Redundant RAID levels primarily maintain availability through a specified number of drive failures; RAID 0 provides no drive-failure protection. RAID mirrors or stripes data across multiple drives so that if one fails, the array keeps running and you retain access. That is all RAID does. It does not protect against accidental deletion, ransomware, silent data corruption, the NAS unit itself failing, fire, theft, or human error. RAID alone does not protect against these events; recoverability depends on additional controls such as snapshots, backups, filesystem-integrity features, and physical separation.
In short: RAID is redundancy, not backup. They solve different problems. RAID keeps your NAS running when a drive fails. Backup provides recovery options for loss scenarios RAID does not cover, provided the backup is current, intact, restorable, and protected from the incident. You need both.
What RAID Actually Does
RAID distributes data across multiple physical drives using one of several configurations. The most common for home and SMB NAS use are RAID 1 (mirroring), RAID 5 (striping with parity), RAID 6 (striping with dual parity), and Synology's SHR is an automated RAID-management system that can use mismatched drive sizes efficiently and provide one- or two-drive redundancy when sufficient drives are installed.
In RAID 1, every write is duplicated across two drives simultaneously. If one drive fails, the other contains a complete copy and the system keeps running in a degraded state until the failed drive is replaced and the array rebuilds. In RAID 5, data is striped across three or more drives with parity information distributed across the array, tolerating a single drive failure. RAID 6 extends this to tolerate two simultaneous drive failures, at the cost of more storage overhead and slower write performance.
The common thread across all RAID levels: RAID exists to keep a storage system running through hardware failures. It is a high-availability mechanism. It is not a data protection mechanism in the broader sense that a backup is.
What RAID Does Not Protect Against
Accidental deletion. If you delete a file, RAID deletes it from all drives simultaneously. The deletion propagates instantly across the entire array. There is no version to recover from. RAID has no concept of history, only of the current state of the drives.
Ransomware. Ransomware encrypts files on the live file system. Because RAID maintains a synchronised copy of everything written to it, the encrypted versions propagate across every drive in the array in real time. A RAID 1 mirror is just as efficient at replicating ransomware-encrypted files as it is at replicating clean ones.
Silent data corruption. Some RAID configurations, particularly those using ext4 without block-level checksums, cannot detect corruption below the file system level. A bit that flips silently can be replicated across the array before anyone notices. ZFS-based systems handle this better through regular scrubs and self-healing, but the underlying issue holds: RAID replicates what is there, good or bad, without verifying the integrity of the data itself.
The NAS unit itself failing. If the NAS enclosure fails, the controller dies, or the operating system becomes corrupted, all drives become inaccessible simultaneously. Even with RAID 1, the data on those drives may be unreadable without an identical NAS to plug them into. RAID protects against one drive failing inside a working device. It does not protect against the device housing those drives failing.
Theft, fire, and flood. Both drives in a RAID 1 array sit in the same enclosure, on the same shelf, in the same room. A single theft, fire, or flood event destroys both copies at the same time.
Human error beyond deletion. Overwriting a file with the wrong version, corrupting a database, running a script in the wrong directory, saving over 12 months of work with a blank template - all of these propagate immediately across the array.
RAID 1 is two copies of your data in one location. That is redundancy, not backup. True backup requires a separate copy, stored separately, ideally with versioning so you can recover from any point in time.
RAID vs Backup: What Each Protects
What RAID covers vs what backup covers
| RAID 1 | RAID 5/6 | 3-2-1 Backup | |
|---|---|---|---|
| Single drive failure | Protected | Protected | Protected (also) |
| Accidental deletion | Not protected | Not protected | Protected (if versioned) |
| Ransomware | Not protected | Not protected | Protected if an unaffected recovery copy is retained, such as an offline or immutable backup |
| NAS device failure | Not protected | Not protected | Protected |
| Fire or flood | Not protected | Not protected | Protected (if offsite) |
| Silent data corruption | Not protected (ext4) | Not protected (ext4) | Protected (if verified) |
| File version recovery | Not possible | Not possible | Possible |
| Protects against two drive failures | No | RAID 6 only | Yes |
The Scenarios That Catch People Out
Ransomware is the most devastating failure mode for RAID-only setups. Ransomware campaigns have targeted internet-exposed NAS devices, and RAID alone does not provide a clean recovery point after encryption. QNAP documented 2022 DeadBolt and Checkmate ransomware attacks against internet-exposed NAS devices; the published attack vectors included software vulnerabilities and weak-password SMB accounts. Recovery outcomes depend on whether an unaffected recovery point, such as a valid snapshot or clean backup, remains available. The lesson is not specific to QNAP - any NAS exposed to the internet without proper access controls is a target, regardless of brand or RAID level.
