Synology isn't about the NAS hardware and OS. Once setup, it doesn't really matter as long as your config is reliable and fast, so there are many competitive options to move to.
The killer feature for me is the app ecosystem. I have a very old 8-bay Synology NAS and have it setup in just a few clicks to backup my dropbox, my MS365 accounts, my Google business accounts, do redundant backup to external drive, backup important folders to cloud, and it was also doing automated torrent downloads of TV series.
These apps, and more (like family photos, video server, etc), make the NAS a true hub for everything data-related, not just for storing local files.
I can understand Synology going this way, it puts more money in their pocket, and as a customer in professional environment, I'm ok to pay a premium for their approved drives if it gives me an additional level of warranty and (perceived) safety.
But enforcing this accross models used by home or soho users is dumb and will affect the good will of so many like me, who both used to buy Synology for home and were also recommending/purchasing the brand at work.
This is a tech product, don't destroy your tech fanbase.
I would rather Synology kept a list of drives to avoid based on user experience, and offer their Synology-specific drives with a generous warranty for pro environments. Hel, I would be ok with sharing stats about drive performance so they could build a useful database for all.
They way they reduce the performance of their system to penalise non-synology rebranded drives is bascially a slap in the face of their customers. Make it a setting and let the user choose to use the NAS their bought to its full capabilities.
On the other had, they have also slowly destroyed their app ecosystem. The photo solution is much worse than it used to be both in terms of features and the now removed support for media codecs. Video station has pretty much been dead for years.
At this point, I'm not that convinced that there's anything that synology offers that isn't handled much better by an app running on docker. This wasn't true 10 years ago.
The photo apps is great because their photo backup app on android is great, and the only thing that works as well as google photo to ensure all your photo and videos are saved, untouched, no duplicate, no missed media.
That's it. For the actual viewing / sorting / album you need something like immich or photoprism, the photos app actually sucks.
Video station has been removed in the latest minor update, not even a major update, they just took it out no warning no replacement. But then again it was not that good, jellyfin is the way to go for me.
Their crown jewels are active backup, hyper backup and synology office. That's where they own their space.
This is sad... I've been using Synology for a very long time (over 15 years?) and have been pretty happy with my experience. The one time I needed their tech support also left me with a good impression...
This however is a deal breaker for me as I'd hate to be locked in to their drives for all the reasons in TFA but also as a matter of principle.
Yeah, same. I have had three Synology boxes over the last 20 or so years, and they have been super reliable, easy to use, and easy to update. The last one is important to me because I would, over time, add more disks and, when the drive bays were all full, replace smaller disks with larger ones.
The first one I bought is still in service at my parents' place, silently and reliably backing up their cloud files and laptops.
I was fully expecting to buy more in the future, but this is a dealbreaker. If a disk goes bad, I want to go to the local store, pick one up, and have the problem fixed half an hour later. I do not want to figure out where I can get approved disks, what sizes are available, how long it will take to ship them, etc.
I've recently installed Unraid on an old PC, and the experience has been surprisingly good. It's not as nice as a Synology, but it's not difficult, either. It's just a bit more work. I've also heard that HexOS plans to support heterogeneous disks, and I plan to check it out once that is available.
Every six years is enough for apple and other companies who have other sources of revenue and have staked out this high quality niche. But androids, as an example, are more of an average 3 year lifespan if I'm not mistaken, which is closer to what Synology would probably want to achieve but cannot.
The comparison to phones is shaky here. Phones bring substantial performance and feature improvements over 6 years, HW and SW. Synology on the other hand still uses a 5-6 year old CPU and 1Gbps connectivity in their home "plus" line. The OS development is mostly security updates with substantial feature releases few and far between. I expect this from a NAS but it's not at all comparable to a phone.
Forcing their drives is a tax on top of an already existing tax. Synology already charges a premium for lower end specs than the competition. If that's not enough to compensate for the longer upgrade cycles, and they want a hand in every cookie jar it's just going to be a hard pass for me.
I upgraded my Synology box every few years and this is exactly the time I was looking to go to the next model. And I'd pull the trigger and buy a current model before they implement the policy but the problem is now I don't trust that they won't retroactively issue an update that cripples existing models somehow. QNAP or the many alternative HW manufacturers that support an arbitrary OS are starting to be that much more attractive.
