Institute of Telecommunications. Piotr Gajowniczek Andrzej Bąk Michał Jarociński

Transcription

1 Institute of Telecommunications Warsaw University of Technology 2015 internet technologies and standards Piotr Gajowniczek Andrzej Bąk Michał Jarociński

2 Link layer; Ethernet LAN

3 link layer services framing, link access: encapsulate datagram into frame, adding header, trailer channel access if shared medium MAC addresses used in frame headers to identify source, dest different from IP address! reliable delivery between adjacent nodes seldom used on low bit-error link (fiber, some twisted pair) wireless links: high error rates

4 where is the link layer implemented? in each and every host link layer implemented in a network interface card (NIC) or on a chip Ethernet card, card; Ethernet chipset implements link, physical layer combination of hardware, software, firmware application transport network link link physical cpu controller physical transmission memory host bus (e.g., PCI) network adapter card

5 multiple access links, protocols two types of links : point-to-point PPP for dial-up access point-to-point link between Ethernet switch, host broadcast (shared wire or medium) Ethernet upstream HFC wireless LAN shared wire (e.g., cabled Ethernet) shared RF (e.g., WiFi) shared RF (satellite)

6 MAC addresses and ARP 32-bit IP address: network-layer address for interface used for layer 3 (network layer) forwarding MAC (or LAN or physical or Ethernet) address: function: used locally to get frame from one interface to another physically-connected interface (same network, in IP-addressing sense) 48 bit MAC address (for most LANs) burned in NIC ROM, also sometimes software settable e.g.: 1A-2F-BB AD hexadecimal (base 16) notation (each number represents 4 bits)

7 LAN addresses (more) MAC address allocation administered by IEEE manufacturer buys portion of MAC address space (to assure uniqueness) analogy: MAC address: like Social Security Number IP address: like postal address MAC flat address portability can move LAN card from one LAN to another IP hierarchical address not portable address depends on IP subnet to which node is attached 5-7

8 ARP: address resolution protocol Question: how to determine interface s MAC address, knowing its IP address? F7-2B LAN A-2F-BB AD D7-FA-20-B0 ARP table: each IP node (host, router) on LAN has table IP/MAC address mappings for some LAN nodes: < IP address; MAC address; TTL> TTL (Time To Live): time after which address mapping will be forgotten (typically 20 min) C-C4-11-6F-E3-98

9 ARP protocol: same LAN A sens datagram to B B s MAC address not in A s ARP table. A broadcasts ARP query packet, containing B's IP address dest MAC address = FF-FF- FF-FF-FF-FF all nodes on LAN receive ARP query B receives ARP packet, replies to A with its (B's) MAC address frame sent to A s MAC address (unicast) A caches (saves) IP-to- MAC address pair in its ARP table until information becomes old (times out) soft state: information that times out (goes away) unless refreshed ARP is plug-and-play : nodes create their ARP tables without intervention from net administrator

10 addressing: routing to another LAN how to send datagram from A to B via R? assume A knows B s IP address assume A knows IP address of first hop router, R (how?) assume A knows R s MAC address (how?) A C-E8-FF-55 R A-23-F9-CD-06-9B B BD-D2-C7-56-2A CC-49-DE-D0-AB-7D E6-E BB-4B B2-2F-54-1A-0F

11 Ethernet LANs ab:cd:ef:01:23:45 ab:cd:ef:01:23:45 shared medium L2 switch (full-duplex, MAC learning) problem how to economically connect a limited numer of hosts in a limited area with high-speed links solution IEEE 802.x standards, especially Ethernet and WiFi Application TCP IP (Data) Link Ethernet Link MAC frame Preamble SFD Destination Source Type Data FCS (7) 0xAB B 32 = broadcast address: Ethernet ff:ff:ff:ff:ff:ff Medium Access Control CSMA /CD, CSMA /CA (Collision Avoidance), used in (wireless LANs) Ethernet switches full duplex transmission, MAC learning

12 802.3 Ethernet standards: link & physical layers many different Ethernet standards common MAC protocol and frame format different speeds: 2 Mbps, 10 Mbps, 100 Mbps, 1Gbps, 10G bps different physical layer media: fiber, cable application transport network link physical 100BASE-TX MAC protocol and frame format 100BASE-T2 100BASE-FX 100BASE-T4 100BASE-SX 100BASE-BX copper (twister pair) physical layer fiber physical layer

