ZFS on NAS Explained — When and Why to Use It

ZFS is the gold standard for data integrity on a NAS, but it is not for everyone. It demands more RAM, more capable hardware, and a genuine understanding of what it does before you commit to it. If you have heard ZFS mentioned in NAS discussions and wondered whether you should care, this guide breaks down exactly what ZFS does, which NAS products support it in Australia, and. Critically. When the overhead is justified versus when a simpler file system like ext4 or Btrfs will serve you just as well.

What Is ZFS and Why Does It Exist?

ZFS (Zettabyte File System) was originally developed by Sun Microsystems in 2001 and released as part of OpenSolaris in 2005. It was designed from the ground up to solve a problem that traditional file systems ignore: silent data corruption. Over time, bits stored on hard drives can flip or degrade without the operating system noticing. On a traditional file system like ext4 or NTFS, a corrupted file looks perfectly normal until you try to open it and find it damaged. By which point your backups may also contain the corrupted version.

ZFS takes a fundamentally different approach. It checksums every block of data and every block of metadata, verifying integrity on every read. If a block does not match its checksum, ZFS knows it is corrupted. In a mirrored or RAID-Z configuration, ZFS can automatically repair the corrupted block by copying the correct version from a redundant drive. This is called self-healing, and no other consumer NAS file system offers it natively.

Beyond data integrity, ZFS combines the roles of a file system and a volume manager into one layer. Where traditional setups require a separate RAID controller or software RAID layer underneath the file system, ZFS handles both. This eliminates the so-called "write hole" vulnerability that affects traditional hardware and software RAID configurations. A scenario where a power loss during a write can leave RAID parity data inconsistent. If you want to understand how ZFS compares to ext4 and Btrfs at a technical level, our NAS file systems explained guide covers all three in detail.

Key ZFS Features That Matter on a NAS

Copy-on-Write and Snapshots

ZFS uses a copy-on-write (CoW) model. When data is modified, ZFS writes the new data to a new location on disk rather than overwriting the old data in place. The old data remains intact until the new write is confirmed. This makes ZFS inherently crash-consistent. If power is lost mid-write, the old data is still complete and valid.

Copy-on-write also enables near-instant snapshots. A ZFS snapshot is a point-in-time record of the file system that costs almost no additional storage when first created, because it simply preserves references to existing data blocks. Snapshots only grow in size as the live data diverges from the snapshot. This makes it practical to take frequent snapshots. Hourly or even every few minutes. For rapid rollback in case of accidental deletion, ransomware, or application errors.

End-to-End Checksumming and Self-Healing

Every block of data written to a ZFS pool is checksummed. On every read, ZFS verifies the checksum. If corruption is detected and redundant copies exist (via mirror or RAID-Z), ZFS automatically replaces the corrupt block with a known-good copy. Scheduled scrubs proactively read every block on every drive to find and repair corruption before it causes problems.

This is the single most important reason to choose ZFS. Hard drives develop bad sectors over time. Enterprise studies have shown that silent data corruption occurs more often than most users realise. On ext4 or Btrfs without checksumming of data blocks (Btrfs checksums metadata but not all data configurations), corruption can go undetected for months or years. ZFS catches it immediately.

Inline Compression

ZFS can compress data inline as it is written to disk using algorithms like LZ4 (fast, low overhead) or ZSTD (higher compression ratio, slightly more CPU usage). For many workloads. Especially databases, virtual machine images, log files, and documents. Compression can save 30-50% of disk space while actually improving performance, because less data needs to be read from and written to the physical drives. LZ4 compression is generally recommended as a default because it is fast enough to be essentially free on modern CPUs.

Inline Deduplication

ZFS can identify duplicate data blocks and store only one copy, referencing it multiple times. This is extremely effective in virtualised environments where multiple VMs share the same base operating system image. Deduplication ratios of 3:1 or higher are common. However, deduplication is very RAM-hungry. The deduplication table (DDT) must be kept in RAM for acceptable performance, and a general rule of thumb is 5GB of RAM per 1TB of deduplicated data. For a 20TB pool with deduplication enabled, you would want approximately 100GB of RAM. On a consumer NAS, this makes deduplication impractical for most users. On a QNAP QuTS Hero system with 32GB+ RAM and a specific virtualisation workload, it can be genuinely transformative.

