Other sample final questions:

1. Define the following terms:

   Serializable schedule

   Recoverable schedule

   Cascading abort

   Deadlock

   Write-Ahead Log

   Checkpoint

2. What are the ACID properties?  Which are ensured by concurrency control
   and which by logging and recovery?

3. Give examples of three different types of serializability problems that can 
   arise because of concurrency.

4. Explain how a lock manager should respond to requests for locks on objects
   assuming that there are two types of locks Share (S) and Exclusive (X)
   locks.

5. What is two phase locking (2PL)?  How does it ensure serializability?
   What is strict 2PL?  What additional property (beyond serializability)
   does it guarantee?

6. What is meant by the term deadlock prevention?  Give two schemes for
   preventing deadlock and talk about the advantages and disadvantages of
   each scheme?

7. How does deadlock detection work?  Why do many systems choose to use
   detection rather than prevention?

8. Why do many systems provide the ability to lock objects at various levels
   of granularity?  Give an example of a locking system that might be used
   in such a database.

9. Using a logging system for recovery, what happens when a commit command
   occurs in a transaction (and in what order)?

10. What are the three phases of crash recovery?  What happens in each phase
    (and in what order)?