Flashcards
SCC 203 (1)
Cards in this deck (200)
- 1What is 203A meme
- 2Transmission methods of Telegraphy and Telephony.Telegraphy - Message Switching
Telephony - Circuit Switching - 3List the Order of the IOS-OSI Models layersApplication
Presentation
Session
Transport
Network
Data Link
Physical - 4Typically, what layers of the ISO-OSI do network routers have?Network
Data Link
Physical - 5Briefly, describe what the Physical Layer abstracts out.Concerned with the transmission and reception of the unstructured raw bit stream over a physical medium.
- 6Briefly, describe what the Data Link layer does.Transfers Datagrams from one node to physical adjacent over a link via Frames.
- 7List the 6 Data Link layer services.Framing
Link Access
Flow Control
Error Detection
Error Correction
Half-duplex and Full-duplex - 8Describe Data Link layer services: Framing, Link Access.Framing: Encapsulate datagram into frames adding Header and Trailer.
Link Access: "MAC" address used in frame headers to identify the source, destination. - 9Describe Data Link layer services: Flow Control, Error Detection.Flow Control: Pacing between adjacent sending and receiving nodes.
Error Detection: Errors caused by signal attenuation noise. The receiver detects the presence of errors. Signals tetransmission or drops frame - 10Describe Data Link layer services: Error Correction, Half-duplex and Full-duplexError Correction: Receiver identifies and correct bit error(s) without resorting to retransmission.
Half-duplex and Full-duplex:
With half-duplex nodes at both ends of the link can transmit but not at same time. - 11Where is the Data Link Layer Implemented?Network Interface Card or on a Chip
- 12When NIC adaptors communicate between each other what does the Sending side and Receiving side do (Respectivly, only in Data Link layer)Sending Side:
Encapsulated Datagram in Frame,
Adds Error Checking bits.
Receiving Side:
Looks For Errors, flow control etc,
Expracts datagram passes to upper layer. - 13What are Three Functions of Data Link layer protocols- Identifies the start and end of a message
(Message Delineation or Framing)
- Detects and Corrects transmission error (Error control)
- Controls when computers transmit (Media Access Control) - 14Why is Medium Access Control needed?Shared Media / Broadcast connections that are typically used for LANs MUST be able to avoid or detect contention
- 15What is Contention?When two or more devices transmit at the same time over same cable / radio channel
- 16Name Two Link TypesPoint-To-Point: Dial-up Access, Point to point link between Ethernet switch and host.
Broadcast: (Shared Wire or Medium)
802.11 Wireless LAN, Upstream HFC - 17What are two Sub-Layers contained within the Data Link layer? (IEEE-802 - LAN Standards)Media Access Control - (MAC)
Logical Link Control - (LLC) - 18Three Broadcast Classes of TaxonomyChannel Partitioning
Random access
Taking Turns - 19Types of Channel ParitioningTime Division Multiplexing (TDM)
Frequency Division Multiplexing (FDM) - 20What do we mean by Time Division MultiplexingAccess channels in "rounds"
Each station gets fixed length slot
Unused slot = idle - 21What do we mean by Frequency Division MultiplexingDivided into frequency bands,
Each station assigned fixed frequency bands,
Unused transmission bands go idle. - 22Random Access Protocols specify...How to detect collisions,
Recover from collisions, - 23Examples of Random Access MAC ProtocolsSlotted ALOHA,
ALOHA,
CSMA, CSMA/CD, CSMA/CA - 24What are "Protocols"Specification of all possible events and actions in a systems.
Expected and required results to every action/event.
The exact detail of every message sent. - 25Two Types of AddressingHierarchical
Flat - 26What are the 3 Phases of Circuit Switched SystemsCall setup
Communicate
Tear down - 27What is Congestion?Lots of data being transmitted over a link/network at a given time at max capacity
- 28How can we counter congestion?Using queues/buffers
- 29Idea of Message/Packet SwitchingEach transmission independent packet of data.
Packet acts like an envelope with Address and Payload.
Not a fixed route to Destination. - 30Idea of Circuit SwitchingA dedicated line between sender and receiver.
