ch8-9-QoS.txt

Card Set Information

Author:
elizaram
ID:
288384
Filename:
ch8-9-QoS.txt
Updated:
2014-11-06 15:47:56
Tags:
VoIP
Folders:

Description:
VoIP chapters 8, 9, and all QoS sources
Show Answers:

Home > Flashcards > Print Preview

The flashcards below were created by user elizaram on FreezingBlue Flashcards. What would you like to do?


  1. What are the 3 enemies of VoIP traffic?
    • Bandwidth (lack of)
    • Delay
    • Packet loss
  2. What is the max delay for voice and video?
    150ms (one-way)
  3. What is the max jitter for voice and video?
    30ms
  4. What is the max packet loss for voice and video?
    1%
  5. What is the bandwidth required for voice and video?
    Depends on the codec
  6. What are the three QoS mechanisms?
    • Best Effort
    • Integrated Services (Int Serv)
    • Differentiated Services (DiffServ)
  7. Define Best Effort.
    • Default behavior of network
    • Traffic is handled first come, first served
  8. Define IntServ.
    Integrated Services: Bandwidth is reserved on every device between endpoints using RSVP.
  9. IntServ is also known as:
    Hard QoS
  10. What is the advantage and disadvantage of IntServ?
    • Advantage: bandwidth is guaranteed.
    • Disadvantages: Not scalable (network can run out of bandwidth to reserve); every device must be configured for it.
  11. Define DiffServ.
    Every device configured to apply a variety of QoS methods based on traffic classification.
  12. DiffServ is also known as:
    Soft QoS
  13. Which QoS mechanism is the standard?
    DiffServ
  14. Five QoS tools are:
    • Classification & Marking
    • Congestion Management
    • Congestion Avoidance
    • Policing and Shaping
    • Link Efficiency
  15. Define classification and marking:
    Mark a packet to be identified by network devices. First device it reaches IDs and tags it; other devices read the tag.
  16. What are three tools a device can use to identify and classify a packet?
    • Access lists
    • Incoming interfaces
    • Deep packet inspection
  17. What are two ways to mark a packet for QoS?
    • Layer 2 (switches can read it)
    • Layer 3 (routers can read it)
  18. "Congestion management" refers to:
    Queuing strategy (the rules a router should apply for forwarding packets when congestion occurs)
  19. "Congestion avoidance" refers to:
    Dropping less essential packets before congestion occurs
  20. What is another term for congestion avoidance?
    Traffic conditioning
  21. What are policing and shaping, and what is the difference?
    • Both policing and shaping limit the bandwidth used by certain types of traffic.
    • Shaping queues the limited traffic, while policing drops it.
  22. "Link efficiency" refers to:
    Delivering traffic in the most bandwidth-efficient way, e.g. compression
  23. How fast is a "speed challenged" link?
    Slower than T1
  24. What are two challenges of "speed challenged" links?
    • Lack of bandwidth limits amount of data that can be sent
    • Serialization delay causes end-to-end delay
  25. What is serialization delay?
    The time it takes to put a packet on a wire (larger packets take longer)
  26. Name three link efficiency mechanisms.
    • Payload compression
    • Header compression
    • Link fragmentation and interleaving
  27. What is payload compression?
    Compression of the application data in a packet
  28. What is header compression?
    Eliminating redundant fields in the header of repeated packets
  29. What does LFI stand for and what is it?
    Link Fragmentation and Interleaving: reduces serialization delay by cutting large packets into smaller pieces, then sending higher priority traffic between the pieces
  30. Why is link efficiency not recommended on lines faster than T1?
    They all have drawbacks: compression adds delay and processor load, LFI increases the bandwidth of data
  31. Name seven queuing algorithms.
    • FIFO
    • Weighted Fair Queuing
    • Class-Based Weighted Fair Queuing
    • Low Latency Queuing
    • Priority Queuing
    • Round Robin
    • Weighted Round Robin
  32. What is priority queuing?
    Packets are sorted into queues based on priority, and no lower priority packets can be sent while there are still packets in a higher priority queue.
  33. What is the problem with priority queuing?
    Lower priority traffic can be starved out on a congested link.
  34. What does WRR stand for and what is it?
    Weighted Round Robin: Switch-based queuing method with priority queues, where traffic is forwarded from each queue in turn in an amount relative to its priority
