Module 4-7 Netacad

Module 5

-          IPV4 Addresses

o   Begin as binary and converted to decimal

-          Binary

o   Each address contains a string of 32 bits

§  Divided into 4 sections called octets

§  Each octet contains 8 bits

·       1 byte

·       4 bits

-          IPV6 Addresses & Hexadecimal

o   8 Hextets

§  Or 8*16 bits

o   Hexadecimal

§  Starts counting after the 10 value, 11 is A, and 12 is B, etc.

§  Each “letter” is 4 bits.

Data Link Layer (Layer 2 of the OSI model)

-          Prepares network data for the physical network.

o   Includes the Network Interface Card (NIC)

-          Accepts Layer 3 Packets (IPv4 or IPv6) and encapsulates them into Layer 2 frames

-          Performs error detection and rejects corrupted frames

Main Task

1.      Adds layer 2 ethernet destination

2.      Adds source NIC information

3.      Converts into a packet supported by the physical layer

Data Link Sublayers

-          IEEE 802 LAN/MAN

-          Logical Link Control (LLC)

o   Talks to networking software at upper layers

o   Talks to device hardware at lower layers

o   Places layer information (for which protocol is used) in frame

o   Allows multiple protocols like IPv4 and IPv6 to use the same interface and media

-          Media Access Control (MAC)

o   Implements this sublayer in hardware

o   Data encapsulation

§  Frame delimiting

·       Identifies fields within a frame

·       Provide synchronization between transmitting and receiving nodes

§  Addressing

·       Source and destination addressing for the Layer 2 frame

§  Error Detection

o   Media access control

§  Allows multiple devices to communicate over a shared medium (half-duplex)

·       Full-duplex do not require access control

o   Data link layer addressing

o   Controls the NIC and other hardware

o   Responsible for sending and receiving data on the LAN/MAN medium

-          Router Functions

-          At each hop along the path, a router performs the following Layer 2 functions:

o   Accepts a frame from a medium

o   De-encapsulates the frame

o   Re-encapsulates the packet into a new frame

o   Forwards the new frame appropriate to the medium of that segment of the physical network

§  LAN packets converted to WAN packets, etc

-          Engineering Organizations

o   IEEE

o   ANSI

o   ITU

o   ISO

-          Network Topologies

o   Hub and Spoke (similar to star)

§ 

§  Single point of failure

o   Mesh

§ 

§  More Expensive

§  High availability

o   Point to Point WAN (PPP)

§  2 nodes placed on the edge of the network

§  Simple, nodes don’t have to make any determination about an incoming frame being for it or not.

-          Legacy LAN Topologies

-          Bus

o   All end systems are chained to each other and terminated in some form on each end. Just like a metal chain.

-          Ring

o   End systems are connected to their respective neighbor forming a ring.

o   Does not have to be terminated

-          Duplexes

o   Half-Duplex

§  Both devices can transmit and receive on the media, but not simultaneously. (restricts)

§  Only allows one device to send or receive at a time on a shared medium (like an ethernet hub).

o   Full-Duplex

§  Both devices transmit and receive on the media at the same time.

-          Access Control Methods

o   Multiaccess Networks

§  Network that can have 2 or more end devices attempting to access the network at the same time.

·       They usually have access rules.

§  Ex: LAN, WLAN.

o   Contention-based access

§  All nodes operating in half-duplex

§  Competing for the use of a medium

§  A process takes place for multiple devices trying to access at once

·       Carrier Sense Multiple Access w/ Collision Detection (CSMA/CD) used on legacy bus-topology Ethernet LANs.

o   PC1 checks if any device is transmitting on the medium. If no signal is detected, it proceeds.

o   Ethernet hub receives and sends the frame. (multiport repeater)

§  If another device wants to transmit, but is currently receiving the mass sent frame, it must wait.

o   The frame has a destination data link address for PC3, so only that device will accept and copy in the entire frame.

o   If a collision is detected (from a higher signal amplitude than normal by the nic), both data would be corrupted and have to be resent.

·       Carrier Sense Multiple Access w/ Collision Avoidance (CSMA/CA) used on wireless LANs.

o   Does not detect collisions but attempts to avoid them by waiting before transmitting.

o   All other devices receive this information and know how long the medium will be unavailable.

o   One the receiver receives the data, it returns an acknowledgement.

o   Do NOT scale well under heavy medium use.

§   

o   Controlled Access

§  Each node has its own time to use the medium

§  Inefficient because devices have to wait their turn

·       Legacy Token Ring

·      

·       Legacy ARCNET

-          Data Link Frame

-          Goals

o   Prepares the encapsulated data (ip packet) for transport by encapsulating it with a header and a trailer to create a frame.

o   Responsible for NIC-to-NIC communication within the same network.

