NAS RAID Explained: RAID 0, 1, 5, 6, 10 and SHR Guide 2026

RAID is the foundation of how a NAS protects your data across multiple drives. This guide explains RAID 0, 1, 5, 6, 10, and Synology's SHR in plain language. What each level does, how many drives you need, how much usable storage you get, and which to choose for your situation.

Choosing the wrong RAID level is one of the most common causes of preventable data loss on a NAS. And most NAS users make the decision without understanding what they're actually trading off. This guide covers every RAID level relevant to NAS use. RAID 0, 1, 5, 6, 10, and Synology's SHR. With real capacity numbers, fault tolerance ratings, rebuild risk profiles, and the decision logic for choosing between them. It draws on drive failure statistics, NAS vendor documentation, and the practical constraints of home and SMB multi-drive setups. By the end, you'll know which RAID level suits your bay count, drive size, and workload. And which configurations carry risks that aren't obvious from marketing specs. Australian consumer rights and retailer notes are in the AU section below.

In short: For a 2-bay NAS, use RAID 1 (or SHR if you're on Synology). For a 4-bay NAS, use RAID 5 or SHR-1. For a 4-bay NAS storing critical data, use RAID 6 or SHR-2. For maximum performance with redundancy, use RAID 10. Whatever RAID level you choose: RAID is not a backup. Always maintain a separate offsite copy of critical data.

What Is RAID and Why Does It Matter on a NAS?

A NAS (Network Attached Storage) device holds multiple hard drives. Without RAID, those drives function independently. Data on one drive is entirely unprotected if that drive fails. RAID solves this by treating multiple drives as a single logical storage pool, distributing data (and in most cases, redundancy information) across all drives in the array.

The core trade-off in every RAID level is between three factors: capacity efficiency (what percentage of raw drive space you can actually use), fault tolerance (how many drives can fail before you lose data), and performance (read and write speeds relative to single-drive operation). No RAID level maximises all three simultaneously. Every choice involves a trade-off.

For a practical example of how RAID capacity plays out: our RAID Capacity Calculator shows usable storage for any drive count and size across all RAID levels. Enter your drives and it calculates exactly what you'd get from RAID 0, 1, 5, 6, 10, and SHR. Useful for comparing options before committing.

RAID is not a backup. RAID protects against drive failure only. It does not protect against ransomware (all drives encrypt simultaneously), accidental deletion (all drives delete simultaneously), fire, theft, flood, or NAS hardware failure. You need a separate backup strategy. Ideally following the 3-2-1 rule. In addition to your RAID configuration.

RAID 0. Striping (No Redundancy)

RAID 0. Striping (No Redundancy)Drive 1Drive 2Stripe AStripe AStripe BStripe BStripe CStripe CStripe DStripe DMIN DRIVES2USABLE SPACE100% of rawFAULT TOLERANCE0 drivesBEST FORSpeed (no data risk)

RAID 0 splits data across all drives in the array (called striping) to maximise performance and use 100% of available capacity. Reading and writing a file uses all drives simultaneously. In theory doubling throughput with two drives, tripling with three, and so on. For workloads that move large files sequentially (4K video editing, database scratch space), the performance is genuinely excellent.

The catch: RAID 0 has zero fault tolerance. There is no redundancy whatsoever. If any single drive in the array fails, the entire array becomes unreadable. Not just the data on that drive. With two drives in RAID 0, you double your performance but also double your probability of a catastrophic failure event. This is the worst RAID level for data safety, which is why most NAS operating systems include a clear warning when you select it.

Usable capacity: 100% of total raw drive space. Four 8 TB drives = 32 TB usable.

Minimum drives: 2

Fault tolerance: None. One drive failure = total data loss.

When it makes sense: Scratch disks, temporary caches, or non-critical high-performance storage where data can be recreated. Never for primary NAS storage. If you're considering RAID 0 for performance reasons, investigate whether a RAID 5 or 10 array on your specific NAS is actually bottlenecked by RAID overhead before sacrificing redundancy.

