I know people out there want 2 drive parity, but I think parity sets would be even more useful

 

ie:

  - 8 drives + 1 parity

  - 10 drives + 1 parity

 

Being able to have two (or more) parity groups would achieve a similar result and lower the fear of having a data loss.

 

Also, I desperately need the 25 drive storage. I basically bought this a while back (http://www.newegg.com/Product/Product.aspx?Item=N82E16811219044) am I'm approaching on the 25 drive limit :-(

 

I know people out there want 2 drive parity, but I think parity sets would be even more useful

 

ie:

  - 8 drives + 1 parity

  - 10 drives + 1 parity

 

Being able to have two (or more) parity groups would achieve a similar result and lower the fear of having a data loss.

 

Also, I desperately need the 25 drive storage. I basically bought this a while back (http://www.newegg.com/Product/Product.aspx?Item=N82E16811219044) am I'm approaching on the 25 drive limit :-(

Limetech have already said that dual parity is going to be a feature of the 6.2 release.  I personally would much prefer this to parity sets as it is has other benefits (like identifying which disk causes a parity mismatch). 

well, every solutions has it's upsides.

 

the 6.2 feature with 2 parity drives is desperately needed for users with lot's of drives, no question. and, limetech works on it, so all is set.

but, sometimes it could be also very useful, to have independent pools, each with its parity drive. just only one quick thought brings up the point, spin up groups, or more generally speaking activity of disks.

lets say a scheduled parity check runs, so that means all drives spin up and run. in different scenarios that could mean alot more energy/heat dissipation then (actually for his purpose) necessary; aside noise level increase.

 

so, it's once again probably the point of having the "right" angle to view the need. 

2 parity sets in one box only protects from 2 simultaneous disk failures if they are in different parity sets, it is nowhere near as safe as dual parity across all the disks.

+1 on this feature request.  And apologies for resurrecting such an old thread; I wasn't able to find a more recent one, and wanted to preserve the previous discussion.  (I guess I could have linked back to this one from a new one.)

 

I get that Dual Parity is more fault tolerant than the same drives divided into two Single-Parity sets.  (But perhaps not 3+ Single-Parity sets . . . granted you'll need more drives to move in that direction.)

 

But now that we have support for Dual Parity, it's not an either-or choice.  Dual Parity could be applied to any/all of the multiple parity sets.

 

Besides improved fault tolerance, other benefits include the potential for decreased Parity Sync / Data Rebuild times, on smaller parity sets (those comprised of 3TB drives versus 4TB drives, for example).  Granted a complete Parity Check on all parity sets wouldn't be faster, but rebuilds are when a parity set is more vulnerable to additional drive failures, and reducing the duration of those processes can help avoid a data-loss situation.

 

Actually I can sort of see a situation where even a Parity Check on multiple parity sets could take less time than if they were all one parity set: take one parity set comprised of 3TB drives, the other of 4TB drives.  In the single-parity-set situation, the 3TB drives are slowing the 4TB drives down for the portion of the Parity Check that they're involved in.  In the two-parity-set situation (both Parity Checks run in parallel of course), the 3TB drives can run at their own pace without slowing down the read rate of the 4TB drives, which then complete their check in less time.

 

I realize that lowering Parity Check times isn't a big concern, but lowering Data Rebuild times *is* something to strive for, IMHO.

 

 

Although I would venture to guess that the bulk of unRAID users run small to very small drive arrays, this would be beneficial to those of us with larger drive arrays. With the few of us being the minority, I don't see this as a likely feature in the near term. Notice I said drive array referring to the physical number of disks rather than the total size/TB they have in their array.

 

Now with that being said, I would love to see multiple arrays, each with their own dual parity. It would bring unRAID up to date to where the other NAS OS's are and be basically multiple  RAID6 arrays.

Besides improved fault tolerance, other benefits include the potential for decreased Parity Sync / Data Rebuild times, on smaller parity sets (those comprised of 3TB drives versus 4TB drives, for example).

 

I realize that lowering Parity Check times isn't a big concern, but lowering Data Rebuild times *is* something to strive for, IMHO.

 

As to data rebuild times, I don't see any benefit here, as the time to rebuild a drive is essentially due to the drive size, of the drive being rebuilt, not the parity drive.  And that doesn't change no matter how the array or arrays are set up.  The only other factor would be a small one, the fact that it can't go faster than the slowest drive of its array.