- 31Characteristics of Correction Oriented Systems- Overhead of connection setup/teardown
- Clear relationship between transmissions
- May Not be possible to continue on fail
- Resources tied up if connection idle
- Bandwidth partitioned by Connecion - 32How are Resources Shared and Used in Packet SwitchingUsers Share capacity:
Packets used full link bandwidth for short period.
Resources only used as when required. - 33What is Statisitcal Multiplexing?Allowing resource demand to exceed supply.
Queue packets until avaliable link capacity.
Congestion can lead to delay or packet loss. - 34Two Modes of Packet SwitchingDatagram (Connection-less):
Applications aware of packet structure
Virtual Circuit (non-physical connection):
Applications unaware of packet structure - 35What do Packet Switching Modes determineStyle of Communication.
How addressing information used for path selection. ( A -> B). - 36Parts and Features of a Datagram- Each packet considered as separate message
- Datagrams like postcards - every short message fully addressed.
- Header hold address
Payload holds application data. - 37What are Virtual CircuitsEmulation of connection oriented system over a connectionless (packet) network.
A phone like connection set-up phase.
Each Packet contains connection tag. - 38Two levels of viewsApplication View (Service provided to apps):
- Datagrams - Connection- less
- Virtual Circuits - Connection Oriented
Physical view (How under net operates):
- Packet Switching - Connection-less
- Circuit switching - Connection Oriented - 39Store and Forward Networks procedureReceive packet
Read destination or ID
Find next step in path
Perform housekeeping
Send packet forward destination - 40Why are Store and Forward networks names as such?The whole packet is Received into local storage, then forwarded.
- 41What Delays can Form in Packet switching?Transmission Delay
- The length of packet/bits per second.
Propagation Delay
- Length of Link/propagation speed in material (some % of speed of light) - 42What do communication protocols specify?- How and when data is transmitted and received
- Allowed Messages and expected replies
- Ordering and timing of transmissions
- Exact format of transmitted data - 43Where can Complexity take placeDistributed Systems:
Dealing with many communicating systems - 44What must communication protocols deal with complexity in Distributed Systems?- Sequence actions
- Ensure every case covered unambiguously - 45What is the 5 Layer internet model?Application
Transport
Network
Link
Physical - 46How does Encapsulation work at every layer of ISO-OSI model?Transmitting adds headers at each respective level.
Receiving removes headers at each respective level. - 47When changing networks Different link types will each have their own...- Protocols
- Headers - 48What parts of the OSI-ISO Model are considered the "Upper Layer Architecture"?Application
Presentation
Session - 49What Does the Network Layer Introduce?Network addresses
(Allowing global network to be partitioned into sub-networks) - 50What Mechanisms are there in the Network Layer?Determine path between sender & receiver
Establish and maintain network connection
Forward Traffic - 51Functions of the Transport Layer...Identify applications within machines
- Process and application addressing
Provides Connectionless and connection oriented services - 52What does each layer transmit to the layer below?Transport - Segments
Network - Packets
Data link - Frames
Physical - Symbols - 53Advantages of Slotted ALOHASingle active node can continuously transmit at full rate of channel
Highly decentralized: Only slots in nodes need to be sync
Simple - 54Disadvantages of Slotted ALOHACollisions, wasted slots, Idle Slots
Nodes may be able to detect collision in less time than transmitting packet
Clock Synchronization - 55How does Pure ALOHA differ from Slotted ALOHANo Synchronization
When frame arrives transmits immediately
Can collide with the end or the start of next slot. - 56What does CSMA Stand for?Carrier Sense Multiple Access
- 57What is the human analogy for CSMA?Listen before speaking
if someone else begins speaking at the same time, Stop speaking!
(Shez) - 58What does CSMA/CD Mean?Carrier Sense Multiple Access with Collision Detection
- 59How can collision still occur with CSMA?Propagation delay over a link.
Two nodes may not hear each other's transmissions. - 60What Happnes to colliding transmissions in CSMA/CD?Aborted.
Resulting in reducing channel wastage - 61Why is CSMA/CD easier in Wired LANs than Wireless?Measuring signal strengths is much easier to compare and then receive.