RAID 1. Mirroring

RAID 1. MirroringDrive 1 (Data)Drive 2 (Mirror)Data Block AData Block AData Block BData Block BData Block CData Block CData Block DData Block DautomirrorMIN DRIVES2USABLE SPACE50% of rawFAULT TOLERANCE1 driveBEST FORCritical data

RAID 1 writes an identical copy of all data to every drive in the array simultaneously. With two drives in RAID 1, both drives contain exactly the same data at all times. If one drive fails, the other continues operating without interruption. The NAS flags the failure, you replace the dead drive, and the system rebuilds the mirror automatically.

Usable capacity: 50% of total raw drive space. Two 8 TB drives = 8 TB usable (not 16 TB).

Minimum drives: 2

Fault tolerance: 1 drive can fail without data loss. With two drives in RAID 1, either drive can fail completely and your data is untouched.

Performance: Read speeds can be slightly improved by serving reads from both drives. Write speeds are limited to the speed of a single drive. Every write must go to all drives before it's confirmed.

When it makes sense: RAID 1 is the standard choice for 2-bay NAS setups. The 50% capacity penalty is the price of redundancy. If you have 16 TB of raw drive space and only 8 TB is usable, you're trading half your capacity for protection. A worthwhile trade for most users. For 4-bay and larger NAS units, RAID 5 offers better capacity efficiency with equivalent single-drive fault tolerance.

RAID 5. Striping with Single Parity

RAID 5. Striping with Single ParityDrive 1Drive 2Drive 3Drive 4Data 1Data 2Data 3ParityData 4Data 5ParityData 6Data 7ParityData 8Data 9ParityData 10Data 11Data 12MIN DRIVES3USABLE SPACE(N−1) ÷ N drivesFAULT TOLERANCE1 driveBEST FORBalanced home NAS

RAID 5 stripes data across all drives in the array and adds parity information distributed across all drives as well. The parity data is a mathematical checksum that allows the NAS to reconstruct any single drive's data from the remaining drives. The result: you lose the equivalent of one drive's worth of space to parity, but gain the ability to survive one drive failure without data loss.

Usable capacity: Total raw capacity minus one drive. Three 8 TB drives in RAID 5 = 16 TB usable (not 24 TB). Four 8 TB drives = 24 TB usable (not 32 TB). Capacity efficiency improves as drive count increases. RAID 5 with 4 drives uses 75% of raw capacity vs 50% for RAID 1.

Minimum drives: 3

Fault tolerance: 1 drive can fail without data loss. The array operates in degraded mode until the failed drive is replaced and rebuilt.

Performance: Good read performance due to striping across all drives. Write performance has some overhead from parity calculation. Noticeable on older ARM-based NAS units, largely irrelevant on modern Intel/AMD models that handle parity in hardware.

When it makes sense: RAID 5 is the most popular choice for 4-bay home and SMB NAS setups. It balances capacity efficiency (75% with 4 drives) against meaningful fault tolerance. For most home users with a 4-bay NAS storing photos, media, and backups, RAID 5 provides the right trade-off. If drives are 12 TB or larger, read the RAID 5 rebuild risk section below before committing.

The RAID 5 Rebuild Risk

RAID 5 has a well-known vulnerability: the rebuild process. When a drive fails and you insert a replacement, the NAS must read every remaining drive in full to reconstruct the missing data. For a 4-bay RAID 5 array with 8 TB drives, that's roughly 24 TB of data read. A process that takes 12-24 hours at sustained HDD read speeds.

During the rebuild, the array is running in a degraded state with no remaining fault tolerance. If any of the remaining drives experiences an Unrecoverable Read Error (URE) during this rebuild process. Which becomes increasingly likely as drives age and as drive capacities increase. The rebuild fails and the array is lost. This is not a theoretical risk: as drives have grown to 8, 12, 16, and 20 TB, the probability of a URE during a RAID 5 rebuild has increased meaningfully.