I don't think mobile is the right comparison. Those ecosystems are explicitly operating on the assumption that they will profit through the software ecosystem (app store revenue).
Synology seems to have gone entirely the other direction here. Most of their software is given away for free, but the hardware is being monetized.
Additionally - the hardware has different operating constraints. I think the big deal for Synology is that they probably assumed that storage need growth would equate to sales growth.
EX - Synology may have assumed that if I need to store 1TB in 2010, and 5TB in 2015, that would equate to me buying additional NAS hardware.
But often, HDD size increases mean that I can keep the same number of bays and just bump drive size.
Which... is great for me as a user, but bad for Synology (this almost single handedly explains this move, as an aside - I just think it's a bad play).
---
I'd rather they just charged for the software products they're blowing all their money on, or directly tie upgrades to the software products to upgrading hardware.
What are "competitive options"? It's a genuine question. Before Synology, I had some DIY server in a Fractal Design case, and noise and, to be honest, bulk were a problem. Also, maintenance of the server wasn't funny.
I switched to Synology about six years ago (918+). The box is small, quiet, and easy to put in the rack together with the network gear. I started with 4TB drives, gradually switched to 8TB over time (drive by drive). I don't use much of their apps (mostly download station, backup, and their version of Docker to run Syncthing, plus Tailscale). But the box acts like an appliance - I basically don't need to maintain it at all; it just works.
I don't like all this stuff with vendor lock-in, so when the time comes for replacing the box, what are alternatives on par with the experience and quality I currently have with Synology?
The problem is that a lot of competitors don't necessarily have great software. For example QNAP on the hardware side is supposed to be good, you have more bang for the bucks in term of performance but they had several major CVEs that really call into question their security practices. I have a friend who is running Unraid on QNAP and is happy though.
Unraid is brilliant if you're interested in BYO hardware. It can be setup with mix and match drives, supports docker and virtual machines. Realistically it's a bit more work than Synology to get up and running, but once it is, the only thing you really need to do is update the software from time to time
I don't mind the idea of BYO hardware, especially if it's an old server with hotswap drive and hotswap power built in.
Increasingly, with the time I have towards the things that interest me, I just want storage and a bit of compute to be like a home appliance, reasonably set and forget it and leave my messing around on a USFF computer.
I think this is just the tradeoff you need to make. I’m not aware of a solution where you can mix-and-match drives but also get the write performance of a traditional RAID array.
that is true, but you can make one fast pool using zfs and one slower one using unraids disk array, if you want to, or just use the zfs part as a cache for performance
Kind of surprising, I went the other way. I started out with ReadyNAS 15 years ago and after that product faded due to lack of support I no longer wanted to be tied down to a manufacturer. I built a custom solution using a U-Nas chassis. Found FreeNAS back in the day and have stuck with it ever since. Maintenance is fairly minimal.
If you heavily rely on apps/services. I've just gone to self managed docker environments for things like that. A very simple script runs updates.
I over-purchased a NAS and ended up with QNAP, even thought Synology provided more power (lower electricity use) to performance ratio.
In hindsight buying a QNAP that was more than the Synology equivalent felt like a good idea but I didn't really get into it quickly enough.
I also got burned by Western Digital's scandal of selling WD Red drives that really weren't that got them caught in a class action lawsuit. Can't see myself buying them again.
Some Intel N100/N105 board from Aliexpress with Fedora or Debian on top should be fine & much more flexible if you decided you want more than just a file server.
Anecdotally, I quickly gave up on their value-add apps, they didn't seem well thought out and had many missing features. My impression was that they were mostly there to tick all the boxes for their marketing material. It's been a few years since I looked at them so I can't give specific examples unfortunately.
Yes, it is the overall ease of configuration, operation - but also for me the app ecosystem.
Well, my Synology NAS is from... 2013 (have upgraded the drives 3-times), so... it is/was time to replace it, and I can tell you that it won't be with another Synology device...
I won't go back to QNAP, which is what I had before Synology, because during an OS update it wiped all my data (yes, there was a warning, but the whole purpose of having a RAID NAS is safe reliable data storage)
May check-out a custom hardware build, combined with Xpenology.