RAID-Z: ZFS-Native Redundancy

ZFS includes its own RAID implementation called RAID-Z, which comes in three levels:

• RAID-Z1. Single parity, tolerates one drive failure (similar to RAID 5)
• RAID-Z2. Double parity, tolerates two simultaneous drive failures (similar to RAID 6)
• RAID-Z3. Triple parity, tolerates three simultaneous drive failures

QNAP’s QuTS Hero also offers RAID-TP (triple parity), which is functionally equivalent to RAID-Z3 and is not available in standard QTS. For a deeper understanding of RAID levels and which one suits your setup, see our NAS RAID explained guide.

The key advantage of RAID-Z over traditional RAID is that ZFS eliminates the write hole problem entirely. Because ZFS is both the file system and the volume manager, it has complete knowledge of every write operation and uses copy-on-write to ensure consistency. A power failure during a RAID-Z write cannot leave parity data in an inconsistent state. A vulnerability that traditional RAID 5 and RAID 6 implementations are susceptible to without a battery-backed cache. To design your vdev layout before committing to a pool configuration, use our ZFS vdev Design Wizard.

ARC: The Adaptive Replacement Cache

ZFS uses a sophisticated read caching algorithm called ARC (Adaptive Replacement Cache) that stores frequently and recently accessed data in RAM. The more RAM you give a ZFS system, the larger the ARC and the faster your read performance. This is why ZFS systems benefit so dramatically from extra RAM. It is not just overhead, it is actively accelerating your storage. On a QNAP QuTS Hero NAS with 16GB or 32GB of RAM, a significant portion of that memory is used by ARC, making hot data access nearly instant compared to reading from spinning drives.

Which NAS Brands Support ZFS in Australia?

Not every NAS supports ZFS. The file system’s hardware requirements. Particularly its appetite for RAM. Mean that only certain NAS platforms offer it. Here is the current state for Australian buyers.

QNAP QuTS Hero

QNAP is the only major pre-built NAS vendor offering ZFS as a first-class option through their QuTS Hero operating system. QuTS Hero runs on many QNAP models with 8GB or more of RAM and provides the full ZFS feature set: self-healing, snapshots, RAID-Z/RAID-TP, inline compression, inline deduplication, WORM (Write Once, Read Many) for compliance, and SnapSync for near-real-time disaster recovery between QuTS Hero units.

QuTS Hero is a separate operating system from QNAP’s standard QTS (which uses ext4). Switching between QTS and QuTS Hero requires a full reinitialisation. All data must be backed up first. The storage pools created by each OS are not cross-compatible. This is a one-way decision for your existing data, so plan accordingly.

TrueNAS (CORE and SCALE)

TrueNAS is the other major ZFS option, and it is the platform most closely associated with ZFS in the NAS world. TrueNAS CORE (FreeBSD-based) and TrueNAS SCALE (Linux-based) are both built entirely around ZFS. It is the only file system they use. TrueNAS is free, open-source software that you install on your own hardware or on iXsystems’ own TrueNAS Mini and TrueNAS enterprise appliances.

For Australian buyers, TrueNAS is a strong option if you are comfortable building or repurposing your own hardware. The trade-off is that TrueNAS lacks the polished, consumer-friendly interface of Synology’s DSM or even QNAP’s QTS. It is a powerful platform, but the learning curve is steeper and there is no local vendor support in Australia. The iXsystems Mini appliances are available internationally but not widely stocked by Australian retailers.

Synology, Asustor, UGREEN, and TerraMaster

Synology uses Btrfs on most of its Plus series and above, and ext4 on its Value series. Synology does not offer ZFS. Btrfs provides some overlapping features. Snapshots, checksumming of metadata, and data scrubbing. But it lacks ZFS’s full end-to-end data checksumming, self-healing on data blocks, and integrated volume management. For most home and SMB users, Btrfs on Synology is excellent and is one of the reasons Synology remains the default recommendation for non-technical buyers.

