Ethernet, VLAN, Ethernet Carrier Grade Dr. Rami Langar LIP6/PHARE UPMC - University of Paris 6 Rami.langar@lip6.fr www-phare.lip6.fr/~langar RTEL 1
Point-to-Point vs. Broadcast Media Point-to-point PPP for dial-up access Point-to-point link between Ethernet switch and host Broadcast (shared wire or medium) Traditional Ethernet 802.11 wireless LAN RTEL 2
Three Ways to Share the Media Channel partitioning MAC protocols: Divide channel into pieces: share channel efficiently and fairly at high load Inefficient at low load: delay in channel access, 1/N bandwidth allocated even if only 1 active node! Taking turns protocols: Passing a token for the right to transmit Eliminates empty slots without causing collisions Vulnerable to failures (e.g., failed node or lost token) Random access MAC protocols: allow collisions, and then recover Efficient at low load: single node can fully utilize channel High load: collision overhead RTEL 3
Channel Partitioning: TDMA TDMA: time division multiple access Access to channel in "rounds" Each station gets fixed length slot in each round Time-slot length is packet transmission time Unused slots go idle Example: 6-station LAN with slots 1, 3, and 4 RTEL 4
Channel Partitioning: FDMA FDMA: frequency division multiple access Channel spectrum divided into frequency bands Each station assigned fixed frequency band Unused transmission time in bands go idle Example: 6-station LAN with bands 1, 3, and 4 time frequency bands RTEL 5
Taking Turns MAC protocols Polling Master node invites slave nodes to transmit in turn Concerns: Polling overhead Latency Single point of failure (master) Token passing Control token passed from one node to next sequentially Token message Concerns: Token overhead Latency Single point of failure (token) RTEL 6
Random Access Protocols When node has packet to send Transmit at full channel data rate R. No a priori coordination among nodes Two or more transmitting nodes => collision, Random access MAC protocol specifies: How to detect collisions How to recover from collisions Examples ALOHA and Slotted ALOHA CSMA, CSMA/CD, CSMA/CA RTEL 7
Key Ideas of Random Access Carrier Sense (CS) Listen before speaking, and don t interrupt Checking if someone else is already sending data and waiting till the other node is done Collision Detection (CD) If someone else starts talking at the same time, stop Realizing when two nodes are transmitting at once by detecting that the data on the wire is garbled Randomness Don t start talking again right away Waiting for a random time before trying again RTEL 8
Slotted ALOHA Assumptions All frames same size Time divided into equal slots (time to transmit a frame) Nodes start to transmit frames only at start of slots Nodes are synchronized If two or more nodes transmit, all nodes detect collision Operation When node obtains fresh frame, transmits in next slot No collision: node can send new frame in next slot Collision: node retransmits frame in each subsequent slot with probability p until success RTEL 9
Slotted ALOHA Pros Single active node can continuously transmit at full rate of channel Highly decentralized: only slots in nodes need to be in sync Simple Cons Collisions, wasting slots Idle slots Nodes may be able to detect collision in less than time to transmit packet Clock synchronization RTEL 10
CSMA (Carrier Sense Multiple Access) Collisions hurt the efficiency of ALOHA protocol At best, channel is useful 37% of the time CSMA: listen before transmit If channel sensed idle: transmit entire frame If channel sensed busy, defer transmission Human analogy: don t interrupt others! RTEL 11
CSMA Collisions Collisions can still occur: propagation delay means two nodes may not hear each other s transmission Collision: entire packet transmission time wasted RTEL 12
CSMA/CD (Collision Detection) CSMA/CD: carrier sensing, deferral as in CSMA Collisions detected within short time Colliding transmissions aborted, reducing wastage Collision detection Easy in wired LANs: measure signal strengths, compare transmitted, received signals Difficult in wireless LANs: receiver shut off while transmitting Human analogy: the polite conversationalist RTEL 13
