Definition
Data Recovery
The process of restoring lost, corrupted, or inaccessible data from backups or other sources.
- Full baseline backup: A complete copy of every data file, kept off‑site so the entire database can be rebuilt after a catastrophic loss.
- Incremental / differential backup: Smaller jobs that capture only data blocks changed since the previous backup, reducing storage and network load while still allowing a full restore.
- Transaction log + point-in-time recovery: The DBMS replays the write-ahead log up to a chosen second, reversing accidental deletes or corruption without losing the day’s valid work.
- Real-time replication with automatic fail-over: Every committed change is streamed to a standby server that can take over instantly, keeping services online even if the primary site fails.
- RAID or similar hardware redundancy: Disks are mirrored or striped with parity so the system survives one (or more) drive failures without data loss or downtime.
- Documented disaster-recovery / business-continuity plan: Step-by-step procedures, roles, and time-targets that guide staff in restoring data and critical services after fire, flood, or cyber-attack.
Example
- Transaction log + point-in-time recovery
- A banking system processes thousands of transactions per second.
- If a software bug accidentally deletes today’s wire transfers, the DBMS can use its transaction log to roll forward all valid changes up to one second before the bug occurred.
- This allows the bank to recover without losing legitimate transactions that happened earlier in the day.
- Real-time replication with automatic fail-over
- E-commerce giants like Amazon use real-time replication to keep a standby database server synchronized with the primary.
- If the primary server in one region fails due to a power outage, the standby server in another region takes over immediately.
- Customers can continue browsing and checking out products with no visible interruption.
Social & ethical considerations 
- High reliability is expensive:
- Redundant hardware, off-site replication, and frequent log shipping raise infrastructure and energy costs.
- Yet organisations that hold mission-critical or sensitive personal data (medical records, payroll, student grades) have an ethical and often a legal duty to protect that information.
- Cutting corners to save money can lead to lost livelihoods, patient harm, or privacy breaches, undermining public trust.
- Digital divide:
- Large corporations can absorb the cost of multi-region fail-over, while small charities or schools may struggle, creating inequality in data protection.
- Ethically, vendors and policymakers must consider affordable managed backup services or grants so vital community data are not left vulnerable.
- Environmental impact:
- Constant replication and always-on standby servers consume power.
- Balancing environmental stewardship against data safety requires smart scheduling (e.g., incremental rather than full nightly backups) and green-energy data centres.
- Transparency and consent:
- Stakeholders should know how their data are safeguarded and the expected recovery time.
- Hiding weak protection to reduce expenditure is deceptive and erodes user confidence.
- Bottom line:
- The more irreplaceable and sensitive the data, the stronger the moral imperative to invest in robust recovery, even if that means higher upfront cost.
- At the same time, organisations should still seek equitable, energy-conscious solutions.
Example
- Reliability Concerns:
- In 2017, the UK’s National Health Service (NHS) was hit by the WannaCry ransomware attack.
- Many hospitals had not invested enough in redundant systems or regular backups due to budget constraints.
- As a result, patient records became inaccessible, surgeries were cancelled, and emergency care was disrupted.
- This case shows that cutting corners on reliability and data protection to save costs can directly harm patients, disrupt services, and erode public trust.
- Environmental Concerns:
- Google Cloud and Microsoft Azure have invested in green-energy data centres to offset the high power consumption of replication and always-on standby servers.
- For instance, instead of running full nightly backups, many clients use incremental backups and smart scheduling, which reduces unnecessary energy usage while still protecting data safety.
- This approach balances environmental stewardship with the need for strong disaster recovery and business continuity.
