CS 416 Exam 3

Spring 2014

See the solutions (6 per page).

100 points – 4 points each

For each statement, select the most appropriate answer. I will give you credit for two incorrect answers.

  1. A weakness of using NAND flash memory for use as a file system is:
    (a) Stored data wears out over time, requiring periodic refreshing.
    (b) Its read speeds are unpredictable, making I/O scheduling difficult.
    (c) It has very long seek times compared to those of a disk.
    (d) It supports a limited number of writes per block.
  2. Checkpointing in a log-structured file system such as YAFFS:
    (a) Periodically writes updates of the file systemÕs superblock.
    (b) Writes data blocks and their associated metadata as a single transaction.
    (c) Replicates data in the log for reliability.
    (d) Avoids the need to reconstruct the file system hierarchy by reading all logs.
  3. A loop device:
    (a) Enables the output of one device to be fed to the input of another.
    (b) Echoes back whataver data is sent to it.
    (c) Enables two local processes to communicate.
    (d) Converts a file to a block device.
  4. The Address Resolution Protocol (ARP) is used to:
    (a) Enable a computer to get a unique IP address.
    (b) Find the IP address that corresponds to a systemÕs domain name.
    (c) Resolve the path that a packet must take to reach a given IP address.
    (d) Convert an IP address to the corresponding ethernet address.
  5. The network device driver is responsible for which layer of the OSI stack?
    (a) Data link.
    (b) Network.
    (c) Transport.
    (d) Presentation.
  6. Sockets are said to be compatible with files because:
    (a) They are implemented under the VFS layer to support all file system operations.
    (b) You can use read/write system calls with them.
    (c) They appear in the file system name space.
    (d) File data can be directed to a socket and vice versa.
  7. A socket buffer:
    (a) Is similar to a buffer cache and stores frequently used network messages.
    (b) Is a block of memory in the kernel that holds all the sockets in the operating system.
    (c) Avoids the need to copy data as it moves up or down the network stack .
    (d) Identifies the set of operations that can be performed on each socket.
  8. LinuxÕs NAPI, the "New API," is best described as an enhancement that:
    (a) Disables network device interrupts during times of high network traffic.
    (b) Avoids the need to poll for incoming packets by taking advantage of interrupts.
    (c) Integrates sockets into the file system API, providing a uniform access interface.
    (d) Reduces the need to copy data between network layers.
  9. NFS does not improve performance via:
    (a) Large reads.
    (b) Use of a stateless protocol.
    (c) Read-ahead.
    (d) Client caching.
  10. A callback in AFS:
    (a) Tells a client that another client is modifying a file.
    (b) Tells the server that a client has changed a file.
    (c) Tells a client that a file was modified on the server.
    (d) Tells a process that the requested download of a file is complete.
  11. Which best describes Microsoft's SMB remote file protocol?
    (a) A stateful file system with a download/upload model.
    (b) A stateful file system with a remote access model.
    (c) A stateless file system with a remote access model.
    (d) A stateless file system with a download/upload model.
  12. Privilege separation is when:
    (a) Users on a system get unique user IDs and privilege levels.
    (b) A process cannot create data that processes running at lower privilege levels can access.
    (c) A process is split into components with limited privileges.
    (d) A process cannot interact with any other process or data at another privilege level.
  13. Permissions stored in a file's metadata are an example of:
    (a) An access matrix.
    (b) An access control list.
    (c) A capabilities list.
    (d) A sandbox.
  14. Multi-level security and the Bell-LaPadula model are an example of:
    (a) Mandatory Access Control (MAC).
    (b) Privilege separation.
    (c) Access Matrix.
    (d) Principle of least privilege.
  15. Password files often store hashes of passwords:
    (a) To make it difficult to steal a password even if an intruder gets the password file.
    (b) To make looking up a password very quick: an O(1) operation.
    (c) To be able to detect if somebody tampered with the password.
    (d) To ensure that a given password is associated with only one user.
  16. For Alice to send a signature of a message to Bob, she would:
    (a) Hash the message and encrypt it with Bob's public key.
    (b) Hash the message and encrypt it with Bob's private key.
    (c) Hash the message and encrypt it with her public key.
    (d) Hash the message and encrypt it with her private key.
  17. The Diffie-Hellman algorithm:
    (a) Uses a one-way function to create cryptographic hashes of data.
    (b) Provides users with a pair of keys for secure communication: encrypt data with one key; decrypt with the other.
    (c) Enables a user to authenticate another user.
    (d) Allows two parties to come up with a common key that they can use for encryption.
  18. A hybrid cryptosystem:
    (a) Encrypts a message N times, once for each of N recipients.
    (b) Encrypts a message with a symmetric algorithm and then encrypts the result again with a different key.
    (c) Encrypts a message with a symmetric algorithm and then encrypts the result with a public key algorithm.
    (d) Uses public key cryptography to encrypt a session key that will be used for symmetric cryptography.
  19. The Challenge Handshake Authentication Protocol (CHAP) relies on:
    (a) Encrypting the password with a key sent by the server.
    (b) Encrypting the user's password with a shared key.
    (c) Having the user authenticate the server before sending a password.
    (d) Hashing a structure containing a user's password and a challenge sent by the server.
  20. A digital certificate cannot be modified because:
    (a) It is encrypted by the certificate owner.
    (b) Only the owner knows the corresponding private key to the public key that is present in the certificate.
    (c) It contains a hash that was encrypted by the issuer.
    (d) It is stored in a trusted database.
  21. A rootkit is:
    (a) Software that makes itself undetectable on a system.
    (b) A set of tools that allow a user to gain administrative access on a computer.
    (c) Malicious software that can gain administrative privileges on a system.
    (d) Software that replicates itself onto other systems.
  22. If it takes you one day to test all 7.2×1016 combinations of a 56-bit key, how long would it take with a 64-bit key?
    (a) 8 days
    (b) 256 days
    (c) 1,024 days
    (d) 1.8×1019 days (264)
  23. A chroot jail:
    (a) Limits the parts of a file system that a process can access.
    (b) Prevents a process from gaining administrative privileges.
    (c) Ensures that a process cannot execute specific system calls.
    (d) Runs a process in a secure environment to detect whether it has a virus.
  24. An interface definition language (IDL) is:
    (a) A machine-independent output language generated by the RPC precompiler.
    (b) Processed by the RPC precompiler to generate stub functions.
    (c) Used to implement remote procedures.
    (d) Used by the operating system to provide user processes with an interface to remote procedure calls.
  25. Why did earlier versions of Linux, BSD, and UNIX not implement access control lists?
    (a) They are redundant with protection mechanisms already in place.
    (b) They opted to use capability lists instead.
    (c) They don't fit within an inode.
    (d) The operating system only supports discretionary access control.