Wireless received signal strength is overwhelmed by local transmission strength - 62What happens when CSMA/CD Detects a collision?Enters Binary Exponential Backoff then
returns to retry transmission. - 63What is Binary Exponential Backoff?Waits for a random amount of time in Microseconds
0 to 51.2 us and increases exponentially per collisions detected. - 64What is Ethernet based on?experimental Pure ALOHA network,
10Mbps CSMA/CD Network. - 65What does IEEE 802.3 Cover?Any 1 - 10Mbps CSMA/CD Network
Different Variations of Ethernet - 66Why is the Star Topology used mainly in modern network structures?If one spoke dies only part of the network is affected
Easy to expand upon (Add more nodes)
Each spoke runs separate ethernet protocol (nodes to not collide with each other - 67What is contained within the Ethernet Frame Structure?- Preamble
- Addresses (Source, Destination MAC addresses)
- Type
- Cyclic Redundancy Check (CRC) - 68What does the Preamble contain and do within an Ethernet Frame?7 Bytes with pattern 0101010 followed by a pattern 10101011,
(0101010-10101011)
Used to Synchronize receiver and sender clock rates. - 69What does the Type Identify in the Ethernet Frame?Higher level protocols (IP, etc..)
- 70What is Meant by Ethernet: Unreliable and ConnectionlessThere is no handshaking between sending and receiving NICs.
Receiving NIC doesn't send ACKs or NACKs - 71What MAC Protocol is used In Ethernet?Unslotted CSMA/CD with Binary backoff
- 724 Features of an Ethernet SwitchLink Layer Device,
Transparent,
Plug-and-Play,
Self Learning. - 73What is meant by Switch FloodingSend out packets on all links if a destination link is unknown.
- 74What are Two "Taking Turn" MAC protocols?Polling
Token Passing - 75what is "Polling" in a MAC protocol context?Each connected node is given chance to transmit by the Master.
Used in Bluetooth. - 76what is "Token Passing" in a MAC protocol context?A token/ticket sent around a ring network sequentially.
When a node has the token it can transmit. - 77What are some Concerns with "Taking turns" MAC protocols?Overhead of many devices,
Latency,
Single point of failure on master or token ring. - 78What layers of the OSI-ISO do switches contain?Data link
Physical - 79What do switches and routers have in common?Store-and-forward.
Forwarding tables. - 80Briefly Describe Port-Based VLANsSwitch with different ports pointing to different network addresses
(1-8 point to net1, 9-15 point to net2) - 81What is Traffic Isolation in Port-Based VLANs?Frames from 1-8 for example can only access ports 1-8
- 82What is Dynamic Membership in Port-Based VLANs?Ports can be dynamically assigned across VLANs
- 83How is forwarding between Port-based VLANs acomplished?Routing or Combined Routers/Switches
- 84What is a Trunk Port?Port that extends a switch. Acts like an index pointer.
Carries frames between VLANs defined over multiple physical switches. - 85What are some Requirements for using a Trunk Port?Must Carry a VLAN ID
(802.1q) - 86IP is Connection-less or Connection OrientedConnection-less
- 87What is HeterogeneityAllowing to work with different network hardware and applications
- 88What is scalability when thinking about internet design decisionsAssuming no centralized system will handle the global load.
Spread load as evenly as possible
Design in performance and design out cost - 89What is the default size of an IPv4 Header?20 Bytes.
Only more when dealing with IP Options - 90IPv4 DS FieldIdentifies priority of traffic.
Voice, File transfer... - 91IPv4 ECN FieldExplicit Congestion Notification,
Flags whether the packet has experienced congestion. - 92How does packet Fragmentation work?Packets too large for the underlying network are split up and duplicate headers are made for each packet.
- 93What is meant by Transparent Fragmentation?Following Devices on a path not aware of fragmentation.
(Packet gets fragmented over path but recombined into original packet after)
-Ties up resources till the last fragment arrives. - 94What is meant by Non-Transparent Fragmentation?Simplifies tasks for routers.
All subsequent devices have to carry and process additional IP Packets.