I have a stack of old harddrives in external enclosures - as I upgrade (non-failed) drives, I buy another enclosure and then do backups that way. So far, over 30-years, it has proven to be more reliable than burned media (I have even had to help clients who had project source code succumb to "bit-rot" on their physical burned media, but I still had a project HD tucked away from 15-years prior)
I’m running two Synology NAS devices, and I wouldn’t consider their app ecosystem to be their strong point. I started by trying to take advantage of the built-in Synology apps when I first got my NAS, but quickly realized how limited they are. Their bi-directional synchronization solution is so slow and archaic compared Syncthing! And the same is true for most of their software offerings. At this point, I’m happy with having Docker support, and don’t particularly care about the rest of their apps.
I still appreciate how easy and maintenance-free was their implementation of the core NAS functionality. I do have a Linux desktop for experiments and playing around with, but I prefer to have all of my actually important data to be on a separate rock solid device. Previously, Synology fulfilled this role and was worth paying for, but if this policy goes live, I wouldn’t consider them fro my next NAS.
I would count supported third-party apps like SyncThing as part of the app ecosystem. You can add the SynoCommunity repository to your Synology and install SyncThing directly, which is pretty nice.
It's a bit more convenient than how other solutions, like Unraid, handle this, where you manually configure a Docker container.
That’s true, but it’s only relevant for the initial setup. I wouldn’t think twice about giving up something so minor compared with the sheer anticompetitive nature of Synology locking down the devices.
Synology did a good job of being relatively turnkey.
QNAP has more configurability for better and worse.
Curious ot hear what other manufactures can compare to them out of the box.
Self-configuring something is a different thing.
I simply do not care any more to rebuild raids and manually swap drives under duress when something is going down. I just replace existing drives with new ones well before they die after they've hit enough years. Backblaze's report is incredibly valuable.
How much of a market is there really for those apps? They are competing against most consumers accepting the ease (and significantly cheaper) of cloud based storage.
We (in the tech space) can scream privacy and risks of the cloud all day long but most consumers seem to just not care.
I have 2 Synology NAS and the only app that I actually use is Synology Drive thanks to the sync app, but there are open source alternatives that would work better and not require a client on the NAS side to work.
I can't imagine any enterprise would be using these features.
Been in the market for a new NAS myself and I am going to be looking into truenas or keep an eye on what Ubiquity is doing in this space (but its a no go until they add the ability to communicate with a UPS).
While true, that assumes that the engineering effort is worth whatever extra market they are getting from it.
I just can't imagine there is that many people that would bother with a "private cloud" that may not already have a use case for a NAS at home for general data storage.
The issue is QNAP has terrible quality/stability at the OS level compared to Synology (also with Apps).
The number of times I’ve broken things on QNAP systems doing what should be normal functionality, only to find out it’s because of some dumb implementation detail is over a dozen. Synology, maybe 1-2.
Roughly the same number of systems/time in use too.
Mind that these are ancient models that are dog slow for anything more than serving files. Not that they are fast in serving files...
I did the procedure on my (now 15yo) TS-410, mostly because the vendored Samba is not compatible with Windows 11 (I had turned-off all secondary services years ago). It took a few days to backup around 8TB of data to external drives. And AROUND 2 WEEKS to restore them (USB2 CPU overhead + RAID5 writes == SLOOOOOW).
Even to get the time down to 2 weeks, I really had to experiment with different modes of copying. My final setup was HDD <-USB3-> RPi4 <-GbE-> TS-410. This relieved TS-410 CPU from the overhead of running the USB stack. I also had to use rsync daemon on TS-410 to avoid the overhead of running rsync over SSH.
So, it's definitely not for the faint of heart, but if you go through the trouble, you can keep the box alive as off-site backup for a few more years.
Having said that, I have to commend QNAP for providing security updates for all this time. The latest firmware update for TS-410 is dated 2024-07-01 [1]. This is really going beyond and above supporting your product when it comes to consumer-level devices.
e.g. QNAP has rare hardware combo of half-depth 1U low-power Arm NAS /w mainline Linux support, 32GB ECC RAM, dual NVME, 4x hotswap SATA, 2x10G SFP, 2x2.5G copper, hardware support for ZFS encryption, https://news.ycombinator.com/item?id=40868855.
In theory, one could fit an Arm RK3588 SBC with NVME-to-PCIe-to-HBA or NVME-to-SATA into half-depth JBOD case. That would give up 2x10G SFP, 2xNVME and ECC RAM.