This risk is why many storage professionals recommend RAID 6 or SHR-2 for arrays using drives 8 TB and larger, or for data that genuinely cannot be lost. For home users with drives under 8 TB and a regular backup schedule, RAID 5 remains practical. Use our RAID Rebuild Time Estimator to calculate expected rebuild time for your specific array.

RAID 6. Striping with Double Parity

RAID 6. Striping with Double ParityDrive 1Drive 2Drive 3Drive 4Data 1Data 2Parity 1Parity 2Data 3Parity 1Parity 2Data 4Parity 1Parity 2Data 5Data 6Parity 2Data 7Data 8Parity 1MIN DRIVES4USABLE SPACE(N−2) ÷ N drivesFAULT TOLERANCE2 drivesBEST FORBusiness / large arrays

RAID 6 works like RAID 5 but with two independent sets of parity information distributed across all drives. This means the array can survive two simultaneous drive failures without data loss. A significant advantage over RAID 5 when rebuilding, since the array now has one failure's worth of tolerance remaining even during a rebuild operation.

Usable capacity: Total raw capacity minus two drives. Four 8 TB drives in RAID 6 = 16 TB usable (not 32 TB). Six 8 TB drives = 32 TB usable. Capacity efficiency improves with more drives. With 6 drives, RAID 6 uses 66% of raw capacity; with 8 drives, 75%.

Minimum drives: 4

Fault tolerance: 2 drives can fail simultaneously without data loss. This also means a RAID 5 rebuild risk is eliminated: if a drive fails during RAID 6 rebuild and the array is degraded, you still have one remaining fault tolerance.

Performance: Read performance is comparable to RAID 5. Write performance is slightly lower than RAID 5 due to double parity calculation. But again, only noticeable on older hardware.

When it makes sense: RAID 6 is the safer choice for 4-bay NAS setups using 8 TB or larger drives, and the recommended level for 6-bay and 8-bay NAS setups storing critical data. If you store irreplaceable files (business documents, RAW photo archives, surveillance footage), the extra parity is worth the 25% capacity cost relative to RAID 5. QNAP's QuTS Hero operating system also offers RAID-TP (Triple Parity), which extends this concept to three simultaneous drive failures. Relevant only for very large arrays.

RAID 10. Mirroring + Striping

RAID 10. Mirroring + StripingMirror Pair AMirror Pair BDrive 1Drive 2Drive 3Drive 4Stripe AMirrorStripe AMirrorStripe BMirrorStripe BMirrorStripe CMirrorStripe CMirrorStripe across pairs for speed1 drive can fail from each mirrored pair. Best balance of speed and protectionMIN DRIVES4 (even number)USABLE SPACE50% of rawFAULT TOLERANCE1 per pairBEST FORPerformance + safety

RAID 10 (sometimes written as RAID 1+0) combines RAID 1 mirroring and RAID 0 striping. Drives are paired into mirrored groups, and those groups are then striped together. The result is both the performance benefit of striping and the redundancy of mirroring.

Usable capacity: 50% of total raw drive space. Four 8 TB drives in RAID 10 = 16 TB usable. The same as RAID 6 with 4 drives, but at better performance and with faster rebuild times.

Minimum drives: 4 (must be an even number)

Fault tolerance: At least 1 drive can fail. Technically, RAID 10 can survive multiple failures as long as they don't both occur in the same mirrored pair. In practice, treat it as tolerating 1 guaranteed failure with additional fault tolerance depending on which drives fail.

Performance: The fastest RAID level that includes redundancy. Reads and writes both benefit from striping across multiple drive pairs, while mirroring ensures reads can be served from either drive in a pair. For workloads that hammer the drives with random read/write operations (databases, VMs, busy multi-user environments), RAID 10's performance advantage over RAID 5/6 is tangible.

