QNAP TS-431K 4-bay NAS $259.00 at Newegg after promo code

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

SamirD

Golden Member
Jun 12, 2019
1,489
276
126
www.huntsvillecarscene.com
Here are some advantages of ZFS solution. First the raid is not locked in your vendor. You can replace the hardware anytime around the disk (that is a big advantage when trying to recover data after your NAS died). Second you can use the actual machine running the raid you don't require a dedicated box so you need to consider that when doing power calculations. Third if you want just the file sever and nothing more you can use low power parts (and modern parts which use less power than 10 year old technology). Last maintenance is not that difficult if you do a few things like label your disk so it is easy to figure out which one died and ZFS has a lot of advantages over older file system by incorporating protection against bit rot. It isn't enough just to be consistent across the raid but actually know which copy is the valid copy. One of the biggest problem with disks is this ability to zero out bad blocks which can create holes in the data. This is something zfs block hash is very good at detecting and fixing.
-
Anyway the choice is the buyer they should just be aware of all the options. To be honest it is #1 that and #2 that well prevent me from never buying a nas box. I don't need the extra hardware and i don't need to worry about trying to reassemble the data if the actual nas box dies.
These are solid points for anyone considering a long-term solution. People forget that if you have a drive, you have to put the drive in something, and if the drive doesn't die but the something does, what do you do? Usually the focus is on drive failure, but no one really thinks about raid controller failure as much as a point of failure, but it truly is.

And this is when being standards based and away from proprietary standards helps. Most nas units are ext3/4 based so you can read the drive using almost any linux live cd. But the exotic synology and qnap proprietary formats can leave you with only one recovery path--theirs.

I treat a nas unit as a multi-point failure device that has no recovery, ie like a drive that failed. If a nas unit fails, assume data gone. So I have multiple nas units from multiple vendors all with the same data replicated. But this doesn't address bit rot, which is damn real.

I first experienced bit rot during yearly comparisons of my photography archive. Just a single file out of hundreds of thousands will not compare correctly with the other two copies of the data, all three copied directly from the source and compared back in that time. As areal densities have increased, the error rate has basically remained the same, and that means bit rot will increasingly become a problem.

ZFS is the only file system I know of that can actively and automatically deal with bit rot. I still don't think I would trust it 100% so I'd have multiple copies of the data, but knowing that this problem is being addressed now will provide a much better solution down the road when the problem is a lot larger too.

And the nice thing about making your own solution is that older essentially free working hardware that is just 'around' can be put to use. I disagree with all the whining over power because no one seems to understand that there are 220/240v 30a circuits on residential ac, ovens, and dryers that use more power than a little server will since the cpu usage is quite little even on a 95w cpu like the i7-2600. The difference in power costs between a 25w nas and a 7w one over the course of a year is the cost of a single meal. And yet no one thinks twice about how many amps cooking at 400F for 35 minutes will cost, or how much it costs to dry clothes--even when each of these appliances are drawing well over 100W during their operation.
 

SamirD

Golden Member
Jun 12, 2019
1,489
276
126
www.huntsvillecarscene.com
This argument was clearly not thought through. :)

I can also take the drives from my Synology - be it RAID, btrfs or SHR (yes, also true for the "locked" Synology storage pool) - and mount them on a Linux machine. QNAP users probably can do the same thing (with ZFS instead of btrfs).
Furthermore, I can take the drives from my Synology NAS and put them into another Synology NAS (not every one - they provide a compatibility chart) and it will be properly recognized.

Moving drives to a NAS from a different vendor NAS or from a DIY server are not guaranteed to work.

In other words: my Synology NAS gives me *more* recovery/migration options, not less, than a DIY alternative. ;)

Well yes, a DIY server is more flexible. But this is a very subjective situation.
Some people keep their NAS in the living room, which probably rules out the option of replacing it with a DIY server (even in a posh mITX case).

Personally, I actually prefer having separate boxes - simply because I can replace and re-purpose the server component whenever I want.
I can turn it off when it isn't needed for a long time, like when I go on vacation (=> power consumption issue later on).
I can also turn the server off for maintenance, change the OS, test stuff etc.
In the meantime, the NAS - which is a crucial element of many workflows for two of us here - remains up and running.

As for the power consumption:
Both my NAS and the server idle at around 5W each (and they are idle for >20h a day). This is important for me because of heat and noise.
I could probably beat this with a DIY setup, but obviously not by a significant margin.

