Question Is this a good NAS build?

[misserinjoan]

I'm trying to build my own NAS for video editing. I won't be working with 4k footage - 2k at the most. The NAS will primarily be for archive but I want to have the option to edit off of it, if needed. I'm fairly certain that this will do the job but my biggest concern is the "networking" part. I've never built a NAS so I'm not sure what is needed to connect it to my PC. Let me know if you have any insights.

PCPartPicker Part List: https://pcpartpicker.com/list/3Pg3L2

CPU: AMD Ryzen 3 3200G 3.6 GHz Quad-Core Processor ($99.99 @ Newegg)
Motherboard: Gigabyte B450 AORUS M Micro ATX AM4 Motherboard ($94.99 @ Newegg)
Memory: Corsair Vengeance LPX 16 GB (2 x 8 GB) DDR4-3200 CL16 Memory ($68.98 @ Amazon)
Storage: Seagate Barracuda Compute 4 TB 3.5" 5400RPM Internal Hard Drive ($90.98 @ Amazon)
Storage: Seagate Barracuda Compute 4 TB 3.5" 5400RPM Internal Hard Drive ($90.98 @ Amazon)
Case: Rosewill SCM-01 MicroATX Mini Tower Case
Power Supply: EVGA BQ 500 W 80+ Bronze Certified Semi-modular ATX Power Supply ($63.98 @ Newegg)
Total: $560.90

Thanks in advance!

Best Regards,
Erin Joan

 

[simas]

are you going to be running the video editing on this machine or only using it for storage of the files (the video editing software will run on something else)? right now what you have is not good for either - not enough CPU power to be editing workstation, slow network and VERY slow drives for storage.

if you want storage, get a real NAS box from someone like QNAP or Synology. at least 2 bays ( i.e. 2-4 HDD bays) , low power consumption , full supportability including security patches , app ecosystem around it. i.e. i use and like QNAP 332X for NAS (https://www.qnap.com/en-us/product/ts-332x) since I wanted 10gbe connectivity , SSD caching, 3 drive bays, all in compact form and low power consumption for always on device.

if you want to go 10Gbe , get a cheap ConnectX3 network card for your workstation off ebay (~$30), get a fiber optic cable from either e-bay or FS.com ($20-30) and use it to connect to your NAS for video editing.
if you are ambitious and want to expose your storage to more than one source (i.e. store movie collection there and read it using whatever media consumption device you have attached to your TV) , you can get something like a brocade 6450 switch (4 10Gbe ports + tons of 1Gb ports) for <$100 in e-bay.

if you do not need 10Gbe , then get any other NAS without a SFP+ port, they are <$200 for two bay system. Network will be one of your bottlenecks as you read the drives for video files.

 

[misserinjoan]

are you going to be running the video editing on this machine or only using it for storage of the files (the video editing software will run on something else)? right now what you have is not good for either - not enough CPU power to be editing workstation, slow network and VERY slow drives for storage.

if you want storage, get a real NAS box from someone like QNAP or Synology. at least 2 bays ( i.e. 2-4 HDD bays) , low power consumption , full supportability including security patches , app ecosystem around it. i.e. i use and like QNAP 332X for NAS (https://www.qnap.com/en-us/product/ts-332x) since I wanted 10gbe connectivity , SSD caching, 3 drive bays, all in compact form and low power consumption for always on device.

if you want to go 10Gbe , get a cheap ConnectX3 network card for your workstation off ebay (~$30), get a fiber optic cable from either e-bay or FS.com ($20-30) and use it to connect to your NAS for video editing.
if you are ambitious and want to expose your storage to more than one source (i.e. store movie collection there and read it using whatever media consumption device you have attached to your TV) , you can get something like a brocade 6450 switch (4 10Gbe ports + tons of 1Gb ports) for <$100 in e-bay.

if you do not need 10Gbe , then get any other NAS without a SFP+ port, they are <$200 for two bay system. Network will be one of your bottlenecks as you read the drives for video files.