-          Frame Fields

o   Breaks the stream into decipherable groupings

o   Control information inserted in the header

o   Frame fields include the following:

§  Frame start and stop indicator flags – Used to identify the edges of the frame

§  Addressing – Indicates source and destination

§  Type – Identifies the L3 protocol

§  Control – Special control flow services such as QoS.

§  Data – Contains the frame payload? (packet header, segment header, and the actual data)

§  Error Detection – Included after the data to form the trailer.

·       Transmitting node creates a logical summary of the contents of the frame (cyclic redundancy check: CRC)

·       CRC is placed in the frame check sequence (FCS) to represent the contents of the frame.

·       In the ethernet trailer, FCS provides a method for the receiving node to determine whether the frame experienced transmission errors.

-          Layer 2 Addresses

o   Data link layer provides addressing used in transporting the frame.

o   Device addresses at this layer are referred to as physical addresses

§  Contained within the frame header

§  Specifies the frame destination node

§  Physical address stays the same if you move networks

o   At each point along the way to the destination, the packet is encapsulated into a new data link frame.

§  Contains the source data link address of the NIC sender

§  Contains destination data link address of NIC receiver

§  No meaning beyond the local network

·       If you must pass onto another network segment, an intermediary device is necessary.

-          LAN and WAN Frames

o   All OSI Layer 2 (data link) protocols work with IP at OSI Layer 3 (Network)

o   Layer 2 protocol used depends.

§  Determined by technology, size, number of hosts, and services.

o   Each protocol performs media access control

§  Number of different network devices can act as nodes that operate at the data link layer when implementing these protocols.

§  Including NICs, Interfaces on routers, and switches.

o   Examples:

§  Ethernet

§  802.11 Wireless

§  Point-to-Point Protocol (PPP)

§  High-Level Data Link Control (HDLC)

§  Frame Relay

Ethernet Switching – Module 7

-          Ethernet Encapsulation

o   One of 2 LAN technologies being used today (other being WLANs)

o   Ethernet operates in the data link layer (2) and the physical layer (1). It is defined by them and their protocols.

o   Ethernet is a family of networking technologies defined by the IEEE 802.2 & 802.3

-          Ethernet uses the LLC and MAC sublayers of the data link layer (2) to operate.

o   LLC (Logical Link Control) recap

§  An IEEE 802.2 sublayer

§  Communicates between the networking software and the device hardware.

§  Places which network layer protocol is being used into the frame.

§  Allows IPv4 and IPv6 to use the same network interface and media (by recognizing the protocol and encapsulating the frame)

o   MAC (media access control) recap

§  IEEE 802.3, 802.11, or 802.15

§  Implemented in hardware

§  Responsible for encapsulation

·       Structure of the Ethernet frame

·       Ethernet addresses (MAC source and destination)

·       Ethernet error detection (FCS trailer)

§  Provides data link layer addressing

-          Ethernet Frame Fields

o   Minimum size is 64 bytes

§  Any frame less than this is considered a “collision fragment” or “runt frame” and is automatically discarded.

o   Expected max is 1518 bytes

§  Includes all bytes from destination MAC address through the FCS field.

§  Frames larger than 1500 bytes of data are considered “jumbo” or “baby giant frames.”

o   Any frame larger/smaller than max/min are automatically discarded. (likely the result of collisions or others)

-          Preamble and Start Frame Delimiter (SFD)

o   Preamble (7 bytes) and SFD (or Start of Frame) fields are used for synchronization of the receiving/sending devices.

o   Used to get the attention of the receiving nodes. (or tells them to get ready)

-          Destination MAC Address

o   6-byte field

o   Identifier for the intended recipient

o   Address in the frame is compared to the MAC address of the device

o   Can be unicast, multicast, or broadcast address

-          Source MAC address

o   6-byte

o   Identifies the originating NIC or interface

-          Type/Length (aka EtherType, Type, or Length)

o   2-byte

o   Identifies upper layer protocol being encapsulated in the Ethernet frame

o   Common values are hexadecimal for IPv4, IPv6, and ARP

-          Data Field

o   46-1500 bytes

o   Contains the encapsulated data from higher layers (commonly an IPv4 packet)

o   Must all be atleast 64 bytes long, padding is added if not

-          Frame Check Sequence

o   4 bytes

o   Uses cyclic redundancy check (CRC)

o   Sending device includes the results of original CRC in the FCS field

o   Receiving device receives the frame and generates a CRC

o   Basically hashing

-          MAC Addresses Specifics

o   Each vendor must register with the IEEE to obtain an unique 6 hexadecimal code called the OUI (organizationally unique identifier)

o   When a vendor assigns a MAC address to a device, it uses its assigned OUI as the first 6 hexadecimal digits, and a unique value in the last 6 digits.