And definitely not if I wanted to run this on a recycled old PC (>40W in idle - unacceptable).

And this is just a long string of stuff I don't have to care about. :)
But once again: most modern NASes offer either ZFS or btrfs.
Synology or any other vendor is not going to give you 'more options' for recovery. That doesn't make business sense. Their proprietary formats are not just going to easily mount in a live cd like a most of my other nas units drives will.

Migration between any nas units is pretty much 10% at best. And there's significant risk doing it. Safe bet? Same as a non-proprietary solution--copy the data again.

I would never keep my data in the living room subject to harm. It stays far away since wired ethernet is everywhere. In fact, it stays replicated in various different locations, even off-site. The data shouldn't live in a place--it should live everywhere you need it and this even starts getting into disaster planning if you're planning it out right.

There definitely is an advantage to keeping things separate so when they do break or need work it doesn't take down the whole operation. The flip side to that is inefficiency since you need separate boxes. This has to be decided on a case by case basis. I think the invention of the NAS as a dedicated storage device versus just sharing a drive on a computer shows that there is a lot of demand for a 'just file serving' device. But as this device has evolved, it's turning back into the device was born from--a full computer running more than just file serving.

As I've already mentioned, power consumption is made into a mountain when it is a mole hill. There are so many more things in a household that the usage of which creates greater swings in power consumption and yet no one even gives them a though compared to the ridiculous amount of cost and time dedicated to bring something down from 40w to 5w. You really see this in homelabbing where people will tell others to throw a server away just because it uses 2x the power even when it costs 50x less or is free. Money is money--whether you are spending it on the hardware or your power bill. The differences is that you can turn hardware off and save power. You can't do anything to recoup the cost of expensive hardware.
 

VirtualLarry

No Lifer
Aug 25, 2001
56,552
10,171
126
I have no idea why anyone even uses exfat, just use fat32 up to 2TB and stop there.
exFAT allows for file sizes > 4GB, which can be important for both disc images, MKV files, and backups. And it's cross-platform, readable on Windows, Linux, and MAC. There's a lot to like about it, compared to regular FAT/FAT32(X).
 

thigobr

Senior member
Sep 4, 2016
234
170
116
exFAT is useful if you need to plug external drives to the NAS for some reason...
For the internal disks I agree there's no point using this file system.
 

piokos

Senior member
Nov 2, 2018
554
206
86
I have older nas units and things have changed dramatically. The very light stable linux base just for file serving has been replaced by a full blown OS that has all the disadvantages of one.
Yes and no.
Yes: NASes changed a lot. They're full of features. But this is what consumers wanted, apparently. Otherwise we would be stick to the simple network sharing idea you refer to. It was easier for NAS makers as well.
I have to say, personally, this is what I want. I wouldn't pay for a basic network sharing device anymore. And BTW: such devices still exist (e.g. Zyxel, D-Link).

Keep in mind that modern, multi-functional NASes are designed for a household with multiple devices (phones, laptops, tablets, TVs, wifi speakers, smart home, IP cameras etc). That's where this functionality shines the most.
If your PC usage at home is built around a single machine, especially a desktop, gains from a NAS are minimal.

It's those extra features that really make paying more for a NAS an easy choice in my case.
Because I use the VPN server, backups, cloud sync, Because I would like to get rid of my desktop. And I would like to have access to things I normally do from at least 2-3 machines.
Some of this is already possible because I moved a lot of things to the cloud. But not everything can be (I mean: I can't afford to, say, edit video on a cloud VM).

And it's just a matter of time before it goes into the territory of too much bloat, etc to the point a new 'lite' version will come out to just be a nas again.
I'm not sure what NASes you used. And how. Even if you buy the cheapest box from Synology (1/2-bay -j series) and don't install extra stuff, it's perfectly smooth.
My point is that saturation of gigabit should be a given in a modern nas that isn't a single drive.
I'm not sure why this is such an important point. 1gbps is actually quite a lot. You need a pretty solid SoC to do that. So the cheapest NASes we have today would not exist. I don't understand why you want to take this choice away from customers...
I'm on the opposite side of the argument. I think the entry-level models from Synology or QNAP are still too expensive. I'd like to see slower models available - making this a fairly cheap and popular options in less wealthy countries.
The emotional aspects are usually left out of that calculation, but are valid point to explore when making a purchase decision.
But this is built around a bad, relative cost measure method.