(Packet gets fragmented but not repackaged after) - 95What happens if a Fragment is lost?Whole packet is retransmitted
- 96How can we avoid Fragmentation?Avoid sending larger packets more likely to be fragmented
- 97What is meant by MTU?Maximum Transfer Unit.
the maximum size of packets that can be sent through a network. - 98How can we Specify that we want a packet Fragmented or Not?Using flags DF (Don't Fragment) and
MF (More Fragments) - 99How is Packet Reconstruction done?Fragment Offset. Tells receiver where to place in reconstituted packet.
8 Bytes - All but the last fragment must be a multiple of 8 Bytes. - 100IPv4 Protocol header/flagIndicates what protocol used to encapsulate following user data.
Determines how payload will be processed
Can be used for De-multiplexing - 101How is the IPv4 Checksum calculated?One's Complement of 16 Bit sum of words.
If received correctly = 0xffff
101011 = 010100 - 102Loopback address127.0.0.1
- 103How is the IPv4 Address split?Network Identifier and Host Identifier
- 104what types of addresses are these 10.0.0.0 172.16.0.0 192.168.0.0Private
(Not to be seen outside the network) - 105255.255.255.255Broadcast address
- 106224.0.0.0 - 239.255.255.255Multicast address
- 107IPv4 Options: (5 Bits) What does each section do?Copy:
0 - copy only to the first fragment
1 - Copy to all fragments
Option Class:
0 - Datagram or net control
2 - Debugging and measurement
- 108What is ICMPInternet Control Message Protocol.
(Encapsulated in IP packet.
Not Considered a Transport Protocol.) - 109What can ICMP be used for?Routers send debugging, error messages to source devices.
(Ping, traceroute, echo are ICMP mechanisms) - 110How can the MTU of a network be found?Using ICMP Protocol to send Packets of large sizes and resizing using Binary chop and the DF flag set. Until success. (Messages replying will determine success)
- 111What addresses do the network and data-link layer deal with?Network - Hierarchical
Data-link - Flat - 112How do we get MAC addresses from IP Addresses?Address Resolution Protocol (ARP)
- 113How do we get IP Addresses from Link-Layer address?Reverse Address Resolution Protocol (RARP)
- 114How does ARP work?1. Broadcasts message "Who has 192.168.8.103".
2. 192.168.8.103 Replies with the devices MAC address - 115When Looking Hop-by-Hop what changes and what stays the same throughout?IP Source and Dest - Constant
MAC Source and Dest - Change - 116How can we avoid using ARP every time we send a message?ARP Table / Cache
Allows hosts to remember recently used mappings.
TTL allows machines changing addresses. - 117What is Gratuitous ARP?Source and Dest set to the same IP.
(ARP Request to itself)
Forces other nodes to update their ARP Caches - 118What Problems does ARP face?Fundamental to communication Systems
Possible/Too easy to spoof or fake ARP replies - 119What does MPLS stand for?Multiprotocol Label Switching
- 120What is the Goal of MPLS?Fast lookup using fixed length identifier.
IP Datagram still keeps it's IP address. - 121What is the Idea of a MPLS Capable Router?Forward packets to outgoing interface based only on label value (don't inspect IP address)
- 122Advantages of MPLS?Flexibility:
- Route flows differently (Traffic Engineering)
- Re-route flows quickly if link fails
(pre-computed backup paths) - 123Whats the difference between how IP and MPLS paths are detemined?IP - Determined by Destination address alone.
MPLS - Based on Source and Destination address - 124What does DHCP stand for?Dynamic Host Configuration Protocol
- 125What does DHCP allow?A computer to join an IP network without having a pre-configured IP address.
- 126What Components are there of DHCP?Server - ( Assigned to specific network,
Configurable params. DHCP Databases, One for manual IP Acquisition, 2nd for pool of available addresses)
Client - (Automatically retrieve DHCP settings) - 127What is a brief overview of the DHCP Procedure?- Client Broadcasts DHCP Discover
Server Answers,
- Server lease address to client
(Client sends Request, Server allocated from address pool, adds lease to database with timeout) - 128Name 8 DHCP Message TypesDISCOVER,
OFFER,
REQUEST,
DECLINE,
ACK,
NACK,
RELEASE,
INFORM - 129Pros of DHCPNo need for manual configuration,
Device can be moved from network to network and obtain a valid IP,
Only Allocated when needed. - 130DHCP LimitationsNeed a server to run the DHCP Server
Used UDP so unreliable
Unauthenticated protocol.