CSMA/CD Collision Detection RTEL 14
Ethernet Dominant wired LAN technology: First widely used LAN technology Simpler, cheaper than token LANs and ATM Kept up with speed race: 10 Mbps 10 Gbps Metcalfe s Ethernet sketch RTEL 15
Ethernet Uses CSMA/CD Carrier Sense: wait for link to be idle Channel idle: start transmitting Channel busy: wait until idle Collision Detection: listen while transmitting No collision: transmission is complete Collision: abort transmission, and send jam signal Random access: exponential back-off After collision, wait a random time before trying again After m th collision, choose K randomly from {0,, 2 n -1} ; n = min (m, 10) Wait for K*512 bit times before trying again (for Ethernet 10Mb/s, 512 bit times = 51.2 µs) The more collisions occured, the more the station may have to wait before attempting retransmission After 16 attempts, the station gives up its transmission RTEL 16
Shared Ethernet and max distance In 10Mbps Ethernet networks, the minimal Ethernet frame length is 64 bytes Why such a limitation? According to the CSMA/CD method, a station must listen to the medium while transmitting in order to detect collisions The frame has to be long enough With a fixed minimal frame length, the maximum distance between 2 stations on a shared network can be calculated. RTEL 17
Maximum Length in Ethernet A! B! latency d! If A emits a minimal frame, the emission lasts 51,2 µs In the worst case scenario B starts emitting just before receiving the first bit emitted by A A will know that a collision occured after a complete round trip between A and B 2 * t prop, with t prop = D AB / Speed Prop After a round trip time, A must still be emitting t trans 2*t prop With Speed Prop = 200 000 km/s, D AB 5,12 km Imposes restrictions on Ethernet length M/D >= 2*L/V M: Minimum frame size : 64 bytes = 512 bits D: Transmission data rate L: Length of the network V: Signal propagation speed along the cable RTEL 18
Ethernet frame Bytes 6 6 2 46-1500 4 DA SA L/ T DATA FCS Bits 24 24 OUI Device Id OUI: Organization Unique Id RTEL 19
Connecting L2&L3 : ARP The Address Resolution Protocol (ARP) -- the mechanisms for translating IP addresses to physical addresses and hide these physical addresses from the upper layers. How it works A requester takes an IP address and searches for a corresponding physical address in a mapping table. Address found, it returns the 48 bit address, such as a workstation or server on a LAN. Address not found, à ARP module sends a broadcast -- ARP request containing the IP address-- onto the network. The machine recognizing its IP returns an ARP reply to the inquiring host. The inquiring host places this address into the ARP cache. RTEL 20
Hubs: Physical-Layer Repeaters Hubs are physical-layer repeaters Bits coming from one link go out all other links At the same rate, with no frame buffering No CSMA/CD at hub: adapters detect collisions twisted pair hub RTEL 21
Interconnecting with Hubs Backbone hub interconnects LAN segments All packets seen everywhere, forming one large collision domain Can t interconnect Ethernets of different speeds hub hub hub hub RTEL 22
Bridge Link layer device Examines frame header and selectively forwards frame based on MAC dest address When frame is to be forwarded on segment, uses CSMA/CD to access segment Transparent Hosts are unaware of presence of bridges RTEL 23
Bridge: Traffic isolation Bridge breaks subnet into LAN segments Bridge filters packets Same-LAN-segment frames not usually forwarded onto other LAN segments Segments become separate collision domains Bridge collision domain hub hub hub collision domain collision domain RTEL 24
From a shared solution to a switched solution switch Ethernet switching: LAN segments contain only one station each, interconnecting with a switch Switches are Plug-and-play, self-learning (do not need to be configured) Speed: 10, 100 Mbps, 1, 10 and 100 Gbps No collision, no limitation of distance RTEL 25