If you just look at the cash that has to be spend instantly on a particular service or object, you're totally ignoring other costs - that may or may not be easy to quantify.

The most obvious example: is personally building a DIY desktop is better value than an OEM one? On basic performance/price - yes. For vast majority of population - no.

Simple performance/price only makes sense when comparing similar, fully exchangeable product - like processors,, GPUs, car fuel types, CocaCola bottles with different volume etc.
See, that's where a lot of people would disagree with you--the learning may have more value than the nas. This is whole premise behind home labbing where people have mini datacenters in their home that are anywhere from 100x to 10000x more powerful than anyone would ever need in a home.
Learning stuff only has value when it is useful for the person. And it's only rational when it brings better value than the alternative.
So for example, learning how to setup a NAS has different value for someone who works (or plans to work) in IT and for a photographer or a biologist.

For most people who may need a NAS, learning how to build and setup one themselves is just purely wasted time.
Similarly, 25 years ago, as a kid, I made the right decision of learning how to build DIY PCs. It let me save on the few machines I made and make some cash helping others. And I had some fun from it as well.
For kids in 2020 this makes absolutely no sense at all. I would like my future kid to grasp proper electronics and to be a decent programmer (whatever career it chooses). But building PCs from components? I absolutely hope he'll never have to do that.
 

piokos

Senior member
Nov 2, 2018
554
206
86
fat32 and ntfs work fine for this too though, with the file size limit on fat32.
NTFS has way more features (permissions, per-file compression and so on) but it makes it incompatible with many devices. Realistically only Windows supports it out of the box.

Ideally, for an external drive, you want a simple, universally compatible FS - like FAT32, but with size limit that matches modern needs.
That's pretty much how exFAT was born and what it is.
 
  • Like
Reactions: VirtualLarry

SamirD

Golden Member
Jun 12, 2019
1,489
276
126
www.huntsvillecarscene.com
Yes and no.
Yes: NASes changed a lot. They're full of features. But this is what consumers wanted, apparently. Otherwise we would be stick to the simple network sharing idea you refer to. It was easier for NAS makers as well.
I have to say, personally, this is what I want. I wouldn't pay for a basic network sharing device anymore. And BTW: such devices still exist (e.g. Zyxel, D-Link).

Keep in mind that modern, multi-functional NASes are designed for a household with multiple devices (phones, laptops, tablets, TVs, wifi speakers, smart home, IP cameras etc). That's where this functionality shines the most.
If your PC usage at home is built around a single machine, especially a desktop, gains from a NAS are minimal.

It's those extra features that really make paying more for a NAS an easy choice in my case.
Because I use the VPN server, backups, cloud sync, Because I would like to get rid of my desktop. And I would like to have access to things I normally do from at least 2-3 machines.
Some of this is already possible because I moved a lot of things to the cloud. But not everything can be (I mean: I can't afford to, say, edit video on a cloud VM).


I'm not sure what NASes you used. And how. Even if you buy the cheapest box from Synology (1/2-bay -j series) and don't install extra stuff, it's perfectly smooth.

I'm not sure why this is such an important point. 1gbps is actually quite a lot. You need a pretty solid SoC to do that. So the cheapest NASes we have today would not exist. I don't understand why you want to take this choice away from customers...
I'm on the opposite side of the argument. I think the entry-level models from Synology or QNAP are still too expensive. I'd like to see slower models available - making this a fairly cheap and popular options in less wealthy countries.

But this is built around a bad, relative cost measure method.

If you just look at the cash that has to be spend instantly on a particular service or object, you're totally ignoring other costs - that may or may not be easy to quantify.

The most obvious example: is personally building a DIY desktop is better value than an OEM one? On basic performance/price - yes. For vast majority of population - no.

Simple performance/price only makes sense when comparing similar, fully exchangeable product - like processors,, GPUs, car fuel types, CocaCola bottles with different volume etc.

Learning stuff only has value when it is useful for the person. And it's only rational when it brings better value than the alternative.
So for example, learning how to setup a NAS has different value for someone who works (or plans to work) in IT and for a photographer or a biologist.

