• 1
  • 9
  • 10
  • 11(current)
  • 12
  • 13
  • 17
FreeNAS versus unRAID as the operating-system for a DIY NAS?
darkscout Wrote:All drives do NOT have to be online to access the data.

How about "All drives have to be powered on (spinning) to access any data within that volume"? Btw, I wonder if it supports drive power down when not in use?

froggit Wrote:unRAID:
-Allows drives to spin down if they are not being accessed, saving power and possible prolonging their life (but what about spin-up/down wear and tear on drive components?)
Downside include: when a non-parity drive dies you will probably lose all data on that drive, unless you are able to recover it using tools like (name them here or point to recovery URL perhaps?), or you have backups. In case of data loss you will need to use backups if you have them, or re-rip your media from original media.

This is a rather silly assumption.
Why would there be a parity drive if it's not for data redundancy?
Why would we bother messing around with unRAID rather than using JBOD\spanned disk on other free distro instead, if that's the case?
Reply
poofyhairguy Wrote:Wow! These pros and cons lists are great. By combining them together we will maximize the knowledge about these two options on the wiki.
I agree, this thread has suddenly become a treasure trove of good information. I looked, and it turns out there is already a great spot in the wiki for this info - and it's in severe need of updating.
http://wiki.xbmc.org/index.php?title=NAS...Storage%29
Reply
darkscout Wrote:Why does everyone keep saying this? It's not true. True, with ANY RAID (even unRAID) you're going to get better performance with matched drives.

Isn't 'perfectly matched drives' on RAID a carry over requirement from hardware RAID days where they control lower level things than what software RAIDs are exposed to?

Quote:If you have mismatched drives, you either give up security or space.

Please elaborate.

The 'security' claim is completely the opposite of the often quoted advice to use drives from different manufacturing batches, so they have less chance of failing together; be it due to similar wear & tear level or to avoid the chance of getting multiple drives from the same bad batch.

We're in the era of software raid w/ consumer drives now. Unless you go by SCSI or Enterprise class SATA/SAS drives, I don't think that advice (to use perfectly matched drives) still stand.
Reply
I added this thread to the wiki as:

# 5 Educational Discussions Regarding NAS
* 5.1 FreeNAS vs. unRAID

Smile
ASrock Ion 330HT - XBMCFreak 10.1 Lucid LiveCD. Everything works but System sounds over HDMI.
Reply
gadgetman Wrote:unRAID:
+ Simple install (embedded distribution on a USB stick).
+ Simple to maintain (most everything is done through the webgui)

These 2 features apply to FreeNAS also. Although I highly recommend an eUSB SSD (Intel Z-U130 clones) over a USB stick. Embedded and full FreeNAS both easily fit on 1GB drive.
Reply
jvdb Wrote:These 2 features apply to FreeNAS also.


Well, the thread got sidetracked to comparing unraid vs zfs, so we end up talking about the raid implementation, not the distribution.... ;(

Moreover.. although zfs is simpler with freenas as its frontend, unraid is still easier to install, setup and maintain.

But thanks, I will update my comparison Smile

PS: If it's FreeNAS vs unRAID, I will have to say that FreeNAS has another advantage: it supports many different form of raids (jbod, 0, 1, 5, raidz variants) and filesystems.
Reply
gadgetman Wrote:PS: If it's FreeNAS vs unRAID, I will have to say that FreeNAS has another advantage: it supports many different form of raids (jbod, 0, 1, 5, raidz variants) and filesystems.

Not just FreeNAS. ZFS supports RAID 0,1 vdevs.. You can also add vdev stripes with a single drive to a ZFS pool. So essentially a single volume with all the space of your drives combined (doesn't matter if they're different sizes). The downside is that this method has no parity/mirror, and if a single drive fails you lose everything.
Reply
yeah, and with unraid you don't care about raid at all... what happens if your motherboard fries and you have to change it? Probably, the raid implementations of the new board won't be compatible and maybe you can't redo the array.

also, I've had some bad experience with bios raid.. like once I swapped motherboards (same models) and forgot to activated bios in raid, booted, then activated and well... bad stuff happened.

I don't know if/how this could affect freenas / zfs. Any ideas?
Reply
PatrickVogeli Wrote:yeah, and with unraid you don't care about raid at all... what happens if your motherboard fries and you have to change it? Probably, the raid implementations of the new board won't be compatible and maybe you can't redo the array.

also, I've had some bad experience with bios raid.. like once I swapped motherboards (same models) and forgot to activated bios in raid, booted, then activated and well... bad stuff happened.

I don't know if/how this could affect freenas / zfs. Any ideas?

I have read the whole thread and it seems like the same concerns regarding Unraid and ZFS reappear quite often. So I will try to explain how they both work from what I gathered through this thread. I have to point out that my knowledge regarding ZFS nearly exclusively stems from this thread. I will edit the information if somebody points out flaws.

Similarities of how Unraid and ZFS function
1. Unraid and ZFS both work via SOFTWARE. So you don't have to worry with either of them too lose all your data if you change your motherboard.

2. Both protect your data via parity drives. You can lose as many drives as you have parity drives without loosing data.

Differences in how they protect data and hard drives are added

1. Unraid
a) Unraid only offers one parity drive.

b) Data is not spread among several drives. As a consequence, you can access the data directly from the drive without Unraid. Additionally, if you access a file, only one drive has to be spinning. As a negative side effect, the read/write speed is worse than with ZFS.

c) If you lose more than one non-parity drive, you lose the data on these drives. However, you can still access the remaining data (see b).

d) As long as the parity drive is as big as your biggest non-parity drive, you can add any hard drive you want and your storage increases by the whole amount of storage offered by the new hard drive. So expending storage is pretty inexpensive. The data stored on the new drive is protected by the parity drive already in place.