When it makes sense: RAID 10 suits workloads where performance is the primary requirement and capacity efficiency is secondary. For home Plex servers and file storage, RAID 5 or RAID 6's better capacity efficiency usually wins. For small businesses running databases, active VMs, or high-transaction workloads on a NAS, RAID 10's performance advantage justifies the 50% capacity cost.

SHR and SHR-2. Synology Hybrid RAID

Synology Hybrid RAID (SHR) is Synology's proprietary RAID implementation, available only on Synology NAS units running DSM. It solves the main practical limitation of standard RAID levels: the requirement that all drives be the same size.

In a traditional RAID 5 array, all drives are treated as the size of the smallest drive in the array. Four drives sized 2 TB, 4 TB, 4 TB, and 4 TB in standard RAID 5 gives you 6 TB usable. The three additional drives all treated as 2 TB to match the smallest. You're wasting 6 TB of raw capacity. SHR instead calculates an optimised layout that maximises usable space from mismatched drive sizes.

How SHR Works

SHR analyses the sizes of all installed drives and creates optimised volume groups that use as much capacity as possible while maintaining fault tolerance. With the 2 TB + 4 TB + 4 TB + 4 TB example above, SHR would give approximately 10 TB usable versus standard RAID 5's 6 TB. A significant efficiency gain from mismatched drives.

The real advantage appears when you upgrade drives over time. If you start with two 4 TB drives and later add a 6 TB drive, SHR accommodates the different size cleanly. Standard RAID 5 would either waste capacity or require all drives to be replaced simultaneously to maintain capacity efficiency. SHR makes incremental drive upgrades practical without capacity waste.

SHR-1: Single-drive fault tolerance (like RAID 5). Minimum 1 drive (single-drive SHR provides no redundancy. Meaningful protection starts with 2 drives). Recommended for most Synology home and prosumer users.

SHR-2: Double-drive fault tolerance (like RAID 6). Minimum 4 drives. Recommended when drives are 8 TB+ or data is critical.

Usable capacity (with matching drives): Same as the standard RAID equivalent. SHR's advantage is specifically for mixed drive sizes.

Should You Use SHR or Standard RAID on a Synology NAS?

For most Synology home and prosumer users, SHR is the right choice. It provides equivalent protection to standard RAID, handles mixed drive sizes gracefully, and simplifies incremental drive upgrades. All without any performance downside. The only scenario where standard RAID is preferable on Synology is if you need to maintain compatibility with non-Synology hardware (SHR arrays can only be read by Synology NAS units) or are operating in an enterprise environment where standardised RAID is required for procedural reasons.

SHR is Synology-only. QNAP, Asustor, TerraMaster, and UGREEN all use standard RAID levels (RAID 1, 5, 6, 10). They don't have an equivalent to SHR. For mixed drive sizes on non-Synology hardware, standard RAID 5 or 6 with uniform drive sizes is the only practical option. Plan to use matching drives from the start.

RAID Level Comparison Table

RAID Level Comparison

RAID 0 RAID 1 RAID 5 RAID 6 RAID 10 SHR-1 SHR-2
Minimum Drives 22344 (even)1 (2 for protection)4
Fault Tolerance None1 drive1 drive2 drives1+ drive (varies)1 drive2 drives
Usable Capacity (4 × 8TB) 32TB8TB24TB16TB16TB24TB16TB
Capacity Efficiency 100%50%75% (4 drives)50% (4 drives)50%75% (4 drives)50% (4 drives)
Read Performance ExcellentGoodGoodGoodExcellentGoodGood
Write Performance ExcellentModerateModerateLowerGoodModerateLower
Rebuild Time N/AFastSlow (large drives)Slow (large drives)FastSlow (large drives)Slow (large drives)
Mixed Drive Sizes LimitedLimitedAll sized to smallestAll sized to smallestAll sized to smallestOptimisedOptimised
Best For Temp/scratch data2-bay NAS4+ bay NAS (home/SMB)4+ bay (critical data)High-performanceSynology 2+ baySynology 4+ bay (critical)

How to Choose the Right RAID Level for Your NAS

2-Bay NAS

Your only practical option with redundancy is RAID 1 (or SHR-1 on Synology). RAID 1 gives you one drive's worth of usable space with protection against either drive failing. This is the standard setup for a 2-bay NAS. Accept the 50% capacity penalty and gain meaningful protection. If you need every byte of space and accept the risk, JBOD or no-RAID is the alternative, but you lose all fault tolerance.