For most people who may need a NAS, learning how to build and setup one themselves is just purely wasted time.
Similarly, 25 years ago, as a kid, I made the right decision of learning how to build DIY PCs. It let me save on the few machines I made and make some cash helping others. And I had some fun from it as well.
For kids in 2020 this makes absolutely no sense at all. I would like my future kid to grasp proper electronics and to be a decent programmer (whatever career it chooses). But building PCs from components? I absolutely hope he'll never have to do that.
Ah consumers mess it up again...

It's okay though as the pendulum will swing back the other way at some point and we'll have 'classic' nas units again. And then it will push back out again and repeat.

Any nas worth its capability should have no issue dealing with 100 connections so long as the drives and cpu can handle it. Modern homes and all their toys don't really stress any nas--even those from 10 years ago.

I would never used the vpn services on a nas unit if security is truly important. It's like using a butter knife to cut steak--sure you can do it but there are more appropriate tools for the job. I think having an rsync server is a great feature to have, but even my oldest nas units have rsync access via the shell. And rsync is all you need for any type of backup or sync operation imo.

People think they can get rid of their desktop but that's like trying to get rid of pen and paper--you can try, but it will always still be the fastest way to get something done.

I use about 20 machines regularly that connect to a simple file server which in turn are backed up to nas units (the server pre-dates the nas units). You don't need a nas to do this actually--it's just file sharing over a network.

Smooth they are--for now. But as the software continues to demand more, it will demand more from the hardware and then you'll be getting into the areas of compromise. The more 'classic' nas units didn't try to do more, so they didn't need more and work well with what they have, even today.

It's an important point because if a nas from 10 years ago can hit 80MB/sec and can be bought used for $30, why in the world should a $250 unit not do better after 10 years of advancement and development? And this is especially important in light of the fact that the processors are faster and so are the drives.

SoCs have gotten much, much more powerful at the same wattage. To the point that gigabit transfers should be trivial on even the smallest of boxes. The problem is that software bloat--that's what keeps the transfer speeds low, and this is unacceptable imo. 'But features! cries the consumer'--and they're the ones behind the issue.

Less wealthy countries if they're smart use hand me down gear that's dirt cheap from the wealthy countries. It's what happens in datacenters and how homelabs exist using servers like my DL380 G5 that originally MSRPd at $7599 and I got for $65. Same thing for NAS units. I don't think I've paid more than $100 for any of my NAS units, and all of these were a couple of hundred new.

Simple price/performance is the only true metric by which to measure. As soon as you start trying to consider every whim and emotion in this calculation, all objectivity goes out the window. But like I said, a good decision needs to factor in more than just price/performance.

Learning stuff ALWAYS has value because one cannot predict when the material learned will be used. And because of this, one cannot predict what one will 'need' so if you don't know it, you're already too late. Case in point--my dad was a real estate agent right after he finished his masters while he was waiting for his first mechanical engineering job. Once he got the engineering job, he never thought he'd need any of his real estate skills again. Fast forward 20 years and he found that commercial real estate development was fast becoming his track in life and all those skill he picked up as a young man served him every day for the next 20 years of his life.

If learning was only valuable when one would find use for it, schools would not need to exist since just 'learning it when you need it' would be good enough in this world. Obviously, it's not.

If anyone needs a nas and learning anything about one would be 'wasted time', Apple makes products for people like that and charges double for that privilege. ;)
 

SamirD

Golden Member
Jun 12, 2019
1,489
276
126
www.huntsvillecarscene.com
NTFS has way more features (permissions, per-file compression and so on) but it makes it incompatible with many devices. Realistically only Windows supports it out of the box.

Ideally, for an external drive, you want a simple, universally compatible FS - like FAT32, but with size limit that matches modern needs.
That's pretty much how exFAT was born and what it is.
exfat also has compatibility issues due to the fact that it's not supported out of the box on a lot of devices. If you have to go that route, ntfs support is usually better ime.
 

you2

Diamond Member
Apr 2, 2002
5,998
1,144
126
In all honestly 1gbps is piss poor for a NAS. As i noted i frequently get 700+MB (5.5gbps) off my raid using mechanical drives and ssd solutions are pretty close to cost effective for drives 1tb and less. 10gbps interfaces are pretty cheap these days and i would expect most new NAS internal buses being able to handle such load. Can one live with 1gbps well that is up to the person and a matter of application.... so there is no straight answer.