  35. What queuing method is supported by Cisco Catalyst switches?
    Weighted Round Robin
  36. What does WFQ stand for and what is it?
    Weighted Fair Queuing: Balances available bandwidth among all senders evenly
  37. What is the default QoS method on Cisco serial interfaces?
    WFQ
  38. What is CBWFQ and what does it stand for?
    Class Based Weighted Fair Queuing: Allows classes of traffic to reserve bandwidth (by absolute amount or percentage) while the rest gets WFQ.
  39. What is LLQ and what does it stand for?
    Low Latency Queuing: Like CBWFQ, but priority traffic gets first shot at bandwidth before any other guarantees are fulfilled
  40. What is another name for LLQ?
    PQ-CBWFQ (Priority Queue - Class Based Weighted Fair Queuing)
  41. What queuing algorithm is preferred for voice?
    LLQ
  42. Of the eight QoS methods (policing, shaping, congestion management, congestion avoidance, classification, marking, fragmentation and interleaving, compression, which ones are applied when traffic enters the router, which are applied when it exits, and which can go either way?
    • Entering router: Classification
    • Either way: Marking, Policing
    • Exiting router: Congestion Management, Congestion Avoidance, Shaping, Compression, Fragmentation/Interleaving
  43. What are four advantages of Cisco AutoQoS?
    • Reduces time of deployment
    • Provides configuration consistency
    • Reduces deployment cost
    • Allows manual tuning
  44. What is a trust boundary?
    A point in the network where you know the QoS markings on the traffic are accurate (not forged)
  45. On which side of the trust boundary do Cisco IP Phones fall?
    Inside
  46. On which side of the trust boundary to PCs fall?
    Outside
  47. What prevents someone from establishing the trust boundary with a Cisco IP Phone, then attaching a different device in its place?
    AutoQoS uses CDP to identify the device and configure QoS
  48. On which interfaces should AutoQoS be applied?
    On interfaces where devices need special treatment
  49. What is the AutoQoS command?
    auto qos voip
  50. What are the six CUCM interfaces?
    • Cisco Unified Communications Manager Administration
    • Cisco Unified Serviceability
    • Disaster Recovery System
    • Cisco Unified Operating System Administration
    • Cisco Unified Reporting
    • Command-Line Interface
  51. What are the nine CUCM Adminstration menus?
    • System
    • Call Routing
    • Media Resources
    • Advanced Features
    • Device
    • Application
    • User Management
    • Bulk Administration
    • Help
  52. What are the five Cisco Unified Serviceabiity Administration menus?
    • Alarm
    • Trace
    • Tools
    • SNMP
    • Help
  53. What is the difference between feature and network services?
    Network services are required for the server to function and cannot be deactivated. Feature services are optional and can be enabled or not depending on need.
  54. What is the purpose of the Cisco Unified OS Admin interface?
    Enables an administrator to interact with the underlying Linux-based OS
  55. What does the Disaster Recovery System interface do?
    Provides scheduled backup and restore capabilities
  56. What is Cisco Unified Reporting?
    Gathers log data and simplifies/formats into reports
  57. What are the three privilege levels that can be assigned for a resource?
    • No Access
    • Read
    • Update
  58. How many Standard roles are default in CUCM 8.0?
    39
  59. How many standard groups are default in CUCM 8.0?
    24
  60. What is the difference between Maximum and Minimum effective access privileges?
    Maximum: most permissive wins when privileges conflict. Minimum: most restrictive wins.
  61. What are the six steps for phone registration in CUCM?
    • 1. Power
    • 2. Loads locally stored firmware
    • 3. Learns VLAN via CDP
    • 4. Uses DHCP to learn IP info and TFTP server address
    • 5. Contacts TFTP server and receives configuration file
    • 6. Registers with primary CUCM server listed in config file
  62. What is the name of the config file on the TFTP server?
    SEP.cnf.xml
  63. What are enterprise parameters?
    Cluster-wide default settings
  64. What are service parameters?
    Application settings and behavior
  65. What are four methods for adding phones in CUCM?
    • Autoregistration
    • Manual configuration
    • Bulk Administration Tool (BAT)
    • Auto Register Phone Tool (TAPS)
  66. What are the four required fields (without default values) for manual configuration?
    • MAC Address