Thanks for your reply!
I would rather build my own NAS because it'll be cheaper and allow me to upgrade in the future. However, while I know a lot about building gaming PCs, I'm definitely in unfamiliar territory in terms of building a NAS. I won't be using this NAS as an editing workstation but I would like to pull the footage off of the NAS (ie. using the files on the NAS for my projects that will be running on my main PC). From what I've heard of the 10Gbe, it's excessive for my needs. At the end of the day, I really just need a huge brick of storage that can be upgraded to something more versatile in the future. Any ideas?

 

[simas]

Thanks for your reply!
I would rather build my own NAS because it'll be cheaper and allow me to upgrade in the future. However, while I know a lot about building gaming PCs, I'm definitely in unfamiliar territory in terms of building a NAS. I won't be using this NAS as an editing workstation but I would like to pull the footage off of the NAS (ie. using the files on the NAS for my projects that will be running on my main PC). From what I've heard of the 10Gbe, it's excessive for my needs. At the end of the day, I really just need a huge brick of storage that can be upgraded to something more versatile in the future. Any ideas?

I understand your approach and was in the similar thinking years (decades) ago - building my own storage, building windows home server (first edition, etc.), building my own routers (pfsense primarily) , etc. all under the guise of "future proof".
Few problems I run into
- the assumption that my time is worthless (has no value and is free), and unlimited to be involved in these things
- the assumption that support is irrelevant (and therefore things like data recovery , data loss, uptime were all irrelevant)
- the needs for future proofing never truly realized.

instead what I currently believe is
- my time has value and is not 'free'. I am not willing to spend huge amounts of time fighting component compatibility, drivers, software updates, security updates, ZFS settings, etc.
- my data has value and is not worthless I want support. if it breaks I want something better than 'hit the forums' and want actual vendor that can provide it.
- i want a tool for a job. I do not expect microwave to be a fridge or the fridge to be a TV with future proofing potential. I want multiple appliances vs one 'server' that tries to be everything (NAS, server, etc) and does nothing well. I do not expect to build my TV with expectation of future proofing it as a fridge (if I want fridge, I buy a fridge). if you want NAS => buy NAS, not be building servers..

NAS
- storage appliance. not a virtualization hub, not a server to run things that require any amounts of compute
- just enough power to do the job. typically ARM chipset
- low power consumption , very reliable

server(s)
- can be use to run virtual guests (which I do)
- can be used to run domain (which I do run at home)
- backs to NAS(s)
- higher capabilities and higher power consumption

two is one, one is none, three is right for me when it comes to data redundancy. any single device is a single point of failure. i have NAS devices that take backups from server and cross replicate the data i care about (movies, documents, pictures). if one of them dies -fine, replace the NAS (if out of warranty). if both die, restore from NAS backup. if server completely dies, boot off USB, automatically connect and pull the image of the server for restore (barebone restore).

If all you want is storage, i do not know how you can get cheaper than one of the QNAP or Synology devices that retail for $100 a bay (so 2 bay solution is $200), all with OS, security patches, support, low power consumption , etc. Compact, easy to run, etc. Your CPU +MB +Case + RAM would be significantly higher than that. Could the homebuilt do more and be a weird hybrid of NAS and server? sure! but if all you need is storage, who cares?

 

[killster1]

I understand your approach and was in the similar thinking years (decades) ago - building my own storage, building windows home server (first edition, etc.), building my own routers (pfsense primarily) , etc. all under the guise of "future proof".
Few problems I run into
- the assumption that my time is worthless (has no value and is free), and unlimited to be involved in these things
- the assumption that support is irrelevant (and therefore things like data recovery , data loss, uptime were all irrelevant)
- the needs for future proofing never truly realized.

instead what I currently believe is
- my time has value and is not 'free'. I am not willing to spend huge amounts of time fighting component compatibility, drivers, software updates, security updates, ZFS settings, etc.
- my data has value and is not worthless I want support. if it breaks I want something better than 'hit the forums' and want actual vendor that can provide it.
- i want a tool for a job. I do not expect microwave to be a fridge or the fridge to be a TV with future proofing potential. I want multiple appliances vs one 'server' that tries to be everything (NAS, server, etc) and does nothing well. I do not expect to build my TV with expectation of future proofing it as a fridge (if I want fridge, I buy a fridge). if you want NAS => buy NAS, not be building servers..