Although I would venture to guess that the bulk of unRAID users run small to very small drive arrays, this would be beneficial to those of us with larger drive arrays. With the few of us being the minority, I don't see this as a likely feature in the near term. Notice I said drive array referring to the physical number of disks rather than the total size/TB they have in their array.

 

Now with that being said, I would love to see multiple arrays, each with their own dual parity. It would bring unRAID up to date to where the other NAS OS's are and be basically multiple  RAID6 arrays.

Yeah, I didn't really expect any action to be taken on this request any time soon.  I just wanted to bring it up again, after Dual Parity was in place, to make sure people didn't think that Dual Parity made multiple parity sets unnecessary (since that seemed to be the outcome of the earlier discussion).

I don't see any real benefit to this.  The concept of splitting your drive sizes into their own "parity sets" (e.g. one set with 3TB drives, one with 4TB drives, etc.) would indeed provide a small benefit in parity check times ... but as already noted several times dual parity provides FAR better fault tolerance than breaking the drives into multiple single-parity sets.

 

If you want multiple sets of disks with their own parity, this is easily achieved by simply building multiple UnRAID servers 

... and with very low power mini-ITX systems this wouldn't necessarily use any more power than a single "monster" server.

(You would also have the ability to make each of these dual parity)

 

 

I don't see any real benefit to this.  The concept of splitting your drive sizes into their own "parity sets" (e.g. one set with 3TB drives, one with 4TB drives, etc.) would indeed provide a small benefit in parity check times ... but as already noted several times dual parity provides FAR better fault tolerance than breaking the drives into multiple single-parity sets.

 

If you want multiple sets of disks with their own parity, this is easily achieved by simply building multiple UnRAID servers 

... and with very low power mini-ITX systems this wouldn't necessarily use any more power than a single "monster" server.

(You would also have the ability to make each of these dual parity)

 

Multiple unRAID servers is what those of us who would like this feature most likely currently do. Atleast for me, it would be nice to be able to have them contained within one unit or one main server. As an example, some of us have jbod chassis that we would prefer to just link together through our SAS expanders rather than have multiple instances of unRAID running. I actually wanted to virtualize unRAID with multiple VM's but I've read and watched videos where unRAID isn't happy when you go to shut one down due to it seeing multiple licenses.

 

Using your suggestion of just running multiple unRAID servers instead of fewer, larger servers.. You would be correct, school systems could buy a bunch of mini vans that set 7 people to get all the kids to school, as it would in fact get the job done as you suggest. But they instead choose to buy a bus that seats 50

 

 

Sent from my iPhone using Tapatalk

...but as already noted several times dual parity provides FAR better fault tolerance than breaking the drives into multiple single-parity sets.

Agreed.  What I'm talking about though is breaking the drives into multiple DUAL-parity sets.

 

If you want multiple sets of disks with their own parity, this is easily achieved by simply building multiple UnRAID servers  ...

And that's what I often do now.  Combining those multiple servers into one saves on a few fronts:

  - unRAID license keys

  - motherboard/RAM/CPU/PSU/chassis overhead (both initial cost and ongoing energy consumption)

 

The latter could range from 10 to 30 watts based on unRAID servers I've built over the years.  More towards 10 watts these days.

 

Admittedly I'd be more likely to use this on my backup servers as opposed to my always-on servers, which I've built to be very energy efficient, so that extra power consumption from splitting one server into multiple servers isn't a big deal.  It's more about the license keys, extra initial cost, and *rack space*; for example, to split this server:

 

https://lime-technology.com/forum/index.php?topic=2031.msg467864#msg467864

 

into two separate servers would likely cause it to need 8 rack spaces instead of the 6 rack spaces it uses as is.  Yes I could go to 1U chassis for the two separate server "heads", but 1U chassis are more difficult to cool as quietly as 2U (or taller) chassis, due to how loud 40mm fans are compared to 80mm fans that move the same amount of air.

 

Agree that if you could have multiple dual-parity sets this would be beneficial for very large arrays.

 

With very large drive counts, I'd be more inclined to use a high-end RAID card that supports multiple arrays (virtually any good card does this) and build multiple RAID-6 sets instead of using UnRAID.  You could even create a few RAID-6 "drives" and then use UnRAID to combine those into a single UnRAID server that used VERY large "drives" if you wanted to use the other UnRAID features (Dockers, VMs, etc.).    You could assign shares to the appropriate "drive" if you wanted to keep them on the same "parity set" (RAID-6 array).  You wouldn't need an UnRAID parity drive, since each of the component "drives" would already be dual fault tolerant.