DNS cannot be used for DHCP configured hosts - 131What does NAT stand for?Network Address Translation
- 132What problem does NAT try to solve?Lack of IPv4 Addresses.
- 133What is the Idea of NAT?Local Network uses just ONE IP address as far as the outside world is concerned.
Change internal IP addresses of devices without notifying outside world.
Devices outside network not explicitly addressable - 134When NAT is implemented what does it do to Outgoing datagrams?Replace {source IP address, port#} with {NAT IP address, new port#}
- Nat Contains a translation table for each. - 135Why is NAT controversial?Routers should only process up to layer 3.
Violates end-to-end.
Could be solved by IPv6. - 136What is the Transport Layers main function?Provide data transport between:
Processes
Different End-systems - 137What are the Requirements for the Transport Layer?Reliability
Efficiency
Low cost
Independent ( From lower layers and specific networks) - 138What are two service types of the Transport Later?Connection Oriented,
Connection-less. - 139Where is the transport layer usually contained?Kernel
- 140What is the Transport Layers Data Unit?TPDU/Message
- 1415 Simple Transport Layer primitives.Listen
Connect
SEND
RECEIVE
DISCONNECT - 142What does TSAP Stand for?Transport Service Access Point
Internet: port - 143How does Initial Connection Protocol find TSAP?Waits for connection requests
Creates appropriate service provider process
Transfers connection and desired service waits for further request - 144What are the 5 Transport Layer functions we learned?Multiplexing/Demultiplexing data for multiple applications. - Uses "port" abstraction.
Connection Establishment (E2E connection)
Error Control. End-to-end flow control..
Congestion control. - 145What do processes need to know to communicate?IP Address
Port Number - 146What are features of UDP?Best Effort message delivery over IP
Unreliable
Connectionless
Message-oriented - 147List 4 Characteristics of UDP.No Connection Establishment
No QoS guarantees
No Flow Control
No Error control or retransmission - 148What 5 Sections are there of the UDP Header?Sender and Receiver Port,
Packet Length,
Checksum,
Data. - 149What is a Socket?An Interface that the OS Provides to its networking subsystem
- 150What layer are Sockets used by?Application layer
(Application layer plugs into a socket)
Returned by socket() call - 151How are sockets determined?Host Address + Port Number
external 102.321.31.21, p21 +
192.69.10.2,p1500
(IDK EITHER) - 152Socket functions that pass address from user to OSbind()
connect()
sendto() - 153Socket functions that pass address from OS to useraccept()
recvfrom() - 154How is the internet formed?A set of routers
(no global view) - 155In internet routing there are two distinct parts to process, what are they?Routing
- Application Level
Forwarding
- System level process. according to learned routes. - 156What Layer does IP Forwarding occur?Network layer
(Kernel Level) - 157How can we find the network address a IP resides within?AND'd the IP address with its subnet mask
- 158What is Super netting?Combining adjacent subnet addresses
Mask pushed into network part of address
Lower bits as host - 159What does CIDR Stand for?Classless Inter-Domain Routing
- 160What does CIDR do?Allows registries to allocate on any bit boundary
Introduces the /prefix-length notation - 161How can the /prefix be worked out?First 24 bits + N bits of variable length.
24 + 3 = /27
/27 = 255.255.255.224
/24 = 255.255.255.0 - 162How can we have multiple devices on same link layer?Virtual routing interfaces
- 163What is the default route of a forwarding table?0.0.0.0/24
- 164What does Longest Prefix Match mean?When scanning through routing/forwarding table we select the most detailed route
10.32.5 > 10.32. - 165What is Route Summarisation?Combining two similar routes
194.80.37 + 194.80.37.64 -> A
194.80.37.128 + 194.80.37.192 - > B - 166Going off the Last card's answer What would be the 2 routes left in the forwarding table?194.80.37.0 -> A
194.80.37.128 -> B - 167What do we need to do as we perform Route Summarisation?Change the mask
- 168How can loops be avoided when summarizing routes?Black-hole any traffic sent to a hole in the address range.