NAS
- storage appliance. not a virtualization hub, not a server to run things that require any amounts of compute
- just enough power to do the job. typically ARM chipset
- low power consumption , very reliable

server(s)
- can be use to run virtual guests (which I do)
- can be used to run domain (which I do run at home)
- backs to NAS(s)
- higher capabilities and higher power consumption

two is one, one is none, three is right for me when it comes to data redundancy. any single device is a single point of failure. i have NAS devices that take backups from server and cross replicate the data i care about (movies, documents, pictures). if one of them dies -fine, replace the NAS (if out of warranty). if both die, restore from NAS backup. if server completely dies, boot off USB, automatically connect and pull the image of the server for restore (barebone restore).

If all you want is storage, i do not know how you can get cheaper than one of the QNAP or Synology devices that retail for $100 a bay (so 2 bay solution is $200), all with OS, security patches, support, low power consumption , etc. Compact, easy to run, etc. Your CPU +MB +Case + RAM would be significantly higher than that. Could the homebuilt do more and be a weird hybrid of NAS and server? sure! but if all you need is storage, who cares?

the only issue i see is how many years will you get out of your nAs before it fails, how hard will it be to recover the data after it fails. im in the same boat, i have backups of everything just wish to have a more centralized place for them. i have extra computer parts extra cases power supplies extra nic's and sas cards. i would want a 5+ bay nas with 10gbe the price looks like 600-800$ but how many years will it last before it fails.

 

[simas]

the only issue i see is how many years will you get out of your nAs before it fails, how hard will it be to recover the data after it fails. im in the same boat, i have backups of everything just wish to have a more centralized place for them. i have extra computer parts extra cases power supplies extra nic's and sas cards. i would want a 5+ bay nas with 10gbe the price looks like 600-800$ but how many years will it last before it fails.

I bought the first NAS from Synology DS115j 5 years ago (in Synology naming typically DS is "DiskStation", first digit is the number of bays =>single bay, next two digits are year of manufacturer and then a model number) . So DS126 is two bay NAS made in 2016 model year.

zero issues in 5 years.. there are no moving parts there other than hard drive so very little to break. now I have QNAP 332x as primary (3 hard drives + M2 drives for faster storage or caching, 16 GB RAM, 10GBe ). I think it was <400 new. very happy with it. my TV appliance (Nvidia shield) is automatically reading the data from NAS into its Plex library, my Kodi is configured to access NFS share on it ,etc.

I still run a server for things like virtualized pi-hole (great appliance) or unifi controller but server is server and NAS is NAS.

 

[killster1]

I bought the first NAS from Synology DS115j 5 years ago (in Synology naming typically DS is "DiskStation", first digit is the number of bays =>single bay, next two digits are year of manufacturer and then a model number) . So DS126 is two bay NAS made in 2016 model year.

zero issues in 5 years.. there are no moving parts there other than hard drive so very little to break. now I have QNAP 332x as primary (3 hard drives + M2 drives for faster storage or caching, 16 GB RAM, 10GBe ). I think it was <400 new. very happy with it. my TV appliance (Nvidia shield) is automatically reading the data from NAS into its Plex library, my Kodi is configured to access NFS share on it ,etc.

I still run a server for things like virtualized pi-hole (great appliance) or unifi controller but server is server and NAS is NAS.

thats cool that you have had zero issues, maybe ill roll the dice but i still have faint worries of the two things i said, how long will it last and when it does break what other choice do i have other then buying a second box exactly the same and installing the drives to recover the data?

Thousands of QNAP NAS devices have been infected with the QSnatch malware

Over 7,000 infections reported in Germany alone. The malware is still spreading.

www.zdnet.com

GN showing multiple synology failures. (exactly why i dont want to buy a nas a DIY will be simple to replace parts and be back up) the qnap 332x 3 hdd isnt going to cut it for my needs

 

[VirtualLarry]

If all you want is storage, i do not know how you can get cheaper than one of the QNAP or Synology devices that retail for $100 a bay (so 2 bay solution is $200), all with OS, security patches, support, low power consumption , etc. Compact, easy to run, etc. Your CPU +MB +Case + RAM would be significantly higher than that. Could the homebuilt do more and be a weird hybrid of NAS and server? sure! but if all you need is storage, who cares?

