How many pools can i have on TrueNAS Core

Nothing for more RAM, however using ECC would be prudent due to the amount of data you have. Regardless of the storage capacity, while you shouldn’t need ECC RAM for a ZFS system, it is one of those safety nets to assist in protecting your data. Recovering that much data takes a lot of time and is not fun. It is just worth noting is all, I’m not trying to tell you that you must make these changes.

I feel your pain as well. I pay for YoutubeTV and the price is crazy high. Almost $90 USD a month. I will be discussing with the wife and cancelling the service. I have more than enough free streaming channels to satisfy my needs, and I also pay for Netflix and Paramount+ which covers all the other crap we watch.

It was good when we had good solid working free rippers (for those of us who owned our media). I use to do that as well, until maintaining a massive library was not financially feasible for me. I figured that there were a handful of movies that I would watch more than a few times, those I would keep closer to the DVD player and I’d just pop in a movie when I really wanted to see it.

The simplest way possible is a proper SAS backplane with expander. You’re very welcome to look for cheaper equivalent to the Supermicro I used as easy to find examples, but do not pretend that your cable spaghetti is “simple”…

You’re gonna turn this thing on and off?

I’m pretty sure a solid, higher power PSU (maybe not even that if the 5A rail is good enough) and a large single case would be cheaper and easier to run, and also take up less space.

And 20TB drives are looking pretty decent on a cost per terabyte, nudging ahead of 16TB drives. When I ran the calculations, a 100TB usable storage stores more than someone can watch in 10 years at an episode or film per day, with no repeats.

Not more often than i have to, but when and if it would be preferable to have all powersupplies sync’ed. Those boards and cables are only a few bucks each.

For me it is as much about keeping everything forever.

Regarding physical media, i stopped buying dvds at around 1200, bluray is at 5100 and 4k at 700.

Not saying ill ever need an 80 disk setup, but knowing it can expand if needed from the get go is calming :wink:

I have no doubt you know your stuff and that it is objectively true :slight_smile:

but if i was going for simplicity, and was prepared to spend that kind of money on the hardware, i could have taken the original suggestion from my friends and gone with $yn0l0gy or equivalent :wink:

For my project think less Linus tech tips and more MacGuyver :slight_smile: :slight_smile:

I’m heading towards the same path I think. My main server case has 20bays along with the backup case and keeping them cool is problematic, along with the aging internals.

I do however have family members who stream off my server, but no one is going to help subsidize the cost of purchasing/running such hardware. On the other hand I have stuff on there you can’t get on streaming services.

But what’s point of using TrueNAS if you cant store zetabytes of data at home :stuck_out_tongue:

1 Like

As your therapist: Consider instead that in 2 years it may make more sense to purchase an additional many-bay chassis, motherboard, etc., and modernize? Technology may be significantly different. Storage density may be different. You might be married and living on a space-yacht.

It is a common trap for smart people, especially software developers and architects, to get caught up in future-proofing.

But you actually don’t know what your needs will be in two years. Simple modules that solve your known needs today are often superior to those that can handle “any” future possibilities.

Mostly I’m saying I’d rather have two or three medium-sized self-contained storage servers instead of one giant one with spaghetti cabling.

Ill settle for a Peta byte first, then we’ll see :smiley:

1 Like

I have owned computers continously since my C64, through A500, A1200, A4000, then pc’s. Ive seen many a media come and go, cassette, floppy, zip, tape backup, cd/dvd burning etc.

The only constant for storage the last 30+ years have been local mechanical harddrives.

SSD will never become economical, ill never trust anyone else to keep my data, so cloud storage is out of the question for other than documents and small sharing.

Sure, i have a bunch of 500GB to 3TB Disks that are taking up more space than they need to, and could be moved to larger disks, but they’re not going anywhere, so as long as i have room for them…

I grew up with VHS and remember all the times i had to overwrite something because i had a limited amount of tapes, only to regret it horribly down the road.

Sure, one can find alot on youtube etc basicallt forever, but the high quality releases are usually dead after a short time, so if i dont get em while theyre fresh, i might miss out.

Since i can afford it, i will never delete anything, ever, and just expand my storage capacity.

I currently have 136 external harddrives full, with about 20 or so of them online divided between my main computers.

So, i think it is safe to say I will be needing mechanical harddrives for storage in 2, 10, probably even 20++ years, and just now making the transition to NAS based storage with redundancy is me keeping up with the newest tech :smiley:

So thanks to everyone, I do believe i got the answers i needed, my plan is plausible, although not the norm…

I will probably buy the 4x SAS cards + what i need for JBOD no 1. when i buy my next 10 drives (when $/TB is acceptable) and ill take it from there i think.

