Scribd Logo
Upload

Welcome to Scribd!

  • 15 views

    Sample-Exam DS

    Uploaded by

    Edem Rockson
    The document provides practice problems to test knowledge of distributed systems topics. It includes 10 questions testing knowledge of concepts like clocks and ordering, distributed systems challenges, threads and processes, virtualization, architecture, fault tolerance, scheduling, and scaling. It also includes a case study problem designing a sensor network to monitor volcanic activity.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Sample-Exam DS

    Uploaded by

    Edem Rockson
    15 views3 pages
    The document provides practice problems to test knowledge of distributed systems topics. It includes 10 questions testing knowledge of concepts like clocks and ordering, distributed systems challenges, threads and processes, virtualization, architecture, fault tolerance, scheduling, and scaling. It also includes a case study problem designing a sensor network to monitor volcanic activity.

    Original Title

    sample-exam DS

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document provides practice problems to test knowledge of distributed systems topics. It includes 10 questions testing knowledge of concepts like clocks and ordering, distributed systems challenges, threads and processes, virtualization, architecture, fault tolerance, scheduling, and scaling. It also includes a case study problem designing a sensor network to monitor volcanic activity.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    15 views3 pages

    Sample-Exam DS

    Uploaded by

    Edem Rockson
    The document provides practice problems to test knowledge of distributed systems topics. It includes 10 questions testing knowledge of concepts like clocks and ordering, distributed systems challenges, threads and processes, virtualization, architecture, fault tolerance, scheduling, and scaling. It also includes a case study problem designing a sensor network to monitor volcanic activity.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 3

    CSCI6421: Distributed Systems Practice Problems

    These practice problems are designed to give you a sense of the type of questions that might
    be asked in your exam. However, it should not be considered a complete coverage of topics you
    might be asked about. Also, these questions were designed for a written test; for an online exam
    the format of the exam may need to be adapted (particularly Q10 would need to be modified).

    1. True or False

    _______ Two programs communicating over sockets must be written in the same programing
    language.
    _______ Peer to Peer services are the only type of distributed system where a node is both a
    client and a server.
    _______ Lamport clocks cannot always put all events into a total order, but vector clocks
    always can.
    _______ Virtualization provides strong isolation, but it is generally slower than running code
    natively.
    _______ Containers are a good choice for operating system kernel developers because they
    can be started very quickly, allowing you to debug as you modify the container's OS
    kernel.
    _______ Most cloud applications use UDP to communicate with clients over the Internet.
    _______ A Decentralized Architecture will always provide better reliability compared to a
    centralized one.

    2. Clocks and Ordering

    Annotate the process timeline with Vector Clocks for each event.

    Using your vector clocks, indicate whether an event happens before ( -> ), after ( <- ) or is
    concurrent ( || ).
    b ________ g g ________ i

    j ________ h f ________ j

    1 of 3
    3. Distributed Systems Challenges

    Pick 3 of the distributed systems challenges below and write 3-4 sentences for each one
    explaining a system or technique we learned about that helps resolve them.

    Challenges: heterogeneity, openness, security, fault handling, concurrency, quality of service,


    scalability, transparency

    4. Threads and Processes


    What is a case where it is preferable to use multiple threads instead of multiple processes?
    What is a case where the reverse is true?

    5. Virtualization
    a) Why do virtual machines use technologies like memory virtualization? Why can't VMs be
    given real physical memory addresses to work with directly?

    b) Why are containers "lighter weight" than VMs?

    6. Architecture
    What is an example of a system that follows a layered architecture? Give an example of one
    Distributed Systems Challenge that is made easier by a layered architecture and one that
    becomes harder.

    7. Fault Tolerance

    You are designing a distributed system that uses replication to mask faults for a stateless
    application. If we assume at most f=4 faults can ever happen at one time, then how many
    replicas are needed in each of these cases:

    a) Replicas either respond correctly, do not respond at all, or can return any incorrect value.

    b) Replicas either respond correctly or do not respond at all.

    d) Replicas either respond correctly or return an error code: NULL.

    2 of 3
    8. Scheduling
    Suppose you are designing a control system for an autonomous car. The car has a variety of
    sensors (cameras, range finders, etc), and each new reading generates an analysis task that
    processes the data and potentially adjusts the car's behavior. What kind of scheduling algorithm
    would you use for these analysis tasks? Why?

    9. Scaling
    Suppose you have a stateful service. What techniques could you use to scale it vertically or
    horizontally? (describe each individually)

    10. Case Study


    For this question you will be judged on the completeness, creativity, and clarity of your answer.
    Note that the case study description is intended to be somewhat vague. You should specify
    what assumptions you are making and how that affects your solution.

    You are building a sensor network to study a mountain chain with several active volcanos. The
    network will be composed of two types of nodes: 1) Small nodes that have a radio with a radius
    of about 100 meters and are low power with limited processing and memory capabilities; 2)
    Large nodes that have a radio range radius of 1000 meters and have greater storage and
    processing capabilities than small nodes. The nodes are designed to be dropped out of an
    airplane over a wide area.

    Each of the nodes will gather data that will be analyzed by scientists to detect potential
    eruptions. Both types are capable of gathering the same data from their sensors. A data point
    takes 1KB of memory. The small nodes have 10KB of total memory available and the large
    nodes have 1,000KB.

    The scientists want to be able to answer questions like: what is the average data reading in a
    geographic region covered by several sensors? However, in the future they may devise more
    complicated programs that will run on each sensor, so they want to have a flexible system.

    Draw diagrams and write an explanation to justify your design for this system. Which of
    the distributed systems challenges below does your system help with? How?
    Challenges: heterogeneity, openness, security, fault handling, concurrency, quality of service,
    scalability, transparency

    3 of 3

    You might also like