13 VLANs port-based VLAN: switch ports grouped (by switch management software) so that single physical switch Virtual LAN switch(es) supporting VLAN capabilities can be configured to define multiple virtual LANS over single physical LAN infrastructure. Electrical Engineering (VLAN ports 1-8) Computer Science (VLAN ports 9-15) operates as multiple virtual switches Electrical Engineering (VLAN ports 1-8) Computer Science (VLAN ports 9-16)

14 802.1Q VLAN frame format type preamble dest. address source address data (payload) CRC frame type preamble dest. source data (payload) address address CRC 802.1Q frame 2-byte Tag Protocol Identifier (value: 81-00) Recomputed CRC Tag Control Information (12 bit VLAN ID field, 3 bit priority field like IP TOS)

15 MPLS Multiprotocol Label Switching

16 MPLS introduction RFC 3031 MPLS architecture virtualization and management of network resources, network services traffic engineering, QOS high availability consolidation of services on a single infrastructure ability to support various services, applications and solutions over a converged network infrastructure BSAN Aggregation BSA BSR BSAN Broadband Service Access Node BSA Broadband Service Aggregator BSR Broadband Service Router VoIP IP/MPLS Core Video Internet

17 MPLS - introduction IP routing: data link layer frame validation network-layer protocol demultiplexing IP packet validation forwarding decision longest prefix match data link frame construction MPLS: Push, Swap& Pop Push Swap Swap Pop Data Data Data Data Label(1) Label(2) Label(3) Data CE iler (PE) LSR (P) LSR (P) eler (PE) CE LSP (Label Switched Path)

18 MPLS IP control plane FEC (Forwarding Equivalence Class) IP routing FEC = IP Prefix; FEC lookup done at each hop MPLS other FEC criteria possible, FEC lookup only at an iler RIB R-table RIB R-table Control Plane LIB Label Info LIB Data Plane LFIB LFIB FIB FIB iler LSR RIB (Routing Information Base) --- LIB (Label IB) FIB (Forwarding Information Base) --- LFIB (Label FIB) IP MPLS

19 MPLS labels and tunnels Data (MPLS Payload) MPLS Header(top) MPLS Header(bottom) L2 Header Labelstack(FrameMode label between L2 and the encapsulated data) iler Voice Data 20 bit 3 bit 1 bit 8 bit Label EXP S TTL LSR Transport Traffic Class(RFC 5462) Important only for external(top) label Pipe Mode and Uniform Mode Tunel A Tunel B eler Voice Data LSR handles only transport tunnels

20 MPLS tunel set-up Upstream Downstream Subnet Prefix A R1 (iler) R2 (LSR) R3 (eler) Subnet Prefix B Subnet Prefix C OSPF OSPF SRC FEC NextHop A - B R2 C R2 DST LDP LDP FEC NextHop FEC NextHop A R1 LFIB A R2 LFIB LFIB FEC i-lbl e-lbl NH B - FEC i-lbl e-lbl NH B R2 FEC i-lbl e-lbl NH C R2 C R3 C R3 C - C MPLS Tunnel MPLS tunnel setup requires: FEC (IP prefixes) reachability = OSPF distribution of label mappings between routers label distribution protocols LDP (Label Distribution Protocol) Downstream Unsolicited RSVP-TE (Resource Reservation Protocol) Downstream on Demand R1 Request(FEC C) R3 Response(FEC C, 100) R2 Response(FEC C, 200)

21 MPLS label distribution protocols LDP (Label Distribution Protocol) tunnels built based on IGP (full-mesh) simple configuration automatic creation of tunnels no traffic engineering convergence time depends on IGP label distribution in downstream unsolicited approach RSVP TE (Resource Reservation Protocol with Traffic Engineering) tunnels can be defined administratively ( outside IGP paths) additional constraints (administratove and TErelated) for advanced path calculation bandwidth reservation for LSP (CAC) advanced LSP protection against failures label distribution in downstream on demand approach

