I don't want to keep my PC on all the time, but I want my media to be available. I also don't want to completely rely on cloud. I also want to share data with my family members, without cloud sync. I also want my IP cameras to record stuff locally without another SaaS subscription.

I'm obviously too conservative.

So I'll always have a NAS.

Maybe there is an alternative to a dedicated NAS via a syncthing machine?

Doesn't rely on cloud

Can share data with family members, no cloud sync needed

IP cameras recording locally is not clear whether it's to a UNC path or to local on device storage. I'm gonna take it as recording over lan to UNC path and say yeah syncthing isn't the best for this but it could sync that path elsewhere.

This is what I do, heterogeneous synced storage through Resilio (but could have chosen Syncthing). One of my servers is even a phone (cracked screen).

I have roughly the same requirements. For me it worked out more time and cost efficient to just leave the PC on. Logistics, management and maintenance were a pain when I had a NAS.

I just split the difference and built my NAS PC with power efficiency in mind. Big fan of the ECC-supporting C3xxx Atoms. I have an eight-core version (which is total overkill here, even) in my NAS with ECC DDR4 and a 8x8TB RAID-Z2: https://www.supermicro.com/en/products/motherboard/A2SDi-8C+...

I'm using a Xeon-D machine, quad cores, with ECC DDR. It's overkill for just storage, but I also run a bunch of VMs on it. It's my router (pfsense), my storage box (samba w/ZFS on SSD array), my ssh utility knife, my "run vendor crapware" (e.g. Lutron's RadioRa software) Windows box, etc. QEMU is fantastic.

My home's IT power draw is about 45 watts, using two access points, that server, and a bunch of little widgets (e.g., Rainforest power monitor, Lutron bridges, etc).

That will go up when I add a big switch, but it's not bad for the utility I get.

The thing is, if you don't want to loose a lot of performance on that storage, either a bought or a built NAS will set you back 1000+ bucks; you can run an idling desktop PC for at least 5 years for that price (80 W idle power consumption, 30 ct/kWh electricity cost).

There are more dimensions to my decision than the raw cost. I sit next to it and enjoy it being relatively silent.

My PC is almost 100% silent.

Did you manage to do that with hard drives?

No which was part of the original point.

Management is a pain? I have a Synology NAS, and in 8 years I can't remember anything I had to do to it beyong initial setup and port-forwarding. You could argue they are underpowered, not cost efficient, or whatever, but technically I had less issues with it than I did with OneDrive.

Your definition of "crazy" huge amount of data is pretty conservative. Your proposed solution would top out at 4TB, while most people I know who spinning up their own RAID at home are looking to store multiple tens of TB. For the cost of the solution you propose (4x 1TB @ $150/ea + $50 NVMe) you could set up an RAID-type solution with 32TB of raw capacity (4x 8TB @ $140/ea + $90 for PCie card and cables).

My photography archive, including about 10 years of high-volume commercial photography, is about 14 years old and at about 23TB. At present, it grows about 5GB a day. I have stopped routinely shooting RAW because storing it all has become a mess.

Most RAID solutions I’ve seen in SOHO environments are 2 bay Qnap etc. Usually top out at 2TiB online. I’d argue anything larger is exceptional. <10% population.

> tens of GB

typo -- tens of TB?

Indeed, fixed. Thanks.

It's still tens of GB, lots and lots of tens of GB.

These things are orthogonal. You can RAID your NVMe storage, or not.

I think you're imagining a dedicated network storage appliance with spinning rust, which is one type of product that might use RAID. But really it's just a computer with some disks in it. You can use RAID in any computer with disks in it.

Hm, I'm not sure i understand.

The 1tb NVMe Sabrent SSD on Amazon is $150: https://www.amazon.com/dp/B07LGF54XR/ref=cm_sw_r_other_apa_i...

You said you could put multiple in the PC so we can assume at least more than 1. So just for the SSDs, not including the NVM PCIe interface we are looking at $300 for 2TB.