You can run two drives as JBOD (Just a Bunch of Disks) or as individual separate volumes without RAID. This gives you 100% of capacity but zero redundancy. Some users do this intentionally when content is easily replaceable (media that can be re-downloaded) and they're running backups to the cloud anyway. For most users, RAID 1 is the right call on a 2-bay.

4-Bay NAS

This is where the decision gets interesting. With four bays, your realistic options are RAID 5, RAID 6, RAID 10, SHR-1 (Synology), or SHR-2 (Synology). The default recommendation for most home users is RAID 5 or SHR-1. 75% capacity efficiency with one-drive fault tolerance is the right trade-off for photo libraries, media servers, and device backups.

If your drives are 8 TB or larger, or if you're storing data that genuinely cannot be recreated (irreplaceable documents, business records), step up to RAID 6 or SHR-2 and accept the 50% capacity efficiency. The rebuild risk with large drives in RAID 5 is real and has caused data loss for real users.

4-bay models like the Synology DS425+ (from $819), QNAP TS-464 (from~$989), and DS925+ (from~$989) all support RAID 5, 6, and 10. Synology models also support SHR. Choose based on your data criticality, drive size, and whether you plan to mix drive sizes over time.

5-Bay and Larger NAS

As bay count increases, RAID 6 becomes increasingly attractive because capacity efficiency improves. Six drives in RAID 6 uses 66% of capacity; eight drives use 75%. The double-drive fault tolerance of RAID 6 also becomes more valuable as total drive count increases, since the probability of a second failure during a rebuild scales with the number of drives in the array.

For a 5-bay NAS like the Synology DS1525+ or an 8-bay like the DS1825+, RAID 6 or SHR-2 is generally the right default for any data you can't afford to lose. Enterprise NAS setups with 12+ bays often use RAID 6, RAID 60 (striped RAID 6), or dedicated hardware RAID controllers with larger parity sets.

RAID and Your Network Speed

In practice, your network connection is almost always the bottleneck for NAS file transfers. Not your RAID configuration. A Gigabit Ethernet connection maxes out at around 110-115 MB/s. Most RAID 5 arrays on modern Intel/AMD NAS hardware exceed this speed by a significant margin. The performance differences between RAID levels that appear in benchmarks rarely matter for typical home NAS workloads because the network is saturated first.

RAID performance differences become more meaningful when you upgrade to 2.5GbE (max ~280 MB/s) or 10GbE (max ~1,000 MB/s). At those speeds, RAID 10's read performance advantage over RAID 5 starts to show up in real transfers. Particularly for large sequential reads of media files. For users with 10GbE infrastructure and demanding workloads (4K video editing to/from NAS, multi-user concurrent access), RAID 10 may justify its 50% capacity cost on performance grounds.

RAID configuration has no meaningful effect on remote access performance. The bottleneck is always your internet upload speed, not the array. Use the Transfer Speed Estimator to see how connection type, LAN speeds, and drive throughput interact for your specific setup.

What Happens When a Drive Fails

Understanding the failure and rebuild process is essential for making informed RAID decisions. Here's what happens step by step when a drive fails in a redundant RAID array:

1. Detection: Your NAS detects a failed or degraded drive, usually within seconds to minutes. The NAS dashboard shows an alert and. If you've configured email or push notifications. Sends an immediate notification. Audible alarms are also available on most NAS models.

2. Degraded operation: If your RAID level has fault tolerance remaining, the NAS continues operating normally in degraded mode. Data is still accessible; the array is just running without redundancy (RAID 5 with one failure) or with reduced redundancy (RAID 6 with one of two failures). Performance may be slightly reduced during this period.