I've got an 8-bay Asustor AS6208T sold by "A+ NAS" in my cart @ Amazon right now, for $699.99. It's a bit cheaper than the 8-bay "Lockerstor 8" (forgot actual model number), but that one is $999.99 without drives. I figure that's the cheapest 8-bay NAS on the market right now, and is probably clearing out old stock.

Note that the Lockerstor unit has dual M.2 NVMe drive slots for caching support, as well as dual 10GbE-T and dual 2.5GbE-T as well. So it may be worth it to pay more for the Lockerstor, if those features are valuable to you.

The AS6208T, should be compatible with the Asustor 2.5GbE-T USB3.0 Type-C adapter (with appropriate Type-C to Type-A physical adapter, like $2 on ebay). (I have one of those connected to my 6104T unit, works fine.)

The native connectivity on the 6208T unit is 4x 1GbE-T, which isn't great for single-workstation performance. For an office-type server setting, it would be fine.

 

[VirtualLarry]

That's really interesting about that QSnatch malware, I think that might be (one) of the agents infecting my network. My QNAP NAS unit(s) started acting "wierd", with things like "System processes" taking like 90% CPU time, while the system was idle (mining crypto, cracking NTLM2 hashes of my other PCs?)

Anyways, I wiped my PCs, but unfortunately WHATEVER seems to persist on my network. I had suspected, actually, that my QNAP units were infected, but had no proof.

Article says a "FACTORY RESET" will fix the issue, but how to do that? With button on back, or through GUI? Button on back, will reset ADMIN password, but keep shares, GUI will erase shares, but keep passwords. Is there some sort of "real" factory reset, that requires re-loading the firmware itself? That's what I would want.

I think that the firmware is stored on the drives, with a small bootloader in flash memory on the unit.

 

[simas]

thats cool that you have had zero issues, maybe ill roll the dice but i still have faint worries of the two things i said, how long will it last and when it does break what other choice do i have other then buying a second box exactly the same and installing the drives to recover the data?

I think I have mentioned that I treat these as appliance - custom hardware build for a purpose with fixed function, so the question is how long do you expect any appliance to last? i.e. I know when I buy a screwdriver what I am getting and I know it would likely outlive me.

the second part of your sentence is somewhat confusing to me regarding "buying a second box exactly the same and installing the drives to recover the data", to me this is for people who do not do backup and have to deal with "data recovery" that way since they had a single point of failure. i would never want to be at the mercy of firmware, "same hardware", etc.

in my professional life I use to be a DBA and for anyone doing database administration, "data recovery" is for people who failed completely at keeping proper backup strategy. so in personal hardware I keep, rotate, and test backups of things I care about and would never be interested in buying a same second one of the same, same drives ,etc. instead I will copy back the data (and settings if I care about that) to whatever new hardware I build/buy/bring replacing the item that failed.

i am also not a fan of fancy designs, extensive RAIDs (RAID is not a backup as saying goes) - rather than having a hardware mirrored drive, I would create two volumes on two bay simple NAS, and have data on each volume that is replicated between the volumes. any one of these drives fails => i did not lose the data, volume corruption =>did not lose the data, i accidentally deleted folder =>did not lose the data. having a fancy RAID setup of 10 drives and corrupt volume still makes all those redundant drives useless for information access..

at work, everything I do is RAID/redundant/snapshots. at home I rather have 2+ copies of information and simple, tested, recovery procedures that do not involve worrying about same drives, firmwares, drivers, hardware, architectures. data is data. YMMV

 

[simas]

I've got an 8-bay Asustor AS6208T sold by "A+ NAS" in my cart @ Amazon right now, for $699.99. It's a bit cheaper than the 8-bay "Lockerstor 8" (forgot actual model number), but that one is $999.99 without drives. I figure that's the cheapest 8-bay NAS on the market right now, and is probably clearing out old stock.

Note that the Lockerstor unit has dual M.2 NVMe drive slots for caching support, as well as dual 10GbE-T and dual 2.5GbE-T as well. So it may be worth it to pay more for the Lockerstor, if those features are valuable to you.

