Backups & Recovery

Why Backups Matter

Data recovery is the process of restoring data after an unexpected event - hardware failure, accidental deletion, ransomware, natural disaster. The difference between a minor inconvenience and a business-ending catastrophe often comes down to one thing: did you have backups?

Planning Your Backup Strategy

Before choosing tools, answer these questions:

What data needs backing up? - emails, databases, financial records, server configs, user files
How much total data? - and how fast is it growing?
Where will backups be stored? - on-site, off-site, or both?
How often? - daily? hourly? real-time replication?
What’s your RTO and RPO?

RTO & RPO

Metric	Full Name	Meaning
RTO	Recovery Time Objective	How long can you be down before it’s a crisis?
RPO	Recovery Point Objective	How much data can you afford to lose? (measured in time)

Example: RTO = 4 hours means you need to be back online within 4 hours. RPO = 1 hour means your most recent backup must be no older than 1 hour.

Types of Backup

Type	What It Copies	Pros	Cons
Full	Everything - all files, every time	Simple to restore	Slow, uses the most storage
Differential	Everything changed since the last full backup	Faster than full, one restore step	Grows larger over time
Incremental	Only changes since the last backup (any type)	Fastest, smallest	Restore requires full + all incrementals in order

File Compression

Backups can be compressed to save storage space, but there’s a trade-off:

Compressed backups take longer to restore (decompression step)
Some data types (video, images) don’t compress well since they’re already compressed

On-Site vs. Off-Site Backups

Approach	Advantages	Disadvantages
On-site (NAS, local drives)	Fast backup & restore, full control	Vulnerable to fire, flood, theft - same physical risk as the originals
Off-site (cloud, remote DC)	Protected from local disasters, accessible anywhere	Slower transfers, ongoing cost, dependency on internet/provider

Backup Tools

Linux - `rsync`

The Swiss army knife of Linux backups. Copies files efficiently, only transferring deltas (changed portions).

# Local backup
rsync -avz /home/user/ /backup/user-backup/

# Remote backup over SSH
rsync -avz -e ssh /home/user/ admin@backup-server:/backups/user/

Key features:

Supports compression (-z)
Transfers over SSH for security
Syncs between remote machines - great for automated cron jobs

macOS - Time Machine

Apple’s first-party backup solution:

Incremental model - backs up changes automatically every hour
Restore entire system or individual files
Restore older versions of files (versioned backups)

Windows - Backup and Restore

Two modes:

File-based: Backs up files to a zip archive - supports full and incremental
System image: Block-level disk copy - supports differential (backs up only changed blocks since last image)

Cloud Sync Services

For user devices and personal files:

Dropbox, Google Drive, iCloud, OneDrive - automatic file sync across devices
Easy for users, reduces IT workload for user data recovery
Particularly useful for remote/mobile workers whose devices aren’t always on corporate network

Archival Storage

For long-term data that doesn’t need frequent access:

Data tapes - slower but much cheaper per GB than disk
Used for regulatory compliance, legal holds, historical archives

RAID ≠ Backup

RAID (Redundant Array of Independent Disks) combines multiple drives for redundancy and/or performance:

Level	Description	Backup Replacement?
RAID 1	Mirrored drives	❌ No - mirrors corruption too
RAID 5	Striped with parity	❌ No - doesn’t protect against deletion or ransomware
RAID 10	Mirror + stripe	❌ No - same reasons

Testing Backups

A backup you’ve never tested is a backup you don’t actually have.

What to Test

Data integrity - can files actually be opened and read after restore?
Full restore process - can you rebuild a server from scratch using only the backup?
Recovery time - does it meet your RTO?
Documentation - is the recovery process documented clearly enough that someone else could do it?

How Often

Test backup restoration at least quarterly
Run a full disaster recovery simulation annually - involve different teams
After any major infrastructure change, verify backups still work

User Backups

Don’t forget end users! Infrastructure backups won’t help if a user’s laptop dies and their files are only local.

Solutions:

Set up cloud sync (Google Drive, OneDrive) on all company devices
Educate users on where to store important files
Configure automatic backup policies on user machines