MUST NOT return traffic to risk loops - 169What does Route Summarisation allow?Main Route + exceptions
- 170What must Broadcasts do to avoid loops?Never leave the interface which it arrives.
- 171Difference between Direct Delivery and Remote Delivery?Direct - Same link
(Use ARP to obtain MAC address for dest & send over connected link)
Remote - off link
(Find next hop in table using longest prefix match & ARP to obtain MAC address of router. - 172What does RIP stand for?Route Information Protocol
- 173What are the two approaches to configuring routes?Non-Adaptive/Static
Adaptive/Dynamic - 174Local Area Networks use what Kind of Routing protocol?Interior Routing
(Distance Vector,
Link State) - 175Wide Area Network use what kind of Routing protocol?Exterior Routing
(Path vector) - 176What is the Distance Vector?Measures number of hops/distance
- 177What is meant by infinite cost?When a link fails the cost across the link is now infinite.
- 178Why can Infinite cost affect route updates?Route Updates only take in new updates that are less than the current.
Since any number is less than Infinite we ignore. - 179What happens when a link fails?Send Normal periodic updates.
Node adjacent to link error sends update on link failure. - 180if a normal periodic update occurs before a failure is sent. what could happen?A node that was adjacent to the broken link could still believe that the link is still alive.
- 181How can a loop/incorrect link/cost occur in RIP?Normal periodic update occurs first. Old Node belives link that was broken is now through the one that sent the update. However the node that sent the update believed that the broken node was the correct way to get to the now broken node
- 182if a link had cost 15 but all costs in all tables are less than 15, what happens?The link that has 15 cost can't be used.
- 183What will happen to a packet sent to node C if A and B have a loop link to C?packet sent will cycle until its TTL hits 0 and expires.
A -> B -> A -> B .... - 184What is the Bouncing Effect?Constantly looping causes costs to be increased per cycle.
- 185What happens if 2 links fail and cause 2 nodes to become isolated, one link send an update before recognizing a link failure?Look is caused between Two nodes.
No good information in system that can correct errors. - 186What does count to infinity mean?Routing updates increase the cost of links each update. This will go to infinity.
Setting "infinity" as a large number but one we will never see as a cost is a way to stop infinite counting. - 187How often are RIP Updates sent?Around every 30 seconds
- 188What happens when a table entry TTL run out?Node considered unreachable
- 189What are problems with having TTL too long or too short?Too long - Stale entries (Increases chance to see bouncing and count to infinity)
Too short - Kill valid routes due to network problems - 190What tries to prevent bouncing and count to infinity?Triggered Updates
Split Horizon - 191What is Split Horizon?If A has route to D via B
A must not tell B it can reach D - 192What are TRIGGERED Updates?Sending updates when a Link fails or Increased link cost.
- 193What is poisoned reverse?(More aggressive variant of split horizon)
When learning of a failed route, suspend split-horizon rules for that route and advertise poison route. - 194What is a Solicited Update?A Device sends RIP-REQUEST to nodes and gets reply RIP-RESPONSE of nodes known routes.
- 195How can all nodes sending RIP updates every 30 Seconds impact the network?Load on the network with each node sending updates. Good to vary around 15 - 45 seconds.
- 196What does RIPv2 support compared to RIPv1?CIDR
Multicast updates
RIPv1 Compatability - 197What does IGMP stand for?Internet Group Management Protocol
- 198What does IGMP serve?A way for clients can register interests in groups/feeds
- 199Why does RIPv2 use next hop?RIPv2 can announce next hops they believe better / more direct.
- 200How does RIP being simple become a problem?No support for multiple or dynamic metrics
Information
- Description
- Flashcards on SCC 203 (1), created by Alex Stout on 05/05/2017.
- Created by
- Alex Stout · May 2017
- Updated
- May 8, 2017
- Views
- 36
- Rating