The AS6208T, should be compatible with the Asustor 2.5GbE-T USB3.0 Type-C adapter (with appropriate Type-C to Type-A physical adapter, like $2 on ebay). (I have one of those connected to my 6104T unit, works fine.)

The native connectivity on the 6208T unit is 4x 1GbE-T, which isn't great for single-workstation performance. For an office-type server setting, it would be fine.

Kind of curious what and how people are filling 8 bay systems with - looking at currently available HDD sizes of reasonable costs (say 8-12TB per drive), for me personally it is not yet the amount of storage I need. with few hundred GB in pictures/documents and half dozen TB total in media (ripped backups of media I own), higher capacities are just not a need I want to meet yet.. and when I do , I am sure drive capacity would increase , technology would move on , etc.

i also really wanted higher network connectivity to my media and went with SFP+ for it given how crazy cheap recycled hardware is. Brocade 6450-24 switch I think was $85 on e-bay for me, connectx3 10Gbe card was $25, active optical cable was $20. while I have cat6 run in my home where it matters , i ended up running fiber link between my rack and primary workstation area as at that time I just did not see the value in copper 10gbe , too hot, cards 4x expensive, switches much more expensive, higher latency.. so that also influenced my decision to pick the qnap model that had spf+ built in .

 

[killster1]

Kind of curious what and how people are filling 8 bay systems with - looking at currently available HDD sizes of reasonable costs (say 8-12TB per drive), for me personally it is not yet the amount of storage I need. with few hundred GB in pictures/documents and half dozen TB total in media (ripped backups of media I own), higher capacities are just not a need I want to meet yet.. and when I do , I am sure drive capacity would increase , technology would move on , etc.

i also really wanted higher network connectivity to my media and went with SFP+ for it given how crazy cheap recycled hardware is. Brocade 6450-24 switch I think was $85 on e-bay for me, connectx3 10Gbe card was $25, active optical cable was $20. while I have cat6 run in my home where it matters , i ended up running fiber link between my rack and primary workstation area as at that time I just did not see the value in copper 10gbe , too hot, cards 4x expensive, switches much more expensive, higher latency.. so that also influenced my decision to pick the qnap model that had spf+ built in .

yea exactly for you, you seem to have very little storage. If you have your nas fail you can easily be back up and running copying the data over again, now if a 5x12tb drive box the recovery process is to buy another nas add one blank drive and then put the old drives in order to get them back up and running. sure you can format them and recover from backups but im not sure what would be more realistic. Of course this isnt my only backup of the data but pretend the nas is the backup of several different computers it would seem that rebuilding the nas would be easier then hand picking the files and folders that where on there. im all for jbod with one being a copy of the next and having the original files as well, i dont like rebuilding raid arrays and praying another drive doesnt fail during the rebuild process.

I dont expect screwdrivers to last forever but if i buy a very expensive screwdriver perhaps having a bit at the end that could be easily replaced instead of throwing the whole thing out would be a better option.

i expect the nas to work for 4 years, if i built a box id expect id be able to get 10 years out of it with a simple part replacement like psu or motherboard etc. My storage needs atm 4x8tb which are backed up with externals, i could easily use triple that amount with a little effort. I purchased a few 14tb wd dc 530 drives but the reviews show they are not reliable so i might be returning them or using them as backup drives that will never power on unless failure.

 

[simas]

yea exactly for you, you seem to have very little storage. If you have your nas fail you can easily be back up and running copying the data over again, now if a 5x12tb drive box the recovery process is to buy another nas add one blank drive and then put the old drives in order to get them back up and running. sure you can format them and recover from backups but im not sure what would be more realistic. Of course this isnt my only backup of the data but pretend the nas is the backup of several different computers it would seem that rebuilding the nas would be easier then hand picking the files and folders that where on there. im all for jbod with one being a copy of the next and having the original files as well, i dont like rebuilding raid arrays and praying another drive doesnt fail during the rebuild process.

I dont expect screwdrivers to last forever but if i buy a very expensive screwdriver perhaps having a bit at the end that could be easily replaced instead of throwing the whole thing out would be a better option.

i expect the nas to work for 4 years, if i built a box id expect id be able to get 10 years out of it with a simple part replacement like psu or motherboard etc. My storage needs atm 4x8tb which are backed up with externals, i could easily use triple that amount with a little effort. I purchased a few 14tb wd dc 530 drives but the reviews show they are not reliable so i might be returning them or using them as backup drives that will never power on unless failure.

