Problem Areas

Problem Areas

1. Distributed Database Design

   • Replication (fully replicated or partially replicated)
   • Fragmentation and optimum distribution of fragments

2. Distributed Query Processing

   • Executing query over network the most cost effective way
   • - Use of parallelism

3. Distributed Directory Management

   • Directory contains information about data items in database
   • Global or local 
   • Centralized or distributed 

4. Distributed Concurrency Control

   • Maintain integrity and consistency of multiple copies of the database
   • Mutual consistency is the condition where multiple copies of every data item to have the same value
   • Pessimistic - synchronize user request before execution
   • Optimistic - execute before checking whether database consistency is compromised
   • Locking - Mutual exclusion of accesses to data items
   • Time-stamping - Transactions are executed in some order

5. Distributed Deadlock Management

6. Reliability of DDBMS

   • Recovery from failure at failed sites

7. Operating System Support

   • Support by operating system does not correspond to the requirements of database management software

8. Heterogeneous Databases

   • Data model (File system, relational, object oriented)
   • Data language (DB2 uses SQL, INGRES uses QUEL)
   • Data translation
   • Data manipulation instructions translation