PARALLEL AND DISTRIBUTED DATABASES
CONTENT
- Parallel Databases
- Architecture for Parallel Databases
- Distributed Databases
- Properties of Distributed Databases
Parallel Databases
Definition of Parallel Databases
A parallel database system employs parallelism to enhance performance across various operations, including data loading, index construction, and query execution.
Architecture for Parallel Databases
The core concept behind a parallel database is the simultaneous execution of tasks to enhance overall performance. Three primary architectures have been proposed for building Database Management Systems (DBMS).
Shared-Memory System
In a shared-memory system, multiple Central Processing Units (CPUs) are interconnected and have access to a common segment of main memory.
Shared-Disk System
In a shared-disk system, each CPU possesses its private memory and enjoys direct access to all disks through an interconnected network.
Shared-Nothing System
In a shared-nothing system, each CPU has its local main memory and disk space, with no two CPUs sharing the same storage area. Communication between CPUs is exclusively through a network connection.
These three architectures can be visualized as follows:
Advantages of Parallel Databases
-
Higher Performance: The availability of multiple CPUs allows for increased speedup and scaleup.
- NEW: Download the entire term's content in MS Word document format (1-year plan only)
- The complete lesson note and evaluation questions for this topic
- The complete lessons for the subject and class (First Term, Second Term & Third Term)
- Media-rich, interactive and gamified content
- End-of-lesson objective questions with detailed explanations to force mastery of content
- Simulated termly preparatory examination questions
- Discussion boards on all lessons and subjects
- Guaranteed learning