I think we are saying the same thing. few more items
- your storage needs would dictate your solutions (disk size, etc)
- hardware is hardware. same hardware would have same life expectancy , there is nothing magical about DYU that would make it last twice as long as off the shelf NAS hardware (the opposite is possible if person would DYU is really bad at what they do, but unlikely). overall, I expect NAS and DYI hardware wise last the same amount of time for components so it is purely financial calculation. Am I better off swapping a $200 device (automatically getting upgrade in the process) in those 10 years if current device fails or spend some amount of PSU/MB (which is unlikely to cost less than that) should it fail over same 10 years?
- as for hard drives, I am in the same boat as you. bought 4 of the 10TB WD HDDs and had one of them fail already in less than 2 years. not impressed.

 

[killster1]

I think we are saying the same thing. few more items
- your storage needs would dictate your solutions (disk size, etc)
- hardware is hardware. same hardware would have same life expectancy , there is nothing magical about DYU that would make it last twice as long as off the shelf NAS hardware (the opposite is possible if person would DYU is really bad at what they do, but unlikely). overall, I expect NAS and DYI hardware wise last the same amount of time for components so it is purely financial calculation. Am I better off swapping a $200 device (automatically getting upgrade in the process) in those 10 years if current device fails or spend some amount of PSU/MB (which is unlikely to cost less than that) should it fail over same 10 years?
- as for hard drives, I am in the same boat as you. bought 4 of the 10TB WD HDDs and had one of them fail already in less than 2 years. not impressed.

but thats the thing with DIY you can easily replace the part that fails, with the nas you will have the psu hanging off the side when you replace it or pay wayyy more then its worth for the part. was it a dc wd hdd? i have been using 8tb reds and hgst 8tb and have had 0 failures. (i dont even know how many i have in use at least 20 through out my pc's.

 

[simas]

but thats the thing with DIY you can easily replace the part that fails, with the nas you will have the psu hanging off the side when you replace it or pay wayyy more then its worth for the part. was it a dc wd hdd? i have been using 8tb reds and hgst 8tb and have had 0 failures. (i dont even know how many i have in use at least 20 through out my pc's.

are we still talking NAS and not server trying to be a file server , some virtualization beast, or 'I have unlimited free time and not family so I want to learn Solaris in my spare time' device?

if we are talking low power NAS (file share available on the network, secure, supportable, etc). low power (since it is always on), with app ecosystem (if you care about this), DYI is not competitive at all , in price, in on-going costs, in functionality , in supportability (external vendor support is included along with warranty).

A very basic QNAP (say QNAP TS-230-US), is currently $180 new. with support, with on-going security patches, with OS, zero brain installation , etc. It will last as long or longer than DYU build and when it breaks could be replaced by another 2 bay device in the same price range with whatever technology is at that time.
a nice enough , very compact , very portable case is not going to be very cheap. add MB (mITX variety), add CPU, add RAM, I do not know if you are going to hit $180 and this is before accounting for your time which to me is not worthless/free. I have things to do with my life that do not entail screwing around on the forums.

Now, can I cludge something together out of rPi for same/similar costs or even cheaper (not accounting for time)? yes, of cause. but it would be not supportable, worse in every aspect
Can I build a file server that would be significantly higher performance? of cause. but I do need file server, I do not care about transcoding, PLEX, passmark scores, etc. If you want a server ->build a server. if you want NAS -> why build a server?

another thing I do not want to do is hunting for right sockets for CPU that failed, right MB, right memory types ,etc.

again, if you do not care about uptime (you do not have kids screaming because they can not watch Moana they want) or the worth of your time (it is unlimited), by all means go DYU on NAS. I do not have that luxury (I do DYU my workstation/CPU as I pick exactly what I want in components and upgrade at different pace for GPU vs CPU+MB)

YMMV

 

[killster1]

are we still talking NAS and not server trying to be a file server , some virtualization beast, or 'I have unlimited free time and not family so I want to learn Solaris in my spare time' device?

if we are talking low power NAS (file share available on the network, secure, supportable, etc). low power (since it is always on), with app ecosystem (if you care about this), DYI is not competitive at all , in price, in on-going costs, in functionality , in supportability (external vendor support is included along with warranty).