Asustor, UGREEN, and TerraMaster all use ext4 as their primary file system. None currently offer ZFS. Some TerraMaster models support Btrfs in limited configurations, but ZFS is not available on any of these platforms through their stock operating systems. If you want ZFS on hardware from these brands, you would need to wipe the stock OS and install TrueNAS. Which voids your warranty and is not recommended for most users.

ZFS vs Btrfs vs ext4. A Quick Comparison

For a much deeper dive into this comparison, including benchmarks and real-world scenarios, read our full NAS file systems explained: ext4 vs Btrfs vs ZFS guide.

When You Should Use ZFS on Your NAS

ZFS is not a universal upgrade. It adds genuine value in specific scenarios and unnecessary complexity in others. Here is an honest breakdown of when ZFS is worth the investment and when it is not.

ZFS Makes Sense For

Business-critical storage where data integrity is paramount. If your NAS stores financial records, client data, legal documents, medical records, or any data where silent corruption is unacceptable, ZFS’s end-to-end checksumming and self-healing provide a level of protection that no other NAS file system matches. For Australian businesses with data retention and compliance obligations, QuTS Hero’s WORM (Write Once, Read Many) feature adds immutability that regulatory frameworks may require.

Virtualisation hosts. If you run virtual machines on your NAS using VMware, Hyper-V, or Docker containers with persistent storage, ZFS’s snapshots, compression, and (where RAM allows) deduplication deliver tangible performance and efficiency benefits. Multiple VMs sharing a common base image can achieve significant storage savings through deduplication, and ZFS snapshots allow near-instant rollback of VM state. See our Docker and virtualisation on NAS guide for more on running VMs and containers.

Video production and large file workflows. Video editors, photographers, and creative professionals working with large files benefit from ZFS’s inline compression and robust snapshot capabilities. The self-healing feature is particularly valuable when a single corrupted frame in a multi-terabyte project file can mean hours of re-rendering. QNAP models with Thunderbolt 4 running QuTS Hero are a natural fit for this use case. See our best NAS for video editing guide.

Disaster recovery with SnapSync. QNAP’s SnapSync feature on QuTS Hero enables near-real-time, snapshot-based replication between two QuTS Hero NAS devices. For businesses that need a low-RPO (Recovery Point Objective) disaster recovery solution without the cost of enterprise SAN replication, this is a compelling use case. Pair a primary QuTS Hero NAS in the office with a secondary unit at a remote site or a co-location facility, and SnapSync keeps them synchronised at the block level.

Long-term archival storage. Data that sits on a NAS for years. Archived projects, historical records, compliance data. Is particularly vulnerable to bit rot (silent data corruption over time). ZFS’s scheduled scrubs and self-healing ensure that archived data remains intact even after years of storage. This is the kind of protection you cannot get from ext4 or standard RAID alone.

ZFS Is Overkill For

Basic home NAS use. If your NAS is primarily for media streaming via Plex, Time Machine backups from your Mac, or storing family photos, ZFS adds complexity and cost without proportional benefit. A Synology with Btrfs or a standard QNAP with QTS (ext4) will handle these tasks efficiently with far less RAM. Spending extra on a high-RAM NAS for ZFS when your use case is Netflix and photo backups is money better spent on an extra drive for redundancy or a proper 3-2-1 backup strategy.

Budget NAS setups. ZFS needs RAM. If your budget limits you to a 2-bay NAS with 2-4GB of RAM, ZFS will not perform well and you will not be able to use its most valuable features. Entry-level NAS devices from under $500 are designed around ext4 and Btrfs. Do not try to force ZFS onto hardware that was not built for it.

Users who just want it to work. ZFS requires more understanding to manage properly than ext4 or Btrfs on a consumer NAS. Concepts like vdevs, zpools, scrub schedules, ARC tuning, and recordsize optimisation are straightforward for technical users but overwhelming for someone setting up their first NAS. If simplicity is your priority, Synology’s DSM with Btrfs gives you snapshots and data protection with a fraction of the learning curve.