I'm not sure why this is such an important point. 1gbps is actually quite a lot. You need a pretty solid SoC to do that. So the cheapest NASes we have today would not exist. I don't understand why you want to take this choice away from customers...
I'm on the opposite side of the argument. I think the entry-level models from Synology or QNAP are still too expensive. I'd like to see slower models available - making this a fairly cheap and popular options in less wealthy countries.
[\QUOTE]
 
  • Like
Reactions: SamirD

VirtualLarry

No Lifer
Aug 25, 2001
56,552
10,171
126
I think the entry-level models from Synology or QNAP are still too expensive. I'd like to see slower models available - making this a fairly cheap and popular options in less wealthy countries.
I agree. That is the reason that I started my "other" NAS hot deal thread, about the IX2-DL and family. Even though they're EOL, for the price they're selling at, they're actually quite useful, even if they don't quite saturate a 1Gbit/sec LAN connection. (If only they allowed FAN control, I've had that complaint.)
 
  • Like
Reactions: SamirD

VirtualLarry

No Lifer
Aug 25, 2001
56,552
10,171
126
Any nas worth its capability should have no issue dealing with 100 connections so long as the drives and cpu can handle it. Modern homes and all their toys don't really stress any nas--even those from 10 years ago.
You are overlooking the effects of RAM and caching, in that statement.

NAS units with more RAM, can handle more connections at once, and do more RAM caching of the disk structures.

I once ran into a (very entry-level) NAS unit that I was using, that couldn't process my directory structure, because... it didn't have enough RAM to cache the necessary disk structures along that path, to reach the files.

I would say, a NAS would need 4GB or maybe 8GB of RAM to handle 100 "active" disk connections. But only even then, an SSD array could handle that, a HDD array (a standard NAS 4-disk array) just couldn't handle the IOPS at a reasonable rate to perform well, I don't believe. Many entry-level NAS units (QNAP, I'm looking at you!) only have an ARM CPU and 1GB of DDR3(L) RAM, which is NOT upgradable. Thus limiting abilities / performance greatly.

And this isn't even getting into allocation some of the NAS's RAM for a VM to run virtualization.
 
  • Like
Reactions: SamirD

VirtualLarry

No Lifer
Aug 25, 2001
56,552
10,171
126
It's an important point because if a nas from 10 years ago can hit 80MB/sec and can be bought used for $30, why in the world should a $250 unit not do better after 10 years of advancement and development? And this is especially important in light of the fact that the processors are faster and so are the drives.
Two words: "QNAP Bloat". Yes, their current firmware has been getting slower and slower, as they add more features.
 
  • Like
Reactions: SamirD

VirtualLarry

No Lifer
Aug 25, 2001
56,552
10,171
126
In all honestly 1gbps is piss poor for a NAS. As i noted i frequently get 700+MB (5.5gbps) off my raid using mechanical drives and ssd solutions are pretty close to cost effective for drives 1tb and less. 10gbps interfaces are pretty cheap these days and i would expect most new NAS internal buses being able to handle such load. Can one live with 1gbps well that is up to the person and a matter of application.... so there is no straight answer.
I'm thinking of investing in a "Asustor Lockerstor 10". 10-bay, Asustor, with 2x 10GbE-T, 2x 2.5GbE-T, and 2x m.2 NVMe caching slots.

If you don't have $1200, then consider a Lockerstor 4, it's very similar, minus the 10GbE-T, and only 4 bays. It differs from the Nimbustor (prior model with 2.5GbE-T), because like it's larger brethren, but unlike the Nimbustor, it does also have the dual m.2 NVMe slots for SSD caching.

I feel that if done properly (and I assume that it is), the NAS will allow you nearly 10GbE-T (or 2x that, with LAGG) write speeds, to the NVMe array, and then lazy-copy that back to the HDD array. (Much in the same way that an SLC cache works on a QLC NVMe drive like the Rocket Q4.)

Thing is, the NVMe drives that I would use in the NAS, are probably going to be 1 or 2TB due to cost concerns, and when I do BIG file copies, I do BIG file-copies, like 4+ TB at a time, and just let them run overnight.

Although, client backups would probably go lightning-fast, if they were primarily hitting just the NVMe array first, and then lazily copied back to the HDD array.
 
