Learning Guide Unit 7
Overview
Unit 7: File processing and external storing
Topics:
Primary versus secondary storage
Disk drives: architecture and access costs
Buffers and Buffer Pools
External Sorting
Learning Objectives:
Understand and be able to articulate the differences between primary and secondary storage including:
0
Computer RAM memory
0
Disk drives and other secondary storage devices
0
The concepts of persistent versus volatile memory
0
Data structures implemented in secondary storage (file structure)
Understand the terminology and architecture of disk drives
Understand the concept of disk access costs associated with head movement and seek time
Understand and be able to implement a buffer cache that incorporates concepts such as:
0
Buffer pool
0
Heuristics including FIFO, LFU, and LRU approaches
0
Virtual memory
Understand and be able to implement External sorts and be familiar with key concepts including:
0
Key sorts and index files
Quicksort and mergesort implementations
0
0
The replacement selection algorithm
0
Merging
Introduction
In unit 7, we explore file processing and external sorting algorithms. It is important to understand what
storage and when it is important to use each.
Internal storage, which is what all of the algorithms and data structures that we have dwelt with so far have used, is the internal memory (often referred to as RAM or random access memory) of the computer. This
to say that RAM is orders of magnitude faster than external storage. External storage, on the other hand, doesn't have as much performance, but it does have characteristics such as size and persistence that
make it equally as important as internal storage. The following table compares internal and external storage
and begins to show why each is so important.
Internal Storage
External
Very fast 100,000 to 1,000,000 times as fast as external storage
Slower performance whe storage
Cost is high per unit of storage is high when compared with external storage
Cost per unit of storage i:
Quantity of storage is relatively low when compared with External storage
Size of storage is virtuall
Data stored is persistent media until erased
Data stored is not persistent ... when the power is turned off the data is lost
By considering these characteristics we can come to some conclusions. First since internal storage is so
internal storage. Sorts conducted in internal storage are much faster than those that must access external storage.
In our text, Shaffer leads us to an understanding of disk drive technology, which is the most common form of external storage. Key topics that are introduced include an understanding of the operation of a disk drive and how rotation, disk head movement contributes to the latency of accessing information from the disk drive. This latency 'cost' of accessing data from a disk drive can be mitigated in part through the use of a
buffer cache or buffer pool. The basic idea behind the buffer cache is that the disk drive will often read more information in each I/O operation than is required by the program. If this informatio