22 MPLS the RSVP protocol DST R4 address can contain Explicit Hops Path State Block HOP -IP address from PATH msg Reservation State Block LABELS (Ingress, Egress) R1 (iler) R2 (LSR) R3 (LSR) R4 (eler) PSB PATH (Lbl Rqst) PSB PATH (Lbl Rqst) PSB PATH (Lbl Rqst) PSB SRC RSB RESV (Lbl 300) RSB RESV (Lbl 200) RSB RESV (Lbl 100) RSB DST Tunel MPLS RFC 3209 RSVP as LDP RSVP TE: LSP definition path calculation outside IGP metrics ( link colors, bandwidth etc.) tunel protection (Secondary Paths, Fast Reroute) resource reservation (CAC) LSP = MPLS tunnel, can be composed of many paths (LSP- Paths). one primary path and seven secondary only one active at a time other RSVP messages: PATH Tear: (downstream), RESV Tear: (upstream) PATH Error, RESV Error: Hello (RSVP heartbeat) Summary Refresh (for less signalling)

23 MPLS Traffic Engineering path calculation Strict LSP manual configuration at source router high signalling overhead APC (Advanced Path Calculation) CSPF (Constrained Shortest Path First) additional criteria bandwidth reservation state administrative criteria (link colors) hop limit TE metric Explicit route ( strict hops, loose hops ) Shared Link Groups reservations are made in the Control Plane actual bandwidth usage in the Data Plane is not considered requires relevant QoS solutions in the Data Plane resiliency allows automated creation of backup paths and detours (Fast Reroute) that are disjoint with the primary path

24 MPLS TE variables OSPF TE the need for additional constraints and link state data has to be reflected in routing protocol OSPF-TE (OSPF Traffic Engineering) RFC 2370: The OSPF Opaque LSA Option Opaque LSAs deneric mechanism for OSPF extensions routers create additional database TED (Traffic Engineering Database) for storing additional link attributes distributed by Opaque LSAs (Type 10) Opaque LSA Flooding activated when: link state (up/down), link configuration of bandwidth reservation state changes periodically (as in IGP) Opaque LSA Type 10 contains Link TLV object, used to advertise information about links handled by RSVP-enabled routers: link type, link ID IP addresses of interfaces on both sides of the link TE metrics maximum bandwidth maximum reservable bandwidth (per LSP priority) unreserved bandwidth (100 = 100%, overbooking possible) administrative group Shared Risk Link Group (SRLG)

25 MPLS CSPF algorithm Signalling information about the route is conveyed in the RSVP PATH message in an ERO (Explicit Route Object) ERO is updated in each intermediate router Bandwidth reservation CSPF algorithm calculates a path with the required amount of unreserved bandwidth using data from TED database at source router downstream: reservation request is signaled in RSVP PATH message each router checks bandwidth availability on outgoing link (CAC) upstream: bandwidth is reserved in each router on path (RSVP RESV message) Unreserved Bandwidth updated and advertised Least-Fill Bandwidth Reservation rule if CSPF has found multiple paths with the same metric relevant QOS policies in the Data Plane are required

26 MPLS priorities and preemption LSP Soft Preemption higher priority LSPs can preempt lower priority paths priorities work in conjunction with knowledge of the Unreserved Bandwidth parameter current values are advertised by OSPF TE for each priority level Setup and Hold priorities (0 to 7, lower value = higher priority) LSP A can preempt LSP B if Setup Priority(A) < Hold Priority(B) LSP priorities are signaled in RSVP PATH message, in SESSION_ATTRIBUTE object RSVP Preemption-Timer & LSP Retry-Timer preemption by MBB (Make Before Break) CSPF tries to find another route for preempted LSP periodically (Retry-Timer) preemption (status = down) after time defined in Preemption-Timer (unless a new route was found earlier)

27 MPLS DiffServ Aware Traffic Engineering Class Type FC # FC High Priority 7 NC (Network Control) Signaling 6 H1 (High-1) Signaling or RT traffic 5 EF (Expedited Forwarding) 4 H2 (High-2) RT services Assured 3 L1 (Low-1) Low loss traffic, network management 2 AF (Assured Forwarding) Low loss traffic Best Effort 1 L2 (Low-2) Best effort data 0 BE (Best Effort) transfer DiffServ-TE (RFC 3564) provides bandwidth reservations for traffic classes provides interworking between FC (Forwarding Class) QOS and MPLS TE allows forwarding packets from different service classes to specific LSPs based on traffic class Prt LSP may be assigned a traffic class (depending on service type) Using RSVP-TE the bandwidth can be reserved based on the TE-Class a combination of traffic class and LSP priority CT