  • Like
Reactions: SamirD

SamirD

Golden Member
Jun 12, 2019
1,489
276
126
www.huntsvillecarscene.com
In all honestly 1gbps is piss poor for a NAS. As i noted i frequently get 700+MB (5.5gbps) off my raid using mechanical drives and ssd solutions are pretty close to cost effective for drives 1tb and less. 10gbps interfaces are pretty cheap these days and i would expect most new NAS internal buses being able to handle such load. Can one live with 1gbps well that is up to the person and a matter of application.... so there is no straight answer.
So when a nas can't even hit 1Gbps, its a bit of shame since my ss4200-e with some 2TB HGST drives can do 80MB/s and that's been discontinued for a decade now, lol.
 

SamirD

Golden Member
Jun 12, 2019
1,489
276
126
www.huntsvillecarscene.com
I agree. That is the reason that I started my "other" NAS hot deal thread, about the IX2-DL and family. Even though they're EOL, for the price they're selling at, they're actually quite useful, even if they don't quite saturate a 1Gbit/sec LAN connection. (If only they allowed FAN control, I've had that complaint.)
Yeah, it's nas units like the IX2-DL that are just fantastic from a price/performance perspective. Mine was $34 shipped from what I recall and it hits 80MB/s just like my older ss4200-e, but has a 'prettier' interface and some more options like an rsync server. For the price it can't be beat even if it can't hit 1Gbps. Of course it might as I haven't tried 2x drives in it yet due to the fan speed issue.
 

SamirD

Golden Member
Jun 12, 2019
1,489
276
126
www.huntsvillecarscene.com
You are overlooking the effects of RAM and caching, in that statement.

NAS units with more RAM, can handle more connections at once, and do more RAM caching of the disk structures.

I once ran into a (very entry-level) NAS unit that I was using, that couldn't process my directory structure, because... it didn't have enough RAM to cache the necessary disk structures along that path, to reach the files.

I would say, a NAS would need 4GB or maybe 8GB of RAM to handle 100 "active" disk connections. But only even then, an SSD array could handle that, a HDD array (a standard NAS 4-disk array) just couldn't handle the IOPS at a reasonable rate to perform well, I don't believe. Many entry-level NAS units (QNAP, I'm looking at you!) only have an ARM CPU and 1GB of DDR3(L) RAM, which is NOT upgradable. Thus limiting abilities / performance greatly.

And this isn't even getting into allocation some of the NAS's RAM for a VM to run virtualization.
Yes, good point about ram caching, although ime it's more valuable for writes than reads. And generally writes are still not at gigabit speeds even if reads are.

Ouch! Those are the early ones that didn't have speeds better than a usb drive connected to a computer, lol. I remember those and passed on them quickly.

Hmmm...if you can tell me an easy way to stress my nas units with a 100 connections, I'll run some tests. The ss4200-e has only 1GB of ram and 4x 2TB hard drives so it would be interesting to see how badly it falls in its face. Plus I've got the Synology, Netgear Readynas, Lenovo ix2, and a Qnap on the way.

Yep, the ram for os and vm is I think where most of the ram goes these days. Hence my reference to bloat.
 

SamirD

Golden Member
Jun 12, 2019
1,489
276
126
www.huntsvillecarscene.com
Two words: "QNAP Bloat". Yes, their current firmware has been getting slower and slower, as they add more features.
See--I knew this would happen! It will be just a matter of time before they start over with another software branch or just keep throwing more hardware at the bad code. Seems to be the way things are done today. :rolleyes:
 

SamirD

Golden Member
Jun 12, 2019
1,489
276
126
www.huntsvillecarscene.com
I'm thinking of investing in a "Asustor Lockerstor 10". 10-bay, Asustor, with 2x 10GbE-T, 2x 2.5GbE-T, and 2x m.2 NVMe caching slots.

If you don't have $1200, then consider a Lockerstor 4, it's very similar, minus the 10GbE-T, and only 4 bays. It differs from the Nimbustor (prior model with 2.5GbE-T), because like it's larger brethren, but unlike the Nimbustor, it does also have the dual m.2 NVMe slots for SSD caching.

I feel that if done properly (and I assume that it is), the NAS will allow you nearly 10GbE-T (or 2x that, with LAGG) write speeds, to the NVMe array, and then lazy-copy that back to the HDD array. (Much in the same way that an SLC cache works on a QLC NVMe drive like the Rocket Q4.)

