CIS 6930/4930 Fall 2000 -- Unix System Administration

Midterm Examination

This closed-book examination is to take 50 minutes. Please answer five of the six questions presented below. Make sure you answer briefly but clearly. Extraneous incorrect information will cause the score for otherwise correct answers to be lowered.

No aid of any kind is allowed on this examination.

Pledge (Must be signed according to UF Honor Code)

On my honor, I have neither given nor received unauthorized aid in doing this assignment.

________________________________ (signature)

Examination Questions

  1. Briefly explain the difference between starting system services from init, an rc file, or inetd. Make sure to tell when a daemon started in each different place will begin execution.

    Also explain why named is started from an rc file rather than by inetd.

    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
  2. Briefly explain the difference between an Internet domain and an Internet zone. Discuss in particular the relationship between the cise.ufl.edu and the unix-ippd.cise.ufl.edu domains and zones.

    Why aren't hierarchical domains enough to handle the task for the whole Internet (mention the problem that cropped up with the hosts.txt file).

    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
  3. Software can be distributed either as source code or as compiled binary object code.

    First mention the tradeoffs (benefits drawbacks of each) for distribution of software as source code and as object code.

    Next describe the fundamental differences in philosophy between distribution of object code as licensed program packages (lpps) or as depotized code. What are the benefits and drawbacks of each?

    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
  4. What is the whole point of using tcp wrappers? What kinds of problems are exhibited by some of the daemons exhibited by inetd and what does tcp wrappers do to alleviate those problems?

    What is the primary difference between the regular way inetd starts a daemon and the way it starts one when using tcp wrappers?

    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
  5. What is the benefit to having a shadow password file?

    Why can't we achieve the same result just by making sure the file /etc/passwd is not readable?

    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
  6. Why is it important to have the PTR records and the in-addr.arpa hierarchy in the Internet domain scheme. What purpose do these records serve and how does the approach taken on the Internet provide an effective mechanism for supporting this capability?
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    

This document is copyright 2000 by Joseph N. Wilson.