In comparison for $100 you can get a 4TB WD red: https://www.amazon.com/dp/B083XVY99B/ref=cm_sw_r_other_apa_i...

That's a lot more $/GB than you could get with HDDs.

I have 4 2TB HDDs in a raid 5 array at home and all 4 disks didn't cost $300 and I have 6TB usable storage space.

So how is it cheaper to go with NVM SDD for home use?

Even this discussion of storage media is entirely orthogonal to RAID. Flash fails too! There is value in redundancy in flash storage applications.

Good point. I think it might be more important with SSDs actually when they fail they typically fail 100%. When HDDs fail you can sometimes get some data off or even use software tools to repair the disk(s) and get all your data back.

There are a variety of real-world SSD failure modes that depend on the specific controller, in addition to environmental factors (NAND doesn't like heat). I would not suggest using this reductive failure model for storage planning.

It’s actually the other way round I have found. I suggest you have a look around at other sources. Rust tends to go clunk at spin up time and that’s it. SSDs soft fail first due to reallocations but are still readable.

Both have very different failure modes depending on what is failing. Controller / electronics failure tends to instantly brick either. Some hard drives experience a single failure event that wipes out a bunch of sectors, but they work fine afterwards and if you overwrite the affected sectors those are back to being fine as well, because they are relocated. For other drives this is ongoing, were more and more of the drive becomes unreadable and even unwritable. I believe both of these are platter surface damage, the latter perhaps R/W head or amplifier degradation. Bearing and motor issues tend to make themselves known in advance, if you listen, but at some point it will stop spinning up or seize while running; instant loss (but a high chance of data recovery, since both media and controller are intact).

For solid state drives a lot of the issues are firmware problems. Things like drives becoming very slow over time have been caused by things like imbalanced internal data structures. Some are media related, e.g. some drives experience slow reads on old writes, because the voltage levels have decayed which necessitates internal retries or slower read speeds. Things like SSDs failing to enumerate are probably controller issues.

In my experience both can fail either slowly, with degraded performance, or fast and total. However, known-good SSDs (most of Samsung, Intel) tend to be very reliable. Good hard drives too.

Purely anecdotal, but I've never had a hard disk just outright fail 100%, they always start going click click click on certain areas of the disk, and avoiding those files you can rescue the rest. SSDs on the other hand, I've never had one properly go into a "read-only" soft fail mode, they just fail to show up one day.

This is all among consumer drives though, enterprise stuff probably behaves better?

I'd love to see some kind of statistics on actual failure modes, but aside from Backblaze, everyone keeps their storage statistics private.

Add a NAS unit to that cost and the recovery time, dire random access time to your estimations and it works out.

I’ve had to bounce around terabytes of crap on spinning rust for years. One nice XFS to XFS copy job took 2 days due to small files. Same thing on SSD was 3 hours.

I’ve also dealt with restoring a hosed 4 TiB NAS from another hard disk array. 4 days recovery sound like fun?

You're still conflating RAID and NAS. These things are orthogonal.

> Unless you need several TiB of contiguous file storage keep them as separate volumes.

Then we have to manually keep track of information such as which volumes contain which files. Isn't that the job of a file system?

It’s not hard to manage it. I managed it before my computers even had hard disks.

Yes, but we shouldn't have to do it. The computer should be tracking this minutiae for us. File systems should be able to automatically use new storage devices in flexible, fault tolerant ways. That way we can keep all our data under one root.

It's normal to have as many file systems as there are storage devices. That's backwards. There should be one file system on top of all devices. This is one reason why ZFS is so good.

And with services like Backblaze offering cheap offsite backup, it's a no-brainer.

My home internet connection is 10:1 upload, but that's still 38Mbps up which is more than enough to backup several TB - sure, the initial backup will take a while, but it then keeps up with incremental changes without any issues.

Backblaze pricing is apparently per computer. Are there any licensing or technical issues with routing backups from all our devices through a single computer like my FreeNAS box?

Exactly that. Only 15mbits up here but that’s fine for incremental backup. Recovery story is better which is the often forgotten bit. Downtime during recovery is a proper pain.