The oldest rule in ZFS stopped being true, and my NAS upgrades got a lot cheaper

Published Sep 4, 2026, 1:00 PM EDT Jasmine is Software and PC Hardware Author at XDA with years of tech reporting experience ranging from AI chatbots right down to gaming hardware, she's covered just about everything. Whether it's breaking news about the latest AMD NPUs or creating video tutorials on social media platforms, Jasmine has contributed to the world of AI and tech in a variety of ways including interviewing the CEO of Razer, AMD's Director of Product Marketing and the VP of Lenovo. Passionate about gaming and PC technology, she has built countless computers, keyboards and other peripherals - knowing them inside and out. When you start to reach a higher capacity on your NAS drive, you might start considering adding one or two more hard drives in order to gain storage, but this might break the bank. In every ZFS forum, documentation page, and subreddit for the past two decades, the law has been seemingly absolute: you cannot expand an existing RAIDZ VDEV by adding a single disk. Instead, in order to grow your pool, you are forced into either ripping and replacing by buying six identical drives all at once and swapping them one by one over several grueling resilvered days and having to eat the cost or buying a second entire six-drive VDEV at once, demanding an expensive 12- to 24-bay chassis. However, the release of OpenZFS RAIDZ expansion has caused a breakthrough. The stubborn inability to expand RAIDZ vdevs drove thousands of budget-conscious users into the arms of Unraid or Synology's proprietary SHR. With native OpenZFS RAIDZ expansion, you can finally buy a single hard drive on sale, slide it into an empty bay, and grow your resilient enterprise-grade pool incrementally without having to replace the entire thing. The architecture behind RAIDZ The problem has finally been solved RAIDZ has been pretty inflexible for the past 20 years. This is thanks to block striping, as in traditional hardware RAID 5/6: data is striped across drives in fixed physical stripe sizes (for example, 64KB chunks). Expanding traditional RAID requires a simple mathematical restriping pass across uniform blocks. ZFS is both a file system and a volume manager. In ZFS RAIDZ uses dynamic variable stripe widths, meaning every single file or record from 4KB up to 1MB has its own unique priority allocation calculated at the right time. This can lead to an allocation problem because every block in a RAIDZ pool has its data and parity blocks laid out explicitly across the exact number of drives present when it was written. Inserting a new drive into the middle of that geometry can break the fundamental on-disk layout assumptions. RAIDZ expansion has solved the problem thanks to the reflow engine. The engineering behind the feature means that ZFS now performs an online reflow of the existing data. It copies and condenses existing logical stripes onto the newly widened drive geometry, creating a contiguous swath of free space at the end of every physical disk in the vdev. This means the pool stays 100% online and accessible while the expansion runs in the background too. You could be saving thousands of dollars in the long run Expand your storage over time This suddenly meant that upgrading a NAS could be significantly cheaper because you could spread capital expenditure over time. Previously, you had to save up to $1,200 to drop on six drives in a single checkout session to avoid running out of capacity. However, with this new paradigm, you can buy a single 16TB or 20TB drive when it drops to an all-time low on Black Friday or Prime Day for sometimes less than $200. Expand the array and repeat a year later, and you can slowly increase your storage so you never risk running out without footing a pricey bill. You are also no longer trapped thanks to the chassis you own. Users with standard 4-bay or 6-bay consumer enclosures are no longer locked into RAID 10 mirrors just to maintain upgrade flexibility. You can run space-efficient RAIDZ2 and still scale disk by disk. Another benefit of expanding your storage over time is the resilver safety over time too. Instead of subjecting an aging pool to six back-to-back high-stress resilvers as you replace drives one by one, you can run a single controlled expansion pass. There are drawbacks Not everything is perfect Credit: RAIDZ expansion is incredible and gives you the ability to expand your storage pool over time. However, it doesn't perform magic, and there are some downsides too that are worth taking note of. The first is the priority ratio ghost space reality. The refloat process widens the physical geometry, but it does not rewrite or recalculate the data-to-parity ratio of existing blocks. This means if a file was written to a 4-wide RAIDZ2 with 50% priority overhead, it keeps that 50% ratio until it's rewritten. Only newly written files benefit from the improved priority efficiency of the wider 5-wide or 6-wide pool. This means that you might have to rebalance your files in order to actually claim 100% of your mathematical capacity from existing data. You have to rewrite the old files onto the expanded pool. You can do this using a recursive snapshot rewrite script or `zfs rebalance`. While you do get the benefit of being able to resilver slowly, the expansion time and I/O load can be quite tough. Reflowing 40 to 80 TB of data across mechanical drives can take days or sometimes even weeks. During expansion, pool I/O performance will drop noticeably. If a drive fails during the expansion, the process pauses automatically to allow degraded pool repair/resilver before resuming. It might be time to commit Don't let the wall put you off For years, enterprise ZFS storage purity came at the expense of home server practicality and thousands of dollars in your wallet. The inability to add one drive at a time was the single biggest reason DIY builders settled for proprietary solutions or compromised unstriped storage. It feels like OpenZFS has finally caught up to the economic reality of the home builder. If you've been putting off your next port expansion or hesitating to commit to ZFS because of the upgrade war, that war is officially gone.