    • Device Pool
    • Phone Button Template
    • Device Security Profile
  67. What is the difference between Reset, Restart, and Apply Config?
    • Reset: reboots both phone firmware and config
    • Restart: unregisters and re-registers the phone
    • Apply Config: intelligently chooses Reset or Restart
  68. What are the three components of BAT?
    • Excel template for new data
    • Server-side templates to configure common data
    • Web page interfaces
  69. What is the challenge with using BAT for large deployments?
    The MAC of each phone must be specified, so the user must get the correct device
  70. How does TAPS improve on BAT?
    Users can partially configure their own phones, including autocapture of MAC address
  71. What extra piece of hardware is required by TAPS?
    IP-IVR server
  72. What is the difference between end users and application users?
    End users type their login and password in to access features. Application users an application interacting with the server that sends authentication information inline.
  73. What is the relationship between application users and LDAP?
    There is none; they must be provisioned locally.
  74. What is the Credential Policy?
    Policies surrounding authentication, such as length/complexity of passwords/PINs, attempts before lockout, etc.
  75. What are three methods for adding users in CUCM?
    • Manual
    • Bulk import with BAT
    • LDAP synchronization/optional authentication
  76. What are the two manual, non-default fields required for user entry?
    • User ID
    • Last name
  77. What is the difference between LDAP Synchronization and LDAP Authentication?
    Synchronization maintains some user data in LDAP and replicates it to CUCM as read-only; passwords are maintained in CUCM. LDAP Authentication maintains passwords in LDAP.
  78. What happens if LDAP Authentication is enabled and LDAP fails?
    Only the Application Administrator account defined at install can log on to CUCM.
  79. What is delay?
    The time required for a packet to reach its destination
  80. What is jitter?
    The uneven arrival of packets (variability in delay)
  81. What is packet loss?
    Packets are dropped when a link is congested and a buffer overflows
  82. Between shaping and policing, which is recommended on higher bandwidth interfaces?
    Policing
  83. Which field in the IPv4 header is used to mark packets?
    The ToS Byte
  84. What are the two types of packet marking?
    IP Precedence and DSCP (Differentiated Service Code Point)
  85. How many bits does IP Precedence use?
    3
  86. how many bits does DSCP use?
    6
  87. What is the standardized set of DSCP priority values?
    Per Hop Behaviors (PHB)
  88. What are the four categories of PHBs?
    • Default
    • Expedited Forwarding (EF)
    • Assured Forwarding (AF)
    • Class Selector (CS)
  89. What is RED?
    Random Early Detection: throwing packets out to avoid congestion. Begins at minimum threshold; all entering packets dropped after maximum threshold.
  90. What is the mark probability denominator?
    In RED, the inverse of the MPD is the chance of packets being dropped at maximum threshold.
  91. How does WRED differ from RED?
    Weighted RED: each priority marking gets its own RED profile.
  92. How does CBWRED differ from WRED?
    Class Based WRED: used in combination with CBWFQ, class based configuration of WRED.
  93. What does ECN stand for and what is it?
    Explicit Congestion Notification: uses last two ToS bits to indicate ECT capability and report congestion.
  94. What is CIR?
    Committed Information Rate: the average bandwidth that policing and shaping aim to achieve.
  95. What is Bc?
    Committed burst: the number of bits sent during a timing interval in policing and shaping.
  96. What is Be?
    Excess burst: traffic allowed to exceed the Bc if bits have been banked earlier from unused allowed bandwidth.
  97. What is the smooth/bursty behavior of data, voice, and video?
    • Data: smooth or bursty
    • Voice: smooth
    • Video: bursty
  98. What is the benign or greedy behavior of data, voice, and video?
    • Data: benign or greedy
    • Voice: benign
    • Video: greedy
  99. What is the drop sensitivity of data, voice, and video?
    • Data: drop insensitive
    • Voice and video: drop sensitive
  100. What is the delay sensitivity of data, voice, and video?
    • Data: delay insensitive
    • Voice and video: delay sensitive
  101. What is another term for delay?
    Latency

What would you like to do?

Home > Flashcards > Print Preview