ZFS Hardware Requirements. What You Actually Need

The most common mistake with ZFS is underestimating the hardware requirements. ZFS can technically run on minimal hardware, but to realise its benefits, you need to give it adequate resources.

The RAM recommendation is not marketing. It is architectural. ZFS uses RAM as its primary read cache (ARC). More RAM means more frequently accessed data lives in memory rather than being read from spinning drives. The difference between 8GB and 32GB of RAM on a ZFS NAS is not marginal; it is transformative for read-heavy workloads. If you are running QuTS Hero on a QNAP NAS, check whether your model supports RAM upgrades and consider maxing it out.

ZFS on QNAP QuTS Hero. The Practical Experience

For most Australian NAS buyers considering ZFS, QNAP’s QuTS Hero is the most accessible path. It provides a full-featured ZFS implementation wrapped in QNAP’s graphical management interface, so you do not need to manage ZFS from the command line (though you can via SSH if you prefer).

What QuTS Hero Gives You Over Standard QTS

Both QTS and QuTS Hero share the same app ecosystem and general interface, but the underlying storage engine is completely different:

• QTS: ext4 file system, standard software RAID (0, 1, 5, 6, 10, 50, 60), Qtier auto-tiering, SSD write caching
• QuTS Hero: ZFS file system, RAID-Z1/Z2/Z3 and RAID-TP (triple parity), self-healing, inline compression and deduplication, WORM, SnapSync, ARC-based read caching

The practical differences that matter most: QuTS Hero’s self-healing catches and repairs silent corruption that QTS cannot detect. QuTS Hero’s RAID-TP tolerates three simultaneous drive failures. Overkill for most home users, but valuable for large arrays in commercial environments where a RAID rebuild under load creates genuine risk of additional drive failure. And QuTS Hero’s ARC makes the NAS feel noticeably faster for read-heavy workloads when you have adequate RAM installed.

Which QNAP Models Suit QuTS Hero?

Any QNAP model that ships with or supports 8GB+ RAM can run QuTS Hero. In practice, the models that make the most sense for ZFS are:

• 4-bay models with 8-16GB RAM. Suitable for small business storage, basic virtualisation, and users who want ZFS data integrity without enterprise cost
• 6- to 8-bay models with 16-32GB RAM. The sweet spot for SMB deployments, video production, and mixed workloads
• Rackmount models with 32-64GB+ RAM. Commercial and enterprise deployments, heavy virtualisation, large storage pools with deduplication

Do not install QuTS Hero on a 2-bay entry model with 4GB of RAM. It will technically work after a RAM upgrade, but the cost of upgrading RAM on a budget NAS to make ZFS functional often exceeds the cost difference to a model that ships with adequate RAM in the first place. For help choosing a QNAP model, see our best QNAP NAS ranked guide.

The QTS to QuTS Hero Switch. Plan Carefully

Switching between QTS and QuTS Hero is possible but destructive. The process reinitialises the NAS, wiping all storage pools and data. You must:

1. Back up all data to an external destination (another NAS, cloud, external drive)
2. Download and flash the QuTS Hero firmware image
3. Reinitialise storage pools under ZFS
4. Restore data from backup

This is not a casual upgrade. Treat it as a migration, not an update. The drives formatted under ext4 (QTS) cannot be read by QuTS Hero’s ZFS, and vice versa. If you are buying a new QNAP NAS and know you want ZFS, install QuTS Hero from the start rather than running QTS first and switching later.

ZFS on TrueNAS. The DIY Option

TrueNAS is the purest ZFS NAS experience available. Both TrueNAS CORE (FreeBSD) and TrueNAS SCALE (Debian Linux) are built from the ground up around ZFS with no alternative file system. If you want maximum control over your ZFS configuration and do not mind building your own hardware, TrueNAS is the platform of choice.

The advantages of TrueNAS over QuTS Hero are depth and flexibility. TrueNAS exposes virtually every ZFS tunable through its web interface or command line. You can configure recordsize per dataset, set up complex pool topologies with mixed vdev types, fine-tune ARC limits, configure SLOG and L2ARC devices, and manage ZFS send/receive replication to remote pools. For users with ZFS experience, TrueNAS is the most powerful option.