A very basic QNAP (say QNAP TS-230-US), is currently $180 new. with support, with on-going security patches, with OS, zero brain installation , etc. It will last as long or longer than DYU build and when it breaks could be replaced by another 2 bay device in the same price range with whatever technology is at that time.
a nice enough , very compact , very portable case is not going to be very cheap. add MB (mITX variety), add CPU, add RAM, I do not know if you are going to hit $180 and this is before accounting for your time which to me is not worthless/free. I have things to do with my life that do not entail screwing around on the forums.

Now, can I cludge something together out of rPi for same/similar costs or even cheaper (not accounting for time)? yes, of cause. but it would be not supportable, worse in every aspect
Can I build a file server that would be significantly higher performance? of cause. but I do need file server, I do not care about transcoding, PLEX, passmark scores, etc. If you want a server ->build a server. if you want NAS -> why build a server?

another thing I do not want to do is hunting for right sockets for CPU that failed, right MB, right memory types ,etc.

again, if you do not care about uptime (you do not have kids screaming because they can not watch Moana they want) or the worth of your time (it is unlimited), by all means go DYU on NAS. I do not have that luxury (I do DYU my workstation/CPU as I pick exactly what I want in components and upgrade at different pace for GPU vs CPU+MB)

YMMV

already said i dont or have use for a 2 bay nas, the nas i want are 750+$ i have zero issues using ebay to find a new motherboard. i do actually plan on running plex but transcoding shouldn't be a issue. the balance of ease and low power is defiantly a thing. i only play on the forums when im getting paid to do so. I also would be open to just a 5 bay usb c box that just powers the drives.

10GbE Solutions

QNAP provides a range of 10GbE-ready NAS as an affordable and reliable storage solution for 10GbE environments. 10GbE connectivity delivers ten times as much bandwidth as 1GbE, allowing for greater deployment of services and providing an immediate boost to overall network performance.

qnapdirect.com

looks like the 832x 5xbay and 4x2.5" bay for 750$.
or perhaps 2x https://qnapdirect.com/collections/...31kx-4-bay-desktop-nas?variant=32042411622451
for 389$ each.

 

[MalVeauX]

@killster1

I think your original goal to build a NAS is a better approach. While pre-made NAS options are easier and fairly good for a "turn it on, it works" approach, it has the issue of not being very friendly to user troubleshooting and it's holding your data. That's not good in my opinion. These little pre-made NAS commercial options are not power savers either, they generally will soak up 50watts idle or so and a DIY NAS using old and modern computer hardware will do the same (depends on number of drives). So the myth of power saving is just that, a myth. If you want to be able to control the hardware options and replace hardware as you need, then DIY is the way to go. You'll get a way more reliable power supply in a DIY than a commercial NAS product that is "affordable." And you have the option of your DIY NAS running on a platform with a superior file structure and potentially ECC memory if you really want your data integrity to be as good as it can be while being affordable, which is not something you'll find in any affordable commercial NAS that isn't already targeting server communities (enterprise level stuff).

You can get a 10G networking card, install it into whatever platform you want to run, and make sure the client machine also has 10G networking.

Beyond that, it's more about if you want raw storage capacity and speed, or if you want a level of redundancy.

And using a DIY machine you can choose what OS to use and what file structure you wish to use. This may matter a lot to you, if you're using large clumps of data and want room to edit, but also room to archive. And as you mentioned, room to expand. Future proofing isn't an issue, expansion is, when it comes to this, having a platform that allows you to expand SATA connections (SAS controller, etc) and a powersupply that can handle more drives (with connections, avoiding adapters that are often cheap nightmare short circuits waiting to happen) and of course physical locations to place harddrives in your case(s). An affordable commercial NAS will lack all of that pretty much unless you go enterprise level.

And a bigger question becomes, do you want to use HDD's or do you want to string up some 1TB SSD's to work from to real-time use over 10G networking?

Very best,