IP datagram divided ( fragmented ) within net Reassembly at host fragmentation: in: one large datagram out: 3 smaller datagrams one datagram becomes several datagrams reassembled only at final destination IP header bits used to identify, order related fragments TDTS06 Lecture 4: flow control and congestion control; ip 3-28 Packet-switching No connection/handshaking phase Routers: No information about state of end-point connections The concept of connection is missing in the layer Packets are forwarded based on destination address Packets between the same source-destination pairs can take different routes application transport link 1. Send data 2. Receive data application transport link IP Fragmentation and Reassembly Example 4000 byte datagram MTU = 1500 bytes length =4000 ID =x fragflag =0 offset =0 One large datagram becomes several smaller datagrams 1480 bytes in data field offset = 1480/8 length =1500 length =1500 length =1040 ID =x ID =x ID =x fragflag =1 fragflag =1 fragflag =0 offset =0 offset =185 offset = TDTS06 Lecture 4: flow control and congestion control; ip

8 Forwarding table, a naïve example 4 billion possible entries Destination Address Range Link Interface through through through otherwise TDTS06 Lecture 4: flow control and congestion control; ip 3-30 Subnets IP address: subnet part (high order bits) host part (low order bits) What s a subnet? device interfaces with same subnet part of IP address can ly reach each other without intervening router subnet consisting of 3 subnets TDTS06 Lecture 4: flow control and congestion control; ip 3-32 A better/commpressed example: Longest prefix matching Prefix Match Link Interface otherwise 3 Examples DA: Which interface? DA: Which interface? TDTS06 Lecture 4: flow control and congestion control; ip 3-31 Subnets How many? TDTS06 Lecture 4: flow control and congestion control; ip

9 IP addressing: CIDR CIDR: Classless InterDomain Routing subnet portion of address of arbitrary length address format: a.b.c.d/x, where x is # bits in subnet portion of address subnet part host part / TDTS06 Lecture 4: flow control and congestion control; ip 3-34 DHCP client-server scenario A DHCP server B E arriving DHCP client needs address in this TDTS06 Lecture 4: flow control and congestion control; ip 3-36 IP addresses: how to get one? Q: How does host get IP address? hard-coded by system admin in a file Wintel: control-panel->->configuration- >tcp/ip->properties UNIX: /etc/rc.config DHCP: Dynamic Host Configuration Protocol: dynamically get address from as server plug-and-play TDTS06 Lecture 4: flow control and congestion control; ip 3-35 DHCP client-server scenario DHCP server: arriving DHCP discover client src : , 68 dest.: ,67 yiaddr: transaction ID: 654 DHCP offer src: , 67 dest: , 68 yiaddrr: transaction ID: 654 Lifetime: 3600 secs DHCP request time src: , 68 dest:: , 67 yiaddrr: transaction ID: 655 Lifetime: 3600 secs DHCP ACK src: , 67 dest: , 68 yiaddrr: transaction ID: 655 Lifetime: 3600 secs TDTS06 Lecture 4: flow control and congestion control; ip

10 IP addresses: how to get one? Q: How does get subnet part of IP addr? A: gets allocated portion of its provider ISP s address space ISP's block /20 Organization /23 Organization /23 Organization / Organization / TDTS06 Lecture 4: flow control and congestion control; ip 3-38 Hierarchical addressing: more specific routes ISPs-R-Us has a more specific route to Organization 1 Organization /23 Organization / Organization /23 Fly-By-Night-ISP Send me anything with addresses beginning /20 Internet Organization /23 ISPs-R-Us Send me anything with addresses beginning /16 or / TDTS06 Lecture 4: flow control and congestion control; ip 3-40 Hierarchical addressing: route aggregation Hierarchical addressing allows efficient advertisement of routing information: Organization /23 Organization /23 Organization / Organization /23 Fly-By-Night-ISP Send me anything with addresses beginning /20 Internet ISPs-R-Us Send me anything with addresses beginning / TDTS06 Lecture 4: flow control and congestion control; ip 3-39 IP addressing: the last word... Q: How does an ISP get block of addresses? A: ICANN: Internet Corporation for Assigned Names and Numbers allocates addresses manages DNS assigns domain names, resolves disputes TDTS06 Lecture 4: flow control and congestion control; ip

11 The Internet Network layer Host, router layer functions: Transport layer: TCP, UDP Network layer Routing protocols path selection RIP, OSPF, BGP forwarding table IP protocol addressing conventions datagram format packet handling conventions ICMP protocol error reporting router signaling Link layer layer TDTS06 Lecture 4: flow control and congestion control; ip 3-42 Traceroute and ICMP Source sends series of UDP segments to dest First has TTL =1 Second has TTL=2, etc. Unlikely port number When nth datagram arrives to nth router: Router discards datagram And sends to source an ICMP message (type 11, code 0) Message includes name of router& IP address When ICMP message arrives, source calculates RTT Traceroute does this 3 times Stopping criterion UDP segment eventually arrives at destination host Destination returns ICMP host unreachable packet (type 3, code 3) When source gets this ICMP, stops. 3 probes 3 probes 3 probes TDTS06 Lecture 4: flow control and congestion control; ip 3-44 ICMP: Internet Control Message Protocol used by hosts & routers to communicate -level information error reporting: unreachable host,, port, protocol echo request/reply (used by ping) -layer above IP: ICMP msgs carried in IP datagrams ICMP message: type, code plus first 8 bytes of IP datagram causing error Type Code description 0 0 echo reply (ping) 3 0 dest. unreachable 3 1 dest host unreachable 3 2 dest protocol unreachable 3 3 dest port unreachable 3 6 dest unknown 3 7 dest host unknown 4 0 source quench (congestion control - not used) 8 0 echo request (ping) 9 0 route advertisement 10 0 router discovery 11 0 TTL expired 12 0 bad IP header TDTS06 Lecture 4: flow control and congestion control; ip 3-43 Traceroute example traceroute: gaia.cs.umass.edu to Three delay measurements from gaia.cs.umass.edu to cs-gw.cs.umass.edu 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 trans-oceanic ( ) 104 ms 109 ms 106 ms 9 de2-1.de1.de.geant.net ( ) 109 ms 102 ms 104 ms link 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 * * * * means no response (probe lost, router not replying) 19 fantasia.eurecom.fr ( ) 132 ms 128 ms 136 ms TDTS06 Lecture 4: flow control and congestion control; ip