28 DiffServ TE link bandwidth allocation Max. Allocation Model each CT (traffic class) is assigned some % of link bandwidth (Bandwidth Constraint) no sharing between classes clear, easy to manage Russian Doll Model bandwidth can be shared between CTs (traffic classes) lower class can use bandwidth unused by higher classes needs priority-based preemption better use of bandwidth but more complex example TE-Class 0 (CT1, Priority 0), 20% of link bandwidth VoIP TE-Class 1 (CT0, Priority 1), 20% of link bandwidth dane data cannot preempt voice voice LSPs cannot preempt other voice LSPs (same for data)

29 Maximum Allocation Model LSP2 200M Data no more Data LSPs possible 1G links -> UnreservedBW(CT0) = 200M UnreservedBW(CT1) = 200M 2 R1 (iler) R5 (eler) no sharing, no preemption LSP1 200M Data R1 (iler) R5 (eler) bandwidth reserved for CT1 cannot be used on shorter route... 1 LSP2 200M Data R1 (iler) R5 (eler) LSP1 200M Data 3 LSP created on shortest path LSP1 200M Data LSP3 200M Voice but can be used for CT1

30 Russian Doll Model 400M free for Data LSPs 1G links -> UnreservedBW(CT0) = 400M UnreservedBW(CT1) = 200M 2 R1 (iler) R5 (eler) sharing and preemption LSP1 200M Data LSP2 200M Data R1 (iler) R5 (eler) LSP2 used CT1 reserved bandwidth 1 LSP2 200M Data R1 (iler) R5 (eler) LSP1 200M Data 3 LSP is created on shortest path LSP1 200M Data LSP3 200M Voice LSP3 preempts LSP2

31 MPLS failure resiliency factors influencing quality of protection failure detection speed OSPF Hello (30 s) RSVP Hello (9 s) Bidirectional Forwarding Detection (<1s)» IP level heartbeat speed of failure advertising service restoration time (switchover speed) PATH ERR PATH ERR Fast Reroute local switchover (<50 ms) PATH ERR conveys only the information about failure detour paths are calculated automatically (CSPF) protects against node and link failures protection types: One-to-One Backup (Detour) Facility Backup (Backup Tunnel) Secondary LSP switchover at source router switchover time depends on PATH ERR message delivery to source router max. 7 standby (Secondary) paths Hot Standby Non-Standby

32 MPLS Point-to-Multipoint LSPs Point-to-Multipoint LSPs MPLS LSP with multiple endpoints PE1 receives IP video stream and encapsulates it into an uni-directional P2MP LSP P routers are responsible for stream replication Advanced MPLS features can be used QOS/TE aware routing RSVP CAC and bandwidth reservation path resiliency control over stream receivers Additional protection level possible two copies of the stream over disjoint LSP P2MP paths in video transport, target Head-End may perform stream conditioning Regional hub B CATV PE3 MPLS;replication Regional hub A i-lbl e-lbl NH TV PE2 Regional hub C Broadband streaming PE4 P P2 LSP A PE2 PE3 PE4 LSP B PE1 Content Provider

33 MPLS P2MP LSP RSVP signalling P2MP LSP a set of P2P LSPs (sub-lsps) each sub-lsp is set up using separate PATH and RESV (ERO objects included in PATH messages targeted to different endpoints are different) PATH and RESV contain new object: Session Object routers have to know the binding between sub-lsps and P2MP path required for proper replication in the data plane LSP tree can be calculated by the source router or offline can be any tree (built under any criteria) flexible solution PE4 P2 PE1 RESV L1 PE5 i-lbl e-lbl NH L1 L3 P3 RESV L2 PE2 L1 PATH ERO(P1, P3, PE3) PATH ERO(P1, PE2) P1 PE2 RESV L2 RESV L3 Common SessionObject P3 RESV L4 PE3 PATH RESV