Page 228 - DCAP402_DCAO204_DATABASE MANAGEMENT SYSTEM_MANAGING DATABASE

Unit 13: Parallel Databases

Notes
Figure 13.2: Speed-up and Scale-up

Speed-up and scale-up are illustrated in Figure 13.2. The speed-up curves show how, for a fixed
database size, more transactions can be executed per second by adding CPUs. The scale-up
curves show how adding more resources (in the form of CPUs) enables us to process larger
problems. The first scale-up graph measures the number of transactions executed per second as
the database size is increased and the number of CPUs is correspondingly increased. An alternative
way to measure scale-up is to consider the time taken per transaction as more CPUs are added to
process an increasing number of transactions per second; the goal here is to sustain the response
time per transaction.

Task Parallel database vs. distributed data base.

13.2 I/O Parallelism

Parallel execution dramatically reduces response time for data-intensive operations on large
databases typically associated with Decision Support Systems (DSS) and Data warehouses.
Symmetric Multiprocessing (SMP), clustered systems, and Massively Parallel Systems (MPP)
gain the largest performance benefits from parallel execution because statement processing can
be split up among many CPUs on a single Oracle system. You can also implement parallel
execution on certain types of Online Transaction Processing (OLTP) and hybrid systems.
Parallelism is the idea of breaking down a task so that, instead of one process doing all of the
work in a query, many processes do part of the work at the same time. An example of this is
when 12 processes handle 12 different months in a year instead of one process handling all 12
months by itself. The improvement in performance can be quite high.
Parallel execution helps systems scale in performance by making optimal use of hardware
resources. If your system’s CPUs and disk controllers are already heavily loaded, you need to
alleviate the system’s load or increase these hardware resources before using parallel execution
to improve performance.

Some tasks are not well-suited for parallel execution.

Example: Many OLTP operations are relatively fast, completing in mere seconds or
fractions of seconds, and the overhead of utilizing parallel execution would be large, relative to
the overall execution time.
There are various ways in which partitioning will be done which are as follows:
1. Horizontal partitioning
2. Vertical partitioning

LOVELY PROFESSIONAL UNIVERSITY 221