Another failure mode is accidental deletion at scale. A sync job runs in the wrong direction and overwrites a source folder with an empty destination. A user deletes the wrong top-level directory. A developer runs a cleanup script on a mounted NAS path instead of a local temp directory. RAID provides no record of what was there before the deletion. Without versioned snapshots or a separate backup, that data is gone permanently.
Another failure mode is failure of the NAS device itself. NAS devices fail. Power surges destroy controllers. Firmware updates corrupt boot volumes. Capacitors fail after years of 24/7 operation. Recovery after enclosure failure depends on vendor and model compatibility, RAID implementation, filesystem, and encryption; some vendors support migration to compatible replacement models without requiring an identical NAS. This is a scenario where even RAID 6 provides no protection at all.
What a Real Backup Strategy Looks Like
The 3-2-1 rule is the standard: three copies of your data, on two different media types, with one copy stored offsite. Applied to a NAS setup, this typically means:
- Copy 1: The NAS itself (this is your working copy, not a backup)
- Copy 2: A second NAS or external drive at the same location (local backup - fast recovery)
- Copy 3: Cloud backup or a drive stored at a separate physical location (protection against site-level events)
Versioning is essential. Versioning materially improves recoverability. A latest-version-only backup can lose its clean recovery point if a later run propagates deleted or encrypted files. Backup solutions like Synology Hyper Backup, QNAP Hybrid Backup Sync, and Veeam support versioned backups that let you recover from retained backup versions within the configured retention window. For a full comparison of NAS backup software, see the NAS backup software guide.
For the complete 3-2-1 strategy breakdown including cloud provider options and hardware recommendations, see the 3-2-1 backup strategy guide for NAS.
How RAID and Backup Work Together
RAID and backup are not alternatives to each other. They are complementary layers of a complete data protection strategy.
RAID handles the hardware availability layer: keeping the NAS running through drive failures, avoiding unplanned downtime, and giving you time to replace a failed drive without scrambling to restore from backup. For a business NAS shared by multiple users, RAID is essential because a non-redundant NAS going offline mid-workday is disruptive and potentially costly.
Backup handles everything RAID cannot: accidental deletion, ransomware, device failure, and site-level events. The correct mental model is this: RAID is availability insurance, backup is recovery insurance. Running RAID without backup is like insuring your car but not your house. Different risks, different coverage.
Common Mistakes Worth Knowing
Thinking RAID 1 gives you double the storage. It does not. RAID 1 gives you the full capacity of one drive, not two. Two 8TB drives in RAID 1 give you 8TB usable. The second drive is entirely consumed by the mirror.
Treating RAID rebuild time as low-risk. During a rebuild, the array is under sustained stress. Every surviving drive is being read in full to reconstruct the failed drive's data. During rebuild, the array has reduced fault tolerance, and sustained I/O can expose latent faults on surviving drives. In RAID 5, a second drive failure during rebuild means total data loss. Replace failed drives promptly and keep current backups at all times, but especially during a rebuild.
Thinking RAID 6 or SHR-2 eliminates the need for backup. RAID 6 tolerates two simultaneous drive failures. It does not protect against ransomware, accidental deletion, or the NAS device itself failing. The distinction is always the same: RAID protects against drive hardware failure. That is the only thing it protects against.
Australian Buyers: What You Need to Know
Ransomware and AU NAS users. QNAP has documented ransomware campaigns against internet-exposed QNAP NAS devices. It did not. The lesson applies to all NAS brands: any device reachable from the internet is a potential target. Use a vendor-supported secure remote-access method and avoid exposing NAS management interfaces directly to the internet; combine access controls with an unaffected recovery copy, such as an offline or immutable backup. For more on hardening your NAS against network threats, see the NAS security and ransomware protection guide.
Cloud backup and NBN upload speeds. NBN upload speeds vary by access technology, retail plan, and provider; check the current advertised upload speed for the specific service before estimating cloud-backup duration. Uploading 4TB of data at 20Mbps takes approximately 18 days of continuous transfer. This makes the offsite cloud backup component of a 3-2-1 strategy a long-horizon project for larger NAS setups. The practical approach is to start cloud backup with your most critical data, complete the initial seed over several weeks, and then let incremental daily backups maintain the copy. After the initial seed, incremental-backup volume depends on how much data changes; feasibility depends on the connection's upload speed, backup window, and data-change rate.