§  Responsibility of the vendor to ensure that none of its devices be assigned the same MAC address.

§  MAC addresses duplicates still exist

§  Sometimes it’s referred to as BIA (burned in address) because it is encoded into the ROM permanently.

·       However, you can use software to change the MAC address

-          Frame Processing

o   When a device forwards a message to an ethernet network, the ethernet header includes these:

§  Source MAC address

§  Destination MAC address

o   When the frame is forwarded, the frame looks like this

§  Destination address

§  Source address

§  Encapsulated Data

o   All the NICs on the network receive the packet, and if their MAC address match, it accepts it. Otherwise, it does not.

-          Unicast MAC addresses

o   In ethernet, different MAC addresses are used for Layer 2 unicast, broadcast, and multicast.

§  The system uses a destination IP address and destination MAC address in combination to deliver data to a specific host.

§  The process that a source host uses to determine the destination MAC address associated with a specific IPv4 address is known as ARP (address resolution protocol).

·       For an IPv6 address, this is known as ND (Neighbor Discovery)

-          Broadcast MAC address

o   Still has a destination MAC address

o   Flooded out ALL the Ethernet switch ports except the incoming port.

o   NOT forwarded by a router.

o   If the encapsulated data is an IPv4 broadcast packet:

§  The packet contains a destination IPv4 address that has all 1s in the host portion.

§  This means that all hosts on the local network will receive and process the packet.

§  When the IPv4 broadcast packet is encapsulated in the ethernet frame, the destination MAC address is “FF-FF-FF-FF-FF-FF” (48 ones in binary)

-          Multicast MAC address

o   The destination MAC address is 01-00-5E for IPv4

o   The destination MAC address for IPv6 is 33-33

§  There are others for STP and LLDP

o   Flooded out all ethernet ports

o   Not forwarded by a router (unless configured to do so)

o   The multicast group is assigned a multicast group IP address

-          Ethernet Switch

o   Uses only L2 MAC addresses to make forwarding decisions

o   Examines its table to make the decision

§  Most switches keep an entry in the table for 5 minutes

§  If the source MAC address exists the timer is reset

§  If it doesn’t exit, it’s added along with the port number it’s connected to.

·       If the destination MAC address is not in the table, it’s blasted to all the ports.

·       The destination does not get added to the MAC table unless it then sends out a packet as the source MAC.

o   When going to the Internet, the destination MAC is the router

o   When coming from the Internet, the IP address is that of the sender off the network. (contained in the DATA section)

§  The source MAC is the router

§  The destination MAC is the computer.

-          Frame Forwarding Methods (on cisco switches)

o   Store-and-forward switching – Recivies the entire frame and computes the CRC (like a hash to determine the number of bits in the frame) to determine if there’s an error.

§  If the CRC is valid, the switch looks up the destination address, which determines the outgoing interface and forwards it through there.

§  Required for QoS.

o   Cut-through switching – Forwards the frame before it is fully received. The destination address of the frame is read before it is forwarded (performs no error detection).

§  Fast-Forward Variant – Lowest level of latency by immediately forwarding the packet after reading the destination address.

§  Fragment Free Variant – Switch stores the first 64 bytes before forwarding (Most network errors occur in the first 64 bytes). It is like the middle between cut-through switching and store-and-forward switching.

-          Memory Buffering on Switches

-          Used when the destination port is busy and the switch stores the frame until it can be transmitted

o   Port-based memory

§  Frames stored in queues linked to specific incoming/outgoing ports

§  Only transmitted when all frames in the queue have been successfully transmitted

§  Possible for one single frame to delay the transmission of all frames in memory because of a busy destination port, even if the other frames could be transmitted to other open destination ports.

o   Shared memory

§  Deposits all frames into one common memory buffer

§  Amount of buffer memory required by a port dynamically allocated

§  Frames in the buffer are dynamically linked to the destination port, allowing a packet to be received on one port and then transmitted on another port without moving it to another queue

§  Results in the ability to store larger frames with potentially fewer dropped frames, important when there are multiple different port speeds.

-          Duplex

o   Duplex mismatch is one of the most common causes of performance on 10/100 Mbps Ethernet links

o   Occurs when one port operates at half-duplex, and other at full duplex.

o   Causes many collisions.

o   Auto negotiation allows them to automatically select the same, best option.

-          Auto MDIX

o   Switch automatically detects the type of cable attached to the port and configures the interfaces accordingly.

o   You should still regardless use the correct cable.