Maybe it's just me, but rare harware isn't something I'd look for in a reliable storage system unless I had a really special need general hardware just couldn't be made to do
Per sibling comment, "unique" is a better descriptor than "rare". The NAS is made in Taiwan and has been readily available from Amazon or QNAP store.
The Marvell CN913x SoC has been shipping for 5 years, following the predecessor Armada SoC family released 10 years ago and used in multiple consumer NAS products, https://linuxgizmos.com/marvell-lifts-curtain-on-popular-nas.... Mainline Linux support for this SoC has benefited from years of contributions, while Marvell made incremental hardware improvements without losing previous Linux support.
Rare more means a unique combination of common hardware products, where other manufacturers don't put all of the features into one piece of hardware like qnap or others might, to keep people buying more devices to get what they want, or buy a device that is way too overkill for their needs.
I ended up doing that with a larger QNAP I had. It did have some odd bugs that I needed to track down, but otherwise was a good (albeit overly expensive) NAS. I used zfs.
Hacker flexibility or consumer take-it-or-leave-it, pick one.
Debian offers flexibility and control, at the cost of time and effort. PhotoSync mobile apps will reliably sync mobile devices with NAS over standard protocols, including SSH/SFTP. A few mobile apps do work with self-hosted WebDAV and CalDAV. XPenology attempts to support Synology apps on standard Linux, without excluding standard Debian packages.
As far as lists of drives to avoid, Synology could certainly do that, but we also already have Backblaze’s reports on their own failure rates. Synology also uses multiple vendors to produce “Synology” branded drives, so as the article states this may also lead to confusion about which Synology branded drives are “good” vs. “bad” in the future, even with seemingly identical specs.
The idea is not so much about which drives fail or whatever. It’s more that certain consumer drives have firmwares that don’t work well with NAS workloads. Long timeouts could be treated as a failed drive rather than a transient error by a desktop drive, for example.
I’d argue that anyone who is buying a NAS for personal use probably does enough research to figure out that NAS-focused/appropriate drives are a thing-though. And if they contact Synology support, it should be very easy for them to identify bad drive types. On top of that, they can (and have) warn about problematic drives.
Seems to have worked well enough for Backblaze for years and years now. Another major vendor publicly announcing that make X model Y has shitty reliability is as much pressure on the storage duopoly as we're likely to get.
Just do it in reverse: a list of drives that they have tested and can confirm work well; at the end of the list they just mention that they cannot recommend any other.
QNAP's ecosystem is decent. There is a third party store by a former QNAP employee that has a lot more selection in it.
Getting a lower powered intel celeron QNAP nas basically lets you run anything you want software or app wise, including docker that just works instead of hunting for ARM64 binaries for anything that is not available off the shelf.
No it can't. Let's be honest Synology's OS is covering more than just storage, and no, spinning up a lot of 3'rd party docker containers that you need to maintain, secure and manage isn't as easy.
What can't TrueNAS do that was listed in the parent comment?
I'd rather have the flexibility offered by TrueNAS, in addition to the robust community. Yes, Synology hardware is convienent in some use cases, but you can generally build yourself a more powerful and versatile home server with TrueNAS Scale. There is a learning curve, so it is not for everyone.
The killer feature for me is the app ecosystem. I have a very old 8-bay Synology NAS and have it setup in just a few clicks to backup my dropbox, my MS365 accounts, my Google business accounts, do redundant backup to external drive, backup important folders to cloud, and it was also doing automated torrent downloads of TV series.
These apps, and more (like family photos, video server, etc), make the NAS a true hub for everything data-related, not just for storing local files.
I can understand Synology going this way, it puts more money in their pocket, and as a customer in professional environment, I'm ok to pay a premium for their approved drives if it gives me an additional level of warranty and (perceived) safety.
But enforcing this accross models used by home or soho users is dumb and will affect the good will of so many like me, who both used to buy Synology for home and were also recommending/purchasing the brand at work.
This is a tech product, don't destroy your tech fanbase.
I would rather Synology kept a list of drives to avoid based on user experience, and offer their Synology-specific drives with a generous warranty for pro environments. Hel, I would be ok with sharing stats about drive performance so they could build a useful database for all.
They way they reduce the performance of their system to penalise non-synology rebranded drives is bascially a slap in the face of their customers. Make it a setting and let the user choose to use the NAS their bought to its full capabilities.