3. Drive replacement: You physically remove the failed drive and insert a replacement of equal or larger capacity. On most NAS units with hot-swap bays, this requires no shutdown. You pull the failed drive, insert the new one, and the NAS detects it automatically. Desktop NAS units (non-hot-swap) require a brief shutdown to swap drives safely.

4. Rebuild: The NAS automatically begins rebuilding the RAID array using the replacement drive. This process reads all remaining drives and writes reconstructed data to the new drive. Rebuild times depend on array size and drive speed. Expect 8-24 hours for a 4-bay RAID 5 with 8 TB drives. The NAS remains accessible during rebuilding, but performance is reduced and the array remains in degraded mode until completion. Use our RAID Rebuild Estimator to calculate expected rebuild time.

5. Full protection restored: Once the rebuild completes successfully, the array returns to full health with complete fault tolerance restored. The entire process. From drive failure to full protection. Typically takes 12-48 hours depending on array size.

Keep a spare drive. The window between a drive failure and its replacement is when your data is most vulnerable. If you're running RAID 5 with one failure and a second drive fails before you replace the first, your data is gone. Keeping a spare NAS-rated drive of the same model on hand eliminates this gap. You can start the rebuild immediately. Enterprise NAS units support hot spare bays that begin rebuilding automatically without manual drive insertion.

RAID Levels by Brand and Platform

Not every NAS brand supports every RAID level. Here's what each major brand supports on their current operating systems:

RAID LevelSynology (DSM)QNAP (QTS/QuTS)Asustor (ADM)UGREEN (UGOS)
RAID 0YesYesYesYes
RAID 1YesYesYesYes
RAID 5YesYesYesYes (4+ bay)
RAID 6YesYesYesYes (4+ bay)
RAID 10YesYesYesLimited
SHR / SHR-2Yes (exclusive)NoNoNo
RAID-TPNoQuTS Hero onlyNoNo
JBODYesYesYesYes

Synology SHR is the default on DSM and is the recommended starting point for most Synology users. Particularly those who plan to expand storage incrementally with drives of different sizes over time.

QNAP QTS supports a wide range of standard RAID levels. QuTS Hero (available on higher-end QNAP models) adds ZFS-based storage with RAID-Z and RAID-TP, plus inline deduplication and compression. For most home and SMB users on QNAP, standard RAID 5 or RAID 6 in QTS is appropriate. See our Synology vs QNAP comparison for a full software breakdown.

Asustor's ADM supports all the common RAID levels. UGREEN's UGOS is a newer platform with improving RAID support. Check specific model specifications before purchasing if RAID 10 is required, as support varies by model.

Common RAID Mistakes to Avoid

Treating RAID as a backup. This is by far the most common and most costly mistake. RAID protects against physical drive failure. It does not protect against ransomware, accidental deletion, software corruption, or the NAS itself being stolen or destroyed. Maintain a separate offsite backup following the 3-2-1 rule. At minimum, Backblaze B2 via Synology Hyper Backup or QNAP HBS3.

Mixing drive speeds and types in a standard RAID array. Standard RAID treats all drives as the smallest drive's size and slowest drive's speed. A 7,200 RPM drive paired with a 5,400 RPM drive in RAID 1 means both drives write at 5,400 RPM speeds. You've paid for a faster drive but don't benefit from it. SHR handles mixed sizes better than standard RAID, but mixed speeds still apply on any platform.

Using desktop drives instead of NAS-rated drives. NAS-rated drives (IronWolf, WD Red Plus) include TLER (Time-Limited Error Recovery), which prevents a drive from taking too long to recover from a read error. A scenario that can cause RAID controllers to mark the drive as failed and drop it from the array. Desktop drives lack TLER and can cause false RAID degradation events in busy NAS environments. See our best NAS hard drives guide for recommended drives by capacity and budget.