Thing is, the NVMe drives that I would use in the NAS, are probably going to be 1 or 2TB due to cost concerns, and when I do BIG file copies, I do BIG file-copies, like 4+ TB at a time, and just let them run overnight.

Although, client backups would probably go lightning-fast, if they were primarily hitting just the NVMe array first, and then lazily copied back to the HDD array.
It's interesting you mention the asus--they even have a 2-bay 2.5G model that has dual nvme slots. So performance is even available in some of the smaller/cheaper units.
 

you2

Diamond Member
Apr 2, 2002
5,998
1,144
126
But it begs the issue. If your NAS cost $1200 (no disks); is it not cheaper to just build a linux box and run zfs ?

I'm thinking of investing in a "Asustor Lockerstor 10". 10-bay, Asustor, with 2x 10GbE-T, 2x 2.5GbE-T, and 2x m.2 NVMe caching slots.

If you don't have $1200, then consider a Lockerstor 4, it's very similar, minus the 10GbE-T, and only 4 bays. It differs from the Nimbustor (prior model with 2.5GbE-T), because like it's larger brethren, but unlike the Nimbustor, it does also have the dual m.2 NVMe slots for SSD caching.

I feel that if done properly (and I assume that it is), the NAS will allow you nearly 10GbE-T (or 2x that, with LAGG) write speeds, to the NVMe array, and then lazy-copy that back to the HDD array. (Much in the same way that an SLC cache works on a QLC NVMe drive like the Rocket Q4.)

Thing is, the NVMe drives that I would use in the NAS, are probably going to be 1 or 2TB due to cost concerns, and when I do BIG file copies, I do BIG file-copies, like 4+ TB at a time, and just let them run overnight.

Although, client backups would probably go lightning-fast, if they were primarily hitting just the NVMe array first, and then lazily copied back to the HDD array.
 
  • Like
Reactions: SamirD

piokos

Senior member
Nov 2, 2018
554
206
86
exfat also has compatibility issues due to the fact that it's not supported out of the box on a lot of devices. If you have to go that route, ntfs support is usually better ime.
I'm not saying exFAT works on everything. But the list of compatible systems is fairly long:
- Windows (since XP with updates),
- MacOS (at least for 10 years),
- Linux (since 5.4 or via FUSE),
- NAS systems from Synology, QNAP and Asustor, probably others as well,
- Android: there's an Android driver in case it's not built-in (but it often is)
- iOS: since 2013 (iOS 7),
- ChromeOS: probably from the beginning.

Most importantly, also being a key factor for popularity in PCs, exFAT is the default file system in cameras (and probably other devices that output large files).

No other file system capable of files >4GB comes even close.
 

piokos

Senior member
Nov 2, 2018
554
206
86
But it begs the issue. If your NAS cost $1200 (no disks); is it not cheaper to just build a linux box and run zfs ?
In cash spent: yes. In total cost - for most people: no.

And in case of enterprises there's also the issue of warranty and support.

Hardly anyone outside of desktop gaming and enthusiast niches builds DIY PCs or servers. Why would they suddenly go for a DIY NAS? :D

And even if you're a DIY enthusiast, you'll have to face the simple matter of a day having just 24h.
So even if you're an absolute genius - you can learn how to fix and build almost everything around you (check: Richard Feynman), you'll have to let go at some point and just buy a product made by someone else.

Treat a NAS as a basic appliance - like a fridge. And spend your "hobby time budget" on something more useful and challenging. :)
 

you2

Diamond Member
Apr 2, 2002
5,998
1,144
126
Hum. Never really consider that most people it would be a time sink. For me it is a few hours (3 ish) to build it, maybe 6 hours to configure and maybe 2 hour every 60 days or so for unexpected maint. At least that is how it has been for the past 9 years the system been running.


In cash spent: yes. In total cost - for most people: no.

And in case of enterprises there's also the issue of warranty and support.

Hardly anyone outside of desktop gaming and enthusiast niches builds DIY PCs or servers. Why would they suddenly go for a DIY NAS? :D

And even if you're a DIY enthusiast, you'll have to face the simple matter of a day having just 24h.
So even if you're an absolute genius - you can learn how to fix and build almost everything around you (check: Richard Feynman), you'll have to let go at some point and just buy a product made by someone else.

Treat a NAS as a basic appliance - like a fridge. And spend your "hobby time budget" on something more useful and challenging. :)