 

 

My very first unRAID server was structured as you outline, only with RAID5 (not 6) built into the enclosures (not in RAID controllers), and I used a whole other enclosure (total of eight) for the unRAID parity drive:

 

http://lime-technology.com/forum/index.php?topic=27327.msg332114#msg332114

 

But all the arguments for using unRAID instead of RAID can be brought to bear against that kind of drive array structure:  can't mix drive sizes, can't incrementally grow a parity set, will lose all data in a parity set instead of just that on the failed drives when an unrecoverable situation occurs, ...

 

I also like, for my always-on servers, that only the drive I'm reading from needs to spin up (plus the parity drives if writing).  For my backup servers I now use Turbo Write mode.

 

I do have a FreeNAS backup server and it has multiple fault-tolerant arrays.  That's part of where the idea for this suggestion came from.

 

I guess the only real reason I can see for multiple parity sets is if you have an enclosure that can hold more drives than the 25 drives UnRAID can maintain in a single array => in that case it would indeed be nice to be able to set up multiple arrays in the same box.    On the other hand, LimeTech may feel that if you're building a 2nd array you should have a 2nd license 

 

But I do see why you'd like to do that -- the "only one drive spinning" feature is indeed a very attractive part of UnRAID's feature set; and the ability to mix drive sizes is also very nice.

 

I suspect, however, that the number of folks who have setups with more than 25 drives is a pretty small subset of UnRAID users ... so the demand for this is probably rather small.  And large setups like that are probably even less likely in the future due to the very high capacity of modern drives.

 

 

I agree on the multiple arrays = multiple licenses; I suspect there would be licensing changes if multiple parity sets were added as an unRAID feature.

 

By the way, I had heard that it's up to 30 drives now.  28 data + 2 parity, with cache drives cutting into that total.

 

The most I've run is 24 drives (not counting my first build with each 'drive' being a RAID5 array), in the 88TB machine in my signature.  And it has two complete backups of my primary server, another source of this suggestion.

 

And definitely agree on the small demand point.  So I don't imagine this would be added soon (if at all).

I actually agree with most of what you said there. To the point of LT feeling if your setting up a second array that you should have a second license. I would even agree with that. So on that note, being able to add a second license to that install thus allowing a second array to be ran would be a perfectly workable solution for me and to be honest, a completely fair solution. I own multiple pro licenses so I'd be all over that if a feature allowing multiple licenses was to be implemented. It would completely solve my current dilemma.

 

As to drive capacities getting larger.. For me this will be offset by the media I store also getting larger. So when I started ripping my collection it was all DVD, then drives for larger and now I only rip Blu Rays. UHD encryption will eventually be broken and those of us that rip our own media will start ripping those. At 100GB a rip (from what I've read) it will end up being the same proportion I have now as far as putting a 25GB rip onto a 2TB drive vs a 100GB rip onto a 8TB drive. So this issue could self correct for the near term but those of us with large disk arrays will be in the same boat in a few years time yet again even with all 8TB or 10TB drives.

 

I currently run (96) 2TB drives in my arrays in (4) SM846 chassis. Half are WD Reds and the other half (backup drives) are WD RE4's that for demoted to backup duty when I started buying the WD Reds. Anyone reading this question why I'm running 2TB drives. Keep in mind that when I started building media servers 2TB drives were these insanely huge drives and 4TB, 6TB, 8TB... drives were not even on the market. Now when it came time to expand since I already had the RE4's naturally I started replacing them with much newer, lower power and cooler running Reds.

 

So if I had to do it all over again and got to start with all that money in my pocket would I go with 2TB Reds? Of course not. But to start over now and sell off all my drives at used eBay prices only to turn around and buy the same capacity worth of 8TB drives would cost me an extra $3,000 on top of what I would get for my used drives to yield the same capacity. As drives start to die, I will replace them with 8TB or 10TB drives.

 

 

Sent from my iPhone using Tapatalk

I should also add that my main driver behind wanting this feature or multiple licenses on the same machine for multiple arrays is I couldn't help myself with the dirt cheap 60 bay SAS3 JBOD HGST chassis on eBay recently. They spoke to me and I had to have them would love to be able to run unRAID on them rather than switching everything back over to Windows and Drivepool.