The disadvantages are equally clear. TrueNAS has no pre-sales support, no local Australian warranty (unless you buy iXsystems appliances internationally), and a steeper learning curve than any pre-built NAS OS. The app ecosystem is improving with TrueNAS SCALE’s Docker and Kubernetes support, but it is not as polished as Synology’s or QNAP’s app centres. For more detail, see our TrueNAS Australia brand guide.

ZFS Best Practices for NAS Users

If you decide to run ZFS on your NAS, follow these practical guidelines to get the most out of it.

Run regular scrubs. Schedule a scrub at least monthly. Weekly if the NAS is in production use. Scrubs read every block on every drive and verify checksums, catching corruption before it affects your data. Scrubs run in the background and have minimal performance impact on modern hardware.

Enable LZ4 compression. There is almost no reason not to. LZ4 is so fast on modern CPUs that it actually improves performance for compressible data by reducing the amount of I/O to the drives. The CPU overhead is negligible.

Do not enable deduplication unless you know you need it. Deduplication sounds appealing but consumes enormous amounts of RAM. If you enable it without adequate resources, performance degrades severely and the deduplication table may spill from RAM to disk, making the entire pool unusably slow.

Leave 15-20% of pool capacity free. ZFS performance degrades as pools approach full capacity because the copy-on-write mechanism needs free space to write new blocks before releasing old ones. A pool at 90%+ capacity will experience significant write performance degradation.

Use a UPS. ZFS is crash-consistent by design, but a UPS still protects against in-flight writes to drives and protects the drives themselves from power-related damage. This applies to any NAS, but ZFS’s ARC holds recent writes in RAM, and a sudden power loss without a clean shutdown can increase recovery time. See our UPS for NAS guide for recommendations.

ZFS is not a backup. Self-healing, snapshots, and RAID-Z protect against hardware failure and corruption, but they do not protect against fire, theft, ransomware that encrypts your snapshots, or catastrophic multi-drive failure. A proper 3-2-1 backup strategy. Three copies, two different media, one offsite. Remains essential regardless of your file system.

The ECC RAM Debate. Do You Need It?

One of the most debated topics in the ZFS community is whether ECC (Error-Correcting Code) RAM is required. The short answer: ECC is recommended but not mandatory.

ECC RAM detects and corrects single-bit memory errors before they are written to disk. Without ECC, a bit-flip in RAM could be written to the ZFS pool as valid data. ZFS would then faithfully protect that corrupted data with checksums, and a scrub would not detect it as an error because the checksum would match the (corrupted) data. This is the theoretical risk that ECC mitigates.

In practice, the risk of a bit-flip in RAM causing meaningful data corruption is low for consumer workloads. Studies from Google and others have found memory error rates of roughly one bit-flip per 8-16GB of RAM per year. For a home NAS with 16GB of RAM, the probability of a memory error corrupting a critical file is small. Not zero, but small enough that the ZFS co-creator Matt Ahrens has stated that ZFS works fine without ECC and that ECC is similarly beneficial for any file system, not just ZFS.

Most QNAP NAS units use non-ECC RAM and run QuTS Hero without issues. TrueNAS documentation strongly recommends ECC, and iXsystems’ own appliances ship with ECC. If you are building a DIY TrueNAS system for business use, investing in an ECC-capable platform (Intel Xeon, AMD EPYC, or AMD Ryzen Pro) is worth the modest premium. For a pre-built QNAP NAS at home, non-ECC is perfectly fine.

Our Drive Failure Risk Estimator models how vdev size and AFR combine to affect the statistical probability of data loss. A useful complement to understanding ZFS checksumming and resilver behaviour.

Related reading: our NAS buyer's guide.

Can I use ZFS on a Synology NAS?

