Backup and Recovery

Backup creates extra copies of data, and recovery brings that data back after a loss. Hardware failures, human mistakes, ransomware, and natural disasters all destroy information. A tested backup plan lets an organization restart quickly and keep customer trust.

The Spare Key Analogy

Imagine locking your house keys inside the house. A spare key at a neighbor's home saves the day, but only if the neighbor lives nearby, remembers where the key sits, and the key actually fits the lock. Backups follow the same rules. Copies must exist, stay reachable, and work when needed.

Why Cloud Data Still Needs Backup

Many beginners believe the provider backs up everything. Providers keep their platform running, and customers own the protection of their own data. A deleted database, an overwritten file, or an encrypted disk stays gone unless the customer holds a separate copy.

Key Terms

Recovery Point Objective (RPO)

RPO measures how much data an organization can afford to lose, counted in time. An RPO of one hour means backups run at least hourly.

Recovery Time Objective (RTO)

RTO measures how long the business can stay offline. An RTO of four hours means systems must return within four hours.

 Last good backup        DISASTER           Systems running again
        |                   |                        |
 ------[X]------------------[!]----------------------[OK]------> time
        |<----- RPO ------->|<--------- RTO -------->|
         (data lost)           (downtime)

The 3-2-1 Rule

The 3-2-1 rule gives a simple recipe for resilient backups:

• Keep 3 copies of important data
• Store them on 2 different kinds of storage
• Place 1 copy in a separate location or account

Live Data --> Copy 1: Same-region snapshot
          --> Copy 2: Different-region storage
          --> Copy 3: Separate account, write-protected

Backup Types

Type	What It Saves	Speed	Storage Cost
Full	Everything	Slow	High
Incremental	Changes since the last backup	Fast	Low
Differential	Changes since the last full backup	Medium	Medium
Snapshot	Disk state at one moment	Very fast	Low to medium

Security Features for Backups

• Encrypt every backup at rest and in transit
• Use immutable storage so nobody can alter or delete copies for a set period
• Restrict backup access to a small group with multi-factor authentication
• Store copies in a separate account so a stolen main account cannot erase them

Real-World Example: A Ransomware Scare

A design agency wakes up to encrypted files and a ransom note. Attackers deleted the backups stored beside the main data. A competitor in the same situation kept immutable copies in a separate account. That team wiped the infected systems, restored the last clean copy from the night before, and lost half a day of work. The agency without protected backups lost three weeks.

Building a Recovery Plan

1. List critical systems and rank them by business importance.
2. Set an RPO and RTO for each system.
3. Choose backup types and schedules that meet those targets.
4. Document restore steps in plain language.
5. Assign a named owner for each system.
6. Test restores on a regular calendar.

Testing Matters Most

An untested backup is a hope, not a plan. Teams discover corrupt files, missing keys, and outdated instructions only during a real restore. Run test restores quarterly, time them, and compare results to the RTO.

Key Takeaways

Backups protect against loss, and recovery plans define how fast a business returns to normal. Follow the 3-2-1 rule, lock copies against tampering, and test restores often.