 

 

Sent from my iPhone using Tapatalk

Understand how you ended up with so many 2TB drives -- I still have quite a few myself.  In fact, I don't have a single 6TB or 8TB drive in any of my 4 UnRAID servers (although I have several in my other computers).    My largest UnRAID server is all 4TB Reds ... it backs up my 3 other servers ... and it only has 18 drives at the moment (soon to be 19 when I add another parity drive).    My main non-media server is a PC-Q25B with 6 3TB Reds; and my media box has 16 drives, mostly 2TB although I upgraded parity to 4TB and added a couple 4TB drives last time I needed more storage.  My 4th unit is an experimental server I frequently change the configuration of -- it's had a couple 8TB Reds in it once; but currently just has a 2TB parity and 5 other mixed size drives.

 

The next time I build a box I'll certainly use 8TB (or larger) drives ... but there's simply no reason to move to those with my current units.  The Q25B is getting close to needing a bit more storage (under 2TB free) ... I may repurpose the box and just build another one with 8TB drives -- haven't decided yet.

 

I agree, however, that our "needs" tend to grow along with capacity.    The first hard drive I bought was a 26MB 14" Seagate "Winchester" unit back ~ 1980  (for a cool $4500).    I thought that was am AMAZING amount of capacity !!  :)

 

 

 

I should also add that my main driver behind wanting this feature or multiple licenses on the same machine for multiple arrays is I couldn't help myself with the dirt cheap 60 bay SAS3 JBOD HGST chassis on eBay recently. They spoke to me and I had to have them would love to be able to run unRAID on them rather than switching everything back over to Windows and Drivepool.

 

 

Sent from my iPhone using Tapatalk

 

60 bays !!!    Hmmm .... 60 8TB WD Reds configured as 2 30-drive UnRAID arrays = 56 x 8 = 448TB of storage for only $18,000 worth of drives !!  [Actually, when you compare it to what I once paid for 26MB it's an amazing STEAL]

 

Haha. These were brand new chassis bought by IXsystems (the FreeNAS guys) to harvest the 8TB Hitachi drives out of.

 

The only way to buy those chassis is fully loaded with drives from Hitachi. Something like $25,000-$50,000/EA. And from what I understand, Hitachi had a big ass sale on these things so IXsystems bought a bunch, harvested the 8TB drives and dumped the chassis on eBay.

 

So basically bought an entire brand new 60 bay chassis for about what it was going to cost me to upgrade one of my SM846 backplanes with a used SAS2 backplane from eBay in order to even run a 8TB red. So for me if was something fun to play with that also serves a purpose.

 

 

Sent from my iPhone using Tapatalk

So although this was a side tangent and I apologize for that, I hope you can see why running more than the allowed 28 data drives in a array would be vital to me. Or multiple licenses off the same head unit. I might have to try running two main U raid servers and have them both connected to one chassis splitting it right down the center as in the left 30 drives to Tower1 and the right 30 drives to Tower2. Just not sure if they would conflict with each other pulling off the same JBOD chassis. Worth a try I guess.

 

 

Sent from my iPhone using Tapatalk

I'd think you'd have no problem running two independent UnRAID servers using drives from the same chassis.

 

The issue would be connecting them.    IF you used a 24-drive RAID card [e.g. http://www.newegg.com/Product/Product.aspx?Item=N82E16816118137 ] you could actually mount 2 mini-ITX motherboards in the same chassis and have enough ports to connect all of your drives [24 with the card, 6 to the motherboard].

 

Not an inexpensive solution -- but then if you're buying 60 8TB drives the cost of those controllers is not really much of a factor 

By the way, I had heard that it's up to 30 drives now.  28 data + 2 parity, with cache drives cutting into that total.

IIRC, the pro licence supports 30 drives, 2 parity, and up to 36 cache drives in a pool with no limit on the # of unassigned devices.

By the way, I had heard that it's up to 30 drives now.  28 data + 2 parity, with cache drives cutting into that total.

IIRC, the pro licence supports 30 drives, 2 parity, and up to 36 cache drives in a pool with no limit on the # of unassigned devices.

Thanks for the clarification.  I tend not to use cache drives so that part of the restrictions I'm not as close to.

 

Note that (I believe) the limit on data drives is 28, *even* if you don't use dual parity.  I.e. 29 data drives with single parity, or 30 data drives with no parity, are *not* options.

By the way, I had heard that it's up to 30 drives now.  28 data + 2 parity, with cache drives cutting into that total.

IIRC, the pro licence supports 30 drives, 2 parity, and up to 36 cache drives in a pool with no limit on the # of unassigned devices.

 

Pro supports:

- up to 30 device assigned to parity-protected array, max 28 data devices

- up to 24 devices assigned to cache pool

- unlimited "unassigned" devices