Choosing RAID 0 because you want maximum space. The correct solution for wanting more space is larger drives or more bays. Not RAID 0. RAID 0 doubles your failure risk with zero benefit other than raw performance, which is network-bottlenecked at Gigabit speeds anyway.

Not configuring notifications. A degraded RAID array is invisible unless you're actively checking the NAS dashboard. Configure email or push notifications for drive health events so you're alerted immediately when a drive fails. Not days later when a second failure causes data loss.


🇦🇺 Australian Buyers: What You Need to Know

Warranty and Consumer Rights

Australian Consumer Law (ACL) provides guarantees that go beyond manufacturer warranties. If a NAS drive fails within what a reasonable person would expect it to last. Generally interpreted as longer than the stated warranty for drives in normal use. You may be entitled to a remedy from the retailer (not just the manufacturer). Keep receipts and purchase records for NAS drives. For drives used in always-on NAS environments, ACL protections have been upheld in cases where drives failed within 2-3 years of the 3-year warranty period.

Before purchasing, it's worth asking your retailer: "If this NAS drive fails in a RAID array during normal use, what's your returns process?" Authorised Synology and QNAP resellers (Scorptec, Mwave, Centre Com) generally have clear returns processes. Grey market or private-seller drives may void manufacturer warranty and leave you relying solely on ACL protections, which are harder to enforce. See our where to buy NAS in Australia guide for retailer recommendations.

NBN and Remote Access

On an Australian NBN connection, your remote NAS access speed is capped by your upload speed. Typically 20 Mbps on NBN100, well below even a single hard drive's sequential speed. RAID configuration has essentially no effect on remote access performance. Use the NBN Remote Access Checker to determine whether your connection type supports reliable remote NAS access, and the Transfer Speed Estimator to model upload-limited scenarios.

Related reading: best NAS picks for Australia, 3-2-1 backup strategy, best NAS hard drives.

Related reading: what is a NAS explainer, Synology brand guide, remote access and VPN guide, NAS vs cloud storage, and NAS power consumption guide.

Related reading: our best M.2 SSD for NAS cache and our RAID guide for UGREEN NAS.

Use our free RAID Rebuild Risk Calculator to estimate the probability of data loss during a RAID rebuild.

