압축 비율

Storage Pressure and the Role of Compression

백업, log archives, and disk images grow faster than most storage budgets.
You can add more disks, but that only delays the next capacity problem.

Compression introduces a smarter option.
Instead of storing every repeated pattern again, you reduce redundancy and keep a smaller representation that still allows full restoration when you need it.

Defining Compression Ratio for Files and Backups

Compression ratio describes how much a compressor shrinks data.
It compares original size to compressed size and expresses the reduction as a simple ratio.

Basic Formula and Examples

The usual formula looks like this:

Compression ratio = original size ÷ compressed size

예시:

• Original 10 GB, compressed 2 GB → 10 ÷ 2 = 5:1

• Original 800 MB, compressed 400 MB → 800 ÷ 400 = 2:1

A 5:1 ratio means the compressed file occupies one fifth of the original space.
Higher ratios indicate stronger size reduction for that dataset.

Interpreting Ratios in Real Workloads

You rarely see the same compression ratio everywhere.
Text logs and database dumps often compress extremely well, while JPEGs or already compressed archives show little gain.

When you tune backup policies, you should track:

• Typical ratio for each data type

• Resulting backup sizes and retention windows

• Restore time from compressed sets

This information helps you set realistic expectations for storage growth and recovery speed.

Factors That Shape Compression Results

The same algorithm can produce very different compression ratios on different inputs.
You gain more reduction when your data has structure and repetition.

Data Type and Redundancy

Highly structured, repetitive content performs best:

• Plain text, CSV, and XML

• Source code and log files

• Virtual machine images with unused blocks

Media files and encrypted data compress poorly:

• JPEG, MP4, MP3, and many game assets

• Encrypted containers and already compressed archives

You should avoid recompressing formats that already use strong compression, especially when speed matters more than tiny extra gains.

Algorithm Choice and Settings

Different algorithms follow different trade-offs:

• Fast general-purpose compressors (like DEFLATE-style) balance speed and reduction

• Newer methods (like modern dictionary or entropy schemes) push higher ratios at higher CPU cost

• Specialized tools target backups, databases, or forensic images with format-aware techniques

Block size, dictionary size, and “level” settings also change behavior.
Higher levels usually increase compression ratio and CPU load at the same time.

Limits with Encrypted or Already Compressed Data

Encrypted data should look random to an observer.
Random data contains almost no patterns, so compressors gain almost nothing.

If a backup includes large encrypted volumes or many compressed media files, you should not expect impressive ratios.
In that scenario, deduplication or selective backup often helps more than aggressive compression.

Balancing Compression Ratio, Speed, and Risk