Discover how snapshots, replication and backups work together to protect your data. Learn the benefits, limitations and best practices for a layered data protection strategy
When it comes to data protection, backups are just one layer. Modern systems also provide storage-centric features like snapshots and replication, which offer far faster recovery than traditional backup alone.
This guide explains how snapshots work, why they can’t fully replace backups, and how replication fits into a comprehensive data protection plan.
What are snapshots?
A snapshot is a metadata record that captures the state of blocks and files in a volume at a specific point in time. Most NAS and SAN arrays include snapshot capability as a built-in feature.
Think of snapshots as a point-in-time index or “table of contents” for your storage. They track which blocks or files existed and where. If you roll back to a snapshot, the system reconstructs the earlier state by re-mapping or removing blocks as needed.
Deleted blocks must also be retained if they’re part of a snapshot, which means the rolled-back version is built from metadata, retained data and the original “parent” copy.
Crucially, snapshots are not backups. They don’t create a full second copy of your data. Individual snapshots take little space, but as deleted or changed blocks accumulate, total usage grows. Array vendors usually limit how many snapshots you can keep.
Benefits and limitations of snapshots
Benefits of snapshots:
- Rapid recovery — roll back to a previous point-in-time much faster than restoring from a backup.
- High frequency — because snapshots are lightweight, you can schedule them every hour or even every few minutes without disrupting production. Backups are typically once per day and outside business hours.
Limitations of snapshots:
- Local storage — snapshots are usually kept on the same device as the production data, so they’re vulnerable to the same failures or ransomware attacks.
- Storage capacity — as deleted or changed blocks accumulate, snapshots consume primary storage.
- Short retention — most organisations delete snapshots after 24–48 hours once backups have been taken.
For these reasons, snapshots work best as a complement to backup, giving you near-term protection while backups provide long-term, off-system resilience.
Snapshot mechanisms: Copy-on-write vs redirect-on-write
Storage arrays use two main snapshot methods:
Copy-on-write (COW)
When a write request changes a block, the original block is copied first and preserved by the snapshot. The snapshot contains pointers to unchanged blocks plus copies of changed ones. This saves space but adds overhead, as each write triggers a copy operation.
Redirect-on-write (ROW)
With ROW, new blocks are written elsewhere immediately and the snapshot references the original data. This reduces double-write overhead but can complicate data access and reconciliation when snapshots are deleted.
Both mechanisms balance space efficiency, performance and complexity differently, and vendors may implement hybrids.
Continuous data protection (CDP)
A special form of snapshotting is continuous data protection. CDP creates a snapshot every time a relevant change occurs (such as file creation, update or deletion). This enables roll-back to virtually any point in time.
The trade-off is performance overhead. Instead of periodic impact, CDP introduces extra I/O with every change. It’s best for critical workloads where granular recovery is essential.
Clones, mirroring and replication
Beyond snapshots, storage platforms offer clones and replication.
- A clone is a full copy of a unit of storage—not just metadata and changed blocks but the entire dataset. Clones can be one-off (e.g. copying a drive for testing) or recurring.
- Replication mirrors data to a second system. It can be:
- Synchronous replication — data is written to two locations at nearly the same time, providing an instantly usable failover copy for mission-critical workloads.
- Asynchronous replication — data is first written to the primary system, then sent to the secondary system later. This is less demanding on network performance but may lag slightly behind the primary.
Replication is common in transaction-sensitive industries like finance, healthcare and e-commerce because it offers instant availability and rapid disaster recovery. But it’s also costly and technically demanding, so many organisations replicate only their most critical datasets.
Why replication and snapshots don’t replace backups
While replication and snapshots are powerful, neither is a full substitute for backup:
- Replication is near real time. If corrupted or infected files appear on the primary system, they’re replicated too. Without backups, there’s no clean version to roll back to.
- Snapshots are local and short-term. If the array fails, snapshots may be lost as well.
A robust data protection plan layers these methods:
- Snapshots for frequent, near-term rollbacks (minutes to hours).
- Replication for instant failover of the most critical data.
- Backups for isolated, long-term recovery (days to years).
This combination ensures that even if snapshots expire or replication spreads corruption, there’s still a clean backup available to restore from.
Best practices for data protection
- Schedule snapshots frequently for critical datasets but monitor storage usage.
- Use replication selectively for high-value workloads that require instant failover.
- Back up everything to a separate system or cloud storage for long-term retention.
- Test your restores regularly to ensure snapshots, replication and backups work as expected.
The takeaway
In storage environments, no single technology covers every scenario. Snapshots provide fast, granular recovery points. Replication ensures a secondary system is ready to take over instantly. Backups provide historical, offline copies immune to primary-system failures.
By combining snapshots, replication and backups, you build a layered data protection strategy that balances cost, performance and risk—safeguarding your data against almost any eventuality.
Read more about data protection
Backup: Don’t leave it to hope. Build a solid data protection strategy. Discover how modern backup strategies protect against ransomware, cover cloud and container environments, and ensure business continuity with RPO and RTO.
RPO and RTO: Their roles in disaster recovery planning. RPO and RTO define acceptable downtime and data loss in disaster recovery. Learn how to calculate them, apply tiers, and align storage and cloud strategies.