1 Like

If you’re going on with the ghetto plan, make at least your life easier with SAS expanders.
One of these in each enclosure. No motherboard needed, only a 4-pin Molex in the bottom right corner for power, and airflow. One or two external SAS cables to the single HBA in the main unit. And break out cables to your drives, or much preferably internal SAS cables to passive backplanes in a cheap rack JBOD (45drives, Silverstone and the like).

Better wire an expander and a handful of passive backplane strips than 20 individual drives…

(Links are examples to illustrate how cheap expanders are, not endorsments of the particular vendors.)

1 Like

I don’t see you commenting on the questions you got related to the number of pools.
Are you still planning on having 8 distinctly separate pools?

Im not planning on a certain number, but the plan is that every time i expand system ill add 10 drives in raidz2, and create a new pool each time yes.

To me that just seems like an easy way to do it, and is something i immediately was drawn to with my short experience with truenas.

My current organizing scheme is newer drive newer mix of: 4k movies, 1080p movies, 4k tv shows, 1080p tvshows, and ‘watched’ iterations of the 4.

So in plex ill have movies4k1 and tvshows4k1 for pool1, movies4k2 and tvshows2k2 for pool2 and so on…

This prevents rewrites which i hope makes system less strained. Single write till full, then move on to next pool…

i get the idea.

ill look some more into this.

no problems with mixing different manufacturers ? lsi/lenovo/fujitsu etc. ?

did buy a fujitsu sata controller once that di not work on generic pcs, so when i found that lsi were plug n play i just stuck with em… thats also part of this. when i find something that works, ill bend over backwards to make it work in weird ways to not have to experiment with something new :wink:

but the photos and product description on that one was not good. no side or rear pictures, no info on different plugs.

That’s going to mean you will need one separate share per pool, at least, if the plan is to use SMB or NFS shares.

Is there a reason why you don’t want to add additional drives as vdevs to first pool?

So on my windows computer they will show as 1 drive each pool under network.

This first is named array1 10x18tb, next will be array2 10xXXtb.

It seems to me having 8 pools 10 drives each, if one drive fails the rebuild is much simpler and only affects the dives in that pool, as opposed to if i have one pool of 80 drives and one fails… no ?

Also, everytime the pool is expanded with more drives it will rewrite to spread data across new drives. And how would that be with every 1-3 years buing bigger and bigger druves, can yku even mix ?

Pool one is 10x18tb, pool 2 might be 10x20tb. Pool 8 might be 10x36tb :wink:

SAS is SAS. For the HBA you want a LSI (possibly rebranded by Dell/Fujitsu/HP/Lenovo/…) on the appropriate IT firmware to match with the drivers in TrueNAS. For expanders, normally anything goes.

Redundancy is per vdev. If one drive fails, the vdev resilvers.
But if you lose a complete vdev, you lose the whole pool.

ZFS does NOT rebalance data, you’d have the trigger a rewrite.

Again, each vdev is an independent unit. Within each vdev, the smallest drive defines the usable size for all drives in this vdev. ZFS has no issue striping vdevs of different sizes. 10*18+10*20+10\22 in raidz2 is 8*(18+20+22) = 480 TB as you expect.

You can get little sas expander cards to put in your ghetto jbods pretty cheap

Not recommending, just illustrating

https://www.ebay.com.au/itm/387710887349

This one supports 8 lanes out of the chassis for 48gbps or use as a daisy chain… and it then also supports 28 drives.

So, one HBA and one of these. And one cable.

You can power it without a PCIe backplane… directly off the PSU.

1 Like

Seems there are some more things here i dont quite understand yet.

So, my current pool which is set up as one vdev is seen by windows as one ‘disk’, and every consequent pool the same. If i then split that pool into 2 vdevs, will ut still show as 1 disk in windows, or 2 ?

If i have a 8+2 10 disk pool and split into 20 vdevs, then a diskfailure will surely affect several pools, no ?

Am i wrong in thinking that disk failure is the only, or atleast main reason ill ever need to rebuild raid/sliver/parity check ??

Here i am thinking that until a disk fails, ill have no issues at all …

And on vdevs, now that inalready copied some 30tb to the pool set up as 1 vdev, can i still split it into several vdevs without loosing data ? And how will i know which data is stored on each vdev ?