For a direct comparison with direct-attached storage, see our guide to NAS vs DAS.
What is the best RAID level for a home NAS?
For a 2-bay NAS: RAID 1 (or SHR-1 on Synology). For a 4-bay NAS with drives under 8 TB: RAID 5 or SHR-1. For a 4-bay NAS with drives 8 TB or larger, or storing critical data: RAID 6 or SHR-2. The right choice balances capacity efficiency against your data's criticality and your risk tolerance during rebuild operations.
Can I change my RAID level after setting up the NAS?
It depends on the platform and direction of change. Synology DSM allows migration between some RAID levels non-destructively. For example, expanding from RAID 1 to RAID 5 by adding a third drive. Not all migrations are supported, and some require adding drives. Changing from RAID 5 to RAID 6 on most platforms requires destroying and recreating the array. Always back up your data before attempting any RAID migration. Even supported migrations carry risk.
Does RAID protect against ransomware?
No. Ransomware encrypts files through the operating system. From the RAID array's perspective, it's simply writing new (encrypted) data to the same locations. All drives in the array are updated simultaneously. RAID provides zero protection against ransomware. The correct protection is immutable snapshots (Synology Btrfs snapshots, QNAP QuTS Hero snapshots) and offsite backups that cannot be reached by ransomware running on the NAS.
What happens if my NAS hardware fails. Not a drive, but the NAS itself?
RAID arrays can generally be moved to a compatible NAS unit if the hardware fails. Synology SHR arrays can be moved to another Synology NAS running DSM. Standard RAID 5/6 arrays may be portable across compatible NAS units from the same brand, though this isn't guaranteed. QNAP and Asustor arrays using standard RAID levels sometimes port to compatible hardware. Always verify compatibility before relying on this. A separate backup is essential for this scenario. Hardware failure + non-portable array = data loss.
Is SHR better than RAID 5?
SHR is not technically better or worse than RAID 5 for matching drive sizes. It provides equivalent fault tolerance and capacity efficiency. SHR's advantage is specifically for mixed drive sizes: it allocates capacity more efficiently than standard RAID 5 when drives differ in size. If all your drives are identical, SHR and RAID 5 give the same result. SHR is only available on Synology NAS units running DSM.
How long does a RAID rebuild take?
Rebuild time depends on array size and drive read speed. A rough estimate: 1 hour per 1 TB of data to rebuild. A 4-bay RAID 5 array with three 8 TB drives containing 16 TB of data takes approximately 12-20 hours to rebuild. Larger drives and busier NAS units (serving users during rebuild) take longer. Use the NAS RAID Rebuild Estimator tool for specific estimates. During this time the array is degraded. Back up your data before inserting the replacement drive if critical files aren't otherwise backed up.
Do I need to use the same brand and model of drive in a RAID array?
You don't need to use the exact same model, but using identical or closely matched drives is strongly recommended. Matching brand and model ensures identical firmware (preventing TLER discrepancies), identical MTBF ratings (so drives are likely to be at similar wear stages), and eliminates capacity waste from mismatched sizes in standard RAID. Synology SHR handles mixed capacities well, but even SHR benefits from consistent drive quality. Never mix NAS-rated and desktop drives in the same array.
Can I pull a drive from one NAS and read it in a computer directly?
Generally no, unless the NAS was using a Linux filesystem (ext4, Btrfs, ZFS) that your computer's OS can read, and you know how to mount it. For RAID arrays, you need all drives from the array connected simultaneously through compatible RAID software. Windows cannot natively read ext4 or Btrfs drives. Mac cannot read them without third-party drivers. Single drives formatted as JBOD with ext4 can be read on Linux. For data recovery from a failed NAS, professional data recovery services that specialise in RAID arrays are the recommended path.
Is RAID 6 overkill for a 4-bay home NAS?
Not if your drives are 8 TB or larger. The risk of a second drive failure during a RAID 5 rebuild. Due to an Unrecoverable Read Error on the surviving drives. Increases significantly with drive size and age. For drives 8 TB+, RAID 6 or SHR-2 provides meaningful additional protection. For drives 4 TB and under in a 4-bay, RAID 5 is reasonable provided you have good backups and replace failed drives promptly. The capacity cost of RAID 6 over RAID 5 at 4 drives is 8 TB. One drive. Which at current Australian prices is $200-300.
What is JBOD and should I use it on a NAS?
JBOD (Just a Bunch of Disks) means drives operate as independent volumes with no RAID. No redundancy, no striping. Each drive appears as a separate volume to the operating system. It uses 100% of each drive's capacity with zero overhead. JBOD is appropriate when content is easily replaceable (you can re-download it), when drives are already full and can't form a RAID array, or as a temporary arrangement before a full RAID reconfiguration. For primary NAS storage containing important data, JBOD is not recommended.
Does RAID affect the speed of my Plex or media server?
For typical Plex streaming workloads, RAID level makes negligible difference. Direct-playing a 4K file requires around 60-100 Mbps of drive throughput. Well within the capability of any NAS RAID level on modern hardware. Transcoding is CPU-bound, not drive-bound. The only scenario where RAID matters for Plex speed is simultaneously streaming many 4K files to multiple clients, where RAID 10's read performance advantage over RAID 5 may become perceptible. For most home Plex setups, RAID 5 and RAID 6 are more than sufficient.

Ready to choose a NAS for your RAID setup? Browse our buying guides for the <a href="/blog/best-nas-australia/">best NAS in Australia</a>, choose the right drives with our <a href="/blog/best-nas-hard-drive-australia/">NAS hard drive guide</a>, and calculate your storage needs with our free <a href="/tools/raid-calculator/">RAID Capacity Calculator</a>.

Learn More