2. ZFS
a) ZFS arrays (combined storage of several had drives) are called vdev.

b) One vdev offers up to 3 parity drives (RAIDZ3= 3 parity drives, RAIDZ2= 2 parity drives, etc.).

c) You store your data in pools which you can assign any name. The data of one pool can be spread among several vdev.

c) Data is spread among several drives. You cannot take out a single drive and read from it. If you read/write data, all drives of the corresponding vdev have to be spinning. Faster than Unraid when it comes to read/write of files.

e) If you lose more non-parity drives in a vdev than you have parity drives, all the data stored in that vdev will be lost.

f) If you create a vdev with drives of different sizes, the smallest drive capacity will be the limit of capacity for the other drives.
Example: If you create a vdev with one 500 GB and three 2 TB drives, you will only be able to use 500 GB per drive.

g) Adding hard drives: You cannot add hard drives to an existing vdev. Instead, you can create a new vdev and add it your pool.
Example: You have already a vdev with RAIDZ3 and six 2 TB hard drives. This means you have 6 TB of storage and another 6 TB for parity. Now you buy 4 new 2 TB drives which you would like to add to your storage. For this, you have to create a new vdev. The question is then if you still want to use RAIDZ3 for the new vdev as that would mean you would only get 2 TB of new storage as 6 TB would be used for parity. Consequently, you either settle for less parity protection or buy more drives. For the example, let us assume you create a RAIDZ2 vdev. So you know have a new vdev with 4 TB of storage and 4 TB of parity. In total you know have 5 disks used for storage and 5 disks used for parity. However, you don't have to lose more 5 disks to lose data. If you lose more than 2 drives on the on the RAIDZ2 vdev, you will lose all of its data. So while there is an additional safety in this setup, it is not as save as 5 parity drives protecting all the data. Additionally, every time you add a new vdev, you have to pay extra for new parity disks and lose Sata ports due to them. This makes expending storage more expensive in relation to Unraid.

Data Protection

1. ZFS is superior regarding data protection when no drives fail because of additional features that prevent bit rot and other issues. This is not debatable.

2. DRIVE FAILURE
Well this is a heated debate throughout this thread and really depends on what the end-user is more afraid of. For an illustrative example, let us assume a data security paranoid individual (seem to be quite a few around here Laugh ) who can either choose between RAIDZ3 vdev from ZFS (3 parity drives) or Unraid. Let us further assume that money is of no issue and no other means of backup are used as they are available for both systems. Then the question boils down to what you are more afraid of:

a) The unlikely event of losing the data of 2 drives in the case of 2 non-parity drives failing with Unraid. You would not lose any data with your RAIDZ3 ZFS vdev in this scenario.
b) The even less likely event of losing the data of 3 drives in the case of 3 non-parity drives failing with Unraid. You would not lose any data with your RAIDZ3 ZFS vdev in this scenario.
c) The very unlikely event of losing all your data of your RAIDZ3 ZFS vdev array if 4 drives fail. You would "only" lose the data of the failed 4 non-parity drives in the case of Unraid.

The risks are seen relatively. I do not know how probable it is that 2,3, or 4 drives fail at the same time. But I think we can all agree that it is more likely for 2 drives to fail at the same time then 4. Furthermore, some people claim that the likeliness of multiple drive failure increases if you buy multiple drives from the same manufacturer and batch as you often do with ZFS. However, all these additional risk considerations are more of a feeling thing than anything else. So far I have not seen any statistical tables exactly quantifying for example the threat of bit rot or multiple drive failures from a bad batch. If anybody reading this is looking for a master thesis topic, data security of media NAS seems interesting Wink
Reply
PatrickVogeli Wrote:yeah, and with unraid you don't care about raid at all... what happens if your motherboard fries and you have to change it? Probably, the raid implementations of the new board won't be compatible and maybe you can't redo the array.

also, I've had some bad experience with bios raid.. like once I swapped motherboards (same models) and forgot to activated bios in raid, booted, then activated and well... bad stuff happened.

I don't know if/how this could affect freenas / zfs. Any ideas?

It doesn't. Hardware raid... sucks when it comes to this department. Very often you had to have not only the same hardware card, but exact firmware version too. Start searching for some old (~5-10yr) hardware RAID cards. If they're out of production you'll sometimes find that it's 10x original resale because someone REALLY NEEDS that particular card.

Quote:The 'security' claim is completely the opposite of the often quoted advice to use drives from different manufacturing batches, so they have less chance of failing together; be it due to similar wear & tear level or to avoid the chance of getting multiple drives from the same bad batch.

The security quote was from the fact that you have mis matched sized drives.

Say you have (for extreme). 1GB, 1GB, 1GB, 2TB drives. How much space does unRAID give you? If it's more than 3GB, something isn't being backed up.

gadgetman Wrote:How about "All drives have to be powered on (spinning) to access any data within that volume"? Btw, I wonder if it supports drive power down when not in use?

Nope. Not even that. I could physically yank those drives and still have access to my volume.
Reply
darkscout Wrote:The security quote was from the fact that you have mis matched sized drives.

Say you have (for extreme). 1GB, 1GB, 1GB, 2TB drives. How much space does unRAID give you? If it's more than 3GB, something isn't being backed up.

Your example is correct. But only because the largest drive is always used for parity. So if you have mismatched drives that are 1GB, 20GB, 500GB, 2TB you will have 521GB of storage (=sum from first three drives) and all of the data is backed up.

darkscout Wrote:Nope. Not even that. I could physically yank those drives and still have access to my volume.

Yes from my understanding you can yank as many drives as you have parity drives in one vdev and you could still access the data. However, from how I read