Switched Ethernet Bridges and Ethernet switch are intelligent equipments Packet switching based on MAC destination address Allows VLAN management Each side of a bridge, or each port of a switch is an independent Ethernet network No distance limitation in switched networks Higher throughput (until 40 Gbps) Main difficulty : Each switch has to know the MAC address of every station on the whole network Frames may pile up in switches : flow control is necessary Pause frame RTEL 26
Ethernet switching 2 generations The old one: we use the Ethernet address as a label. The signaling technique is implicit. When an Ethernet switch receive a packet, it verifies that the switching table possesses a line with this label and the output interface. If there is no output line defining how to go to the destination MAC @, the switch sends a multicast packet (ARP packet) to find where is the destination node and update its switching table. A B 1 2 3 4 RTEL 27
Ethernet switching The new generation: A pure switching technique with a field transporting the label. The MPLS label is used as the reference for switching: this is the MPLS Ethernet forwarding. IP signaling What happens in case of link failure? With MAC address switching : sending of a multicast packet to find the station or a switch that knows where to forward the packet With MPLS Ethernet forwarding : Back up paths can be used RTEL 28
Virtual Local Area Network Network traffic consists of a high percentage of broadcasts and multicasts Reduce the need to send such traffic to unnecessary destinations Use VLAN VLAN is a broadcast domain Group of hosts (ports) on the switch with a common set of requirements Group of hosts communicate as if they were attached to the same wire VLAN has the same attributes as a physical LAN VLAN allows grouping to the end stations, services and devices, which do not need to locate on the same LAN segment Traffic can be switched between VLANs with a router RTEL 29
VLAN Operations VLAN has a switched network that is logically segmented Each switch port can be assigned to a VLAN Ports assigned to the same VLAN share broadcasts. Ports that do not belong to that VLAN do not share these broadcasts This improves network performance because unnecessary broadcasts are reduced RTEL 30
Difference of VLAN and LAN RTEL 31
How does it works? Switch receives data from a workstation, it tags the data with a VLAN identifier Tagging can be based on : The port from which it came (Layer 1 VLAN) The source Media Access Control (MAC) field (Layer 2 VLAN) The source network address (Layer 3 VLAN) Or some other field or combination of fields RTEL 32
Virtual networks (physical ports) VLAN 1: #1,3 STATION A STATION B STATION C VLAN 2: # 2,4,5 5 1 2 3 Switch A 4 STATION D STATION F RTEL 33
Virtual networks (MAC addresses) STATION A STATION B STATION C VLAN 1 Switch A STATION D Switch B STATION F3 VLAN 2 STATION F1 STATION F2 RTEL 34
VLAN Tagging First switch adds tag containing VLAN id to all incoming packets Intermediate switches do not recompute the VLAN id Last switch removes tags from all outgoing packets Tag is not swapped at every hop like VC Id or labels RTEL 35
Ethernet Frame for VLAN (IEEE 802.1Q) Octets CRC Octets Tag Protocol ID (TPID) : 0x8100 to identify the frame as a VLAN tagged frame. This value indicates to a L2 device that the frame has an 802.1Q tag. 3 priority bits 8 levels of priority from 0 (lowest) to 7 (highest) The priority bits are the reason why 802.1Q is often referred to as 802.1p/Q. CFI (Canonical Format Indicator): Bit order of address info in Token Ring / FDDI frames. Always set to zero for Ethernet switches. VLAN ID identify the VLAN (12 bits => 4094 different VLANs): A value of 0 means that the frame does not belong to any VLAN; in this case the 802.1Q tag specifies only a priority and is referred to as a priority tag. bits RTEL 36
Ethernet Carrier Grade Characteristics of Carrier Grade: Scalability: billons of users Availability Five«9» = 99,999 pourcentage of availability Time before re-establising a session SONET/SDH : 50 ms n:m:l technology (n paths backuped by m path, backuped by l paths) «Hard» QoS: Strong guaranty on some services How to add these guaranties to get a carrier grade Ethernet? RTEL 37
Reference Provider Transport Architecture PE Provider Network PE RB CE CE CE Customer Network CE RB Customer Network Regular bridges PE: Provider Edge CE: Customer Edge