No. Synology uses Btrfs on its Plus series and above, and ext4 on its Value series. Synology does not offer ZFS on any of its NAS products. If you specifically want ZFS, your options in the pre-built NAS space are QNAP with QuTS Hero, or building your own system with TrueNAS. If Synology’s Btrfs features (snapshots, metadata checksumming, data scrubbing) meet your needs, they may be sufficient without the added complexity of ZFS.

How much RAM do I actually need for ZFS on a NAS?

The commonly quoted figure is 1GB of RAM per 1TB of storage, but this is an oversimplification. The real answer depends on your workload. For a basic file server with 4-20TB of storage, 16GB of RAM provides a good balance of ARC caching and general performance. If you enable deduplication, budget approximately 5GB of RAM per 1TB of deduplicated data. For most home and small business ZFS NAS setups without deduplication, 16GB is the sweet spot. 8GB is the minimum but leaves very little room for ARC, reducing one of ZFS’s key performance advantages.

Is ZFS more reliable than RAID 5 or RAID 6?

ZFS with RAID-Z1 (equivalent to RAID 5) or RAID-Z2 (equivalent to RAID 6) provides the same drive failure tolerance as traditional RAID, but adds data integrity verification that traditional RAID does not. Traditional RAID 5 and RAID 6 have no mechanism to detect silent data corruption. If a block goes bad on a drive, the RAID array has no way to know unless the drive itself reports the error. ZFS checksums every block independently and can detect and repair corruption automatically. ZFS also eliminates the RAID write hole, which is a known vulnerability in traditional RAID 5 and RAID 6 implementations.

Can I expand a ZFS pool by adding drives later?

This is one of ZFS’s historically biggest limitations. Traditional RAID and Synology’s SHR allow you to add a drive to an existing array and expand capacity. ZFS pools are made up of vdevs (virtual devices), and you can add new vdevs to a pool, but you cannot add individual drives to an existing RAID-Z vdev. For example, if you start with a 4-drive RAID-Z1 vdev, you cannot later add a fifth drive to that vdev. You would need to either add a new vdev (such as another set of drives in a mirror or RAID-Z configuration) or replace all drives with larger ones and let the pool resize. OpenZFS 2.3 (released late 2024) introduced RAID-Z vdev expansion, allowing you to add a single drive to an existing RAID-Z vdev. QNAP QuTS Hero and TrueNAS SCALE are expected to support this feature as they adopt newer OpenZFS versions. Check your specific firmware version for availability.

Will ZFS slow down my NAS?

Not if you have adequate RAM. With 16GB or more of RAM, ZFS’s ARC read cache typically makes read-heavy workloads faster than ext4, because frequently accessed data is served from RAM rather than spinning drives. Write performance is comparable to ext4 in most scenarios, and enabling LZ4 compression can actually improve write speeds for compressible data. ZFS will feel slow if you run it on a NAS with only 4-8GB of RAM, because the ARC has insufficient space and the system may use swap. The performance difference between ZFS with adequate RAM and ZFS without it is dramatic.

Should I choose QTS or QuTS Hero for my QNAP NAS?

Choose QTS (ext4) if your NAS is for home use, basic SMB file sharing, media streaming, or backup. And particularly if your NAS has less than 8GB of RAM. Choose QuTS Hero (ZFS) if data integrity is a priority, you plan to run virtualisation workloads, you need compliance features like WORM, or you want SnapSync disaster recovery between two QNAP units. QuTS Hero requires 8GB minimum and ideally 16GB+ to take advantage of ZFS features. Remember that switching between the two requires a complete reinitialisation and data migration, so choose the right one from the start if possible.

Does ZFS protect against ransomware?

ZFS snapshots are one of the best defences against ransomware on a NAS. Because snapshots are read-only, ransomware cannot encrypt or modify them. If ransomware encrypts the live data on your NAS, you can roll back to a pre-infection snapshot and recover your files. However, snapshots are only useful if they exist before the attack. Configure automatic snapshots at regular intervals (hourly or more frequently for critical data) and retain multiple days of snapshots. For maximum protection, combine ZFS snapshots with offsite backup. Ransomware that gains admin access to your NAS could potentially delete snapshots, so an offsite copy remains essential. See our NAS security and ransomware protection guide for a comprehensive approach.