Warranty covers hardware, not data. Australian Consumer Law rights apply independently of any manufacturer's warranty. If a NAS fails to meet a consumer guarantee, the seller must provide the applicable remedy, which depends on whether the failure is major or minor and may be a repair, replacement, or refund. Warranty covers the hardware purchase. It does not cover the contents. A hardware warranty does not guarantee restoration of lost data. Separate backups remain essential, while any ACL entitlement to compensation for associated loss depends on the circumstances, including whether the loss was reasonably foreseeable. This is the most important reason to maintain a real backup independent of the NAS or its warranty status.
Power events. A compatible, correctly configured UPS can give the NAS time to save data and shut down safely during a power failure, reducing the risk of data loss or filesystem corruption. RAID does not protect against this. For UPS selection and sizing, see the UPS for NAS Australia guide.
Related reading: our NAS buyer's guide.
Free tools: RAID Calculator and Backup Storage Calculator. No signup required.
See also: our complete QNAP NAS Australia guide.
See also: our full Backup and Disaster Recovery guide.
Does RAID count as a backup?
No. RAID is redundancy, not backup. It protects against drive hardware failure only. It does not protect against accidental deletion, ransomware, NAS device failure, or physical events like fire or theft. A real backup requires a separate copy of your data, stored separately from the NAS, ideally with versioning.
Do I still need RAID if I have a good backup strategy?
Yes, for different reasons. RAID keeps your NAS running through a drive failure without needing to restore from backup, which can take hours or days depending on data volume. If the NAS serves multiple users or holds data you access constantly, RAID prevents unplanned downtime. Backup is for data recovery after a loss event. RAID is for availability through hardware failure. They address different risks and you benefit from having both.
Can ransomware be recovered from if I have RAID?
No. Ransomware encrypts files on the live file system. Because RAID maintains an exact synchronised copy of everything, the encrypted versions propagate across all drives in the array. RAID alone does not reverse ransomware encryption. Recovery requires an unaffected recovery point, such as a valid snapshot or clean backup; offline or immutable copies reduce the risk that recovery data is also altered.
What is the difference between RAID 1 and RAID 5 for home use?
RAID 1 mirrors two drives, giving 100% redundancy with the simplest rebuild process. For a two-bay NAS, RAID 1 is a common standard RAID option for one-drive fault tolerance; supported alternatives vary by model and may include SHR, Basic, JBOD, or RAID 0. RAID 5 uses three or more drives with distributed parity, giving you the capacity equivalent of N-1 drives with a single drive failure tolerance. RAID 5 is more storage-efficient than RAID 1 on larger arrays but more complex to rebuild. Synology's SHR is a flexible variant that handles mixed drive sizes. For a full comparison of RAID levels, see the NAS RAID explained guide.
Is ZFS better than RAID for data protection?
ZFS adds block-level checksums that detect silent corruption; when a valid redundant copy is available, it can self-heal detected damage during reads or scrubs. QNAP's QuTS Hero and TrueNAS both use ZFS. But ZFS still does not protect against deletion, ransomware, or device failure. ZFS improves the quality of your redundancy. It does not replace backup.
How often should I test my backups?
At minimum, quarterly. A backup job can appear to run successfully yet still fail to restore, so restoration should be tested regularly. Test restoration of a sample of files from your backup every three months. Set a calendar reminder. A backup you have never tested is a backup you cannot rely on.
What backup software works with Synology and QNAP NAS devices?
On compatible Synology models, Hyper Backup supports local and cloud destinations with versioning; Active Backup for Business backs up PCs and servers but requires a supported x64 Synology NAS using Btrfs. QNAP includes Hybrid Backup Sync, which supports a wide range of local, cloud, and NAS-to-NAS destinations. Both support cloud destinations including Backblaze B2, Amazon S3, and others accessible from Australia. For a full comparison, see the NAS backup software guide.
RAID is the first layer. The next step is building the backup layer that protects against everything RAID cannot handle. The 3-2-1 backup strategy guide covers local backup hardware, cloud provider options for Australian NBN speeds, and how to set up versioned backups on Synology and QNAP.