Ethernet Carrier Grade Three solutions: MPLS Ethernet forwarding MEF (Metro Ethernet Forum) Ethernet GVLAN (Generalized VLAN) A VLAN may correspond to a path establishment if point to point VLAN may allow QoS with traffic monitoring, but the number of possible VLAN is too small (4094) for scalability RTEL 39
1 st solution : MPLS Ethernet forwarding IP signaling to open a path (LSP) Label all along the path (MPLS Label or Shim Label) Ethernet switching along the path RTEL 40
2nd solution : Metro Ethernet Forum (MEF) MEF (founded in 2001), international industry consortium, dedicated to worldwide adoption of Carrier Ethernet networks and services. Composed of leading service providers, network equipment vendors, and other networking compagnies. It had 160 members as of Feb. 2010. MEF : Metro Ethernet Forum RTEL 41
3rd solution: Ethernet Carrier Grade Q-tag (defined in IEEE 802.1Q) allowing to define VLAN Management and performance are improved Three solutions for this approach: IEEE 802.1ad : known as Q-in-Q, stacked VLANs or Provider Bridging, that extend the first definition of VLAN. Creation of a VLAN field for the service provider Service provider can only use 4094 VLANs, still insufficient IEEE 802.1ah : known as MAC-in-MAC, or Provider Backbone Bridge Client MAC is encapsulated in the service provider MAC address The operator only needs to know its own MAC to switch the frame Provider Backbone Transport (PBT) : Creation of a MPLS tunnel MPLS references corresponds to the network ends MPLS allows much larger VLAN differentiation: add a 48 bit reference => total 24 + 48 = 72 bits VLAN identifier. RTEL 42
Q-in-Q (802.1ad) DA SA DA SA DA SA Payload 802.1 VID Payload S-VID C-VID 802.1q SA: Source Address DA: Destination address VID: VLAN ID (802.1q Tag) S-VID: Service VID (802.1q Outer Tag or Metro Tag) C-VID: Customer VID (802.1q Inner Tag) Payload 802.1ad RTEL 43
3rd solution: Ethernet Carrier Grade Q-tag (defined in IEEE 802.1Q) allowing to define VLAN Management and performance are improved Three solutions for this approach: IEEE 802.1ad : known as Q-in-Q, stacked VLANs or Provider Bridging, that extend the first definition of VLAN. Creation of a VLAN field for the service provider Service provider can only use 4094 VLANs, still insufficient IEEE 802.1ah : known as MAC-in-MAC, or Provider Backbone Bridge Client MAC is encapsulated in the service provider MAC address The operator only needs to know its own MAC to switch the frame Provider Backbone Transport (PBT) : Creation of a MPLS tunnel MPLS references corresponds to the network ends MPLS allows much larger VLAN differentiation: add a 48 bit reference => total 24 + 48 = 72 bits VLAN identifier. RTEL 44
MAC-in-MAC (802.1ah) DA SA Payload 802.1 SA: Source Address DA: Destination address VID: VLAN ID S-VID: Service VID C-VID: Customer VID B-SA: Backbone SA B-DA: Backbone DA B-VID: Backbone VID I-SID: Service Instance ID DA SA VID Payload 802.1q DA SA S-VID C-VID Payload 802.1ad B-DA B-SA B-VID I-SID DA SA S-VID C-VID Payload 802.1ah RTEL 45
3rd solution: Ethernet Carrier Grade Q-tag (defined in IEEE 802.1Q) allowing to define VLAN Management and performance are improved Three solutions for this approach: IEEE 802.1ad : known as Q-in-Q, stacked VLANs or Provider Bridging, that extend the first definition of VLAN. Creation of a VLAN field for the service provider Service provider can only use 4094 VLANs, still insufficient IEEE 802.1ah : known as MAC-in-MAC, or Provider Backbone Bridge Client MAC is encapsulated in the service provider MAC address The operator only needs to know its own MAC to switch the frame Provider Backbone Transport (PBT) : Creation of a MPLS tunnel MPLS references corresponds to the network ends MPLS allows much larger VLAN differentiation: add a 48 bit reference => total 24 + 48 = 72 bits VLAN identifier. RTEL 46
PBT DA SA Payload 802.1 SA: Source Address DA: Destination address VID: VLAN ID S-VID: Service VID C-VID: Customer VID B-SA: Backbone SA B-DA: Backbone DA B-VID: Backbone VID MPLS-SL: Service Label DA SA VID Payload 802.1q DA SA S-VID C-VID Payload 802.1ad B-DA B-SA B-VID MPLS-SL DA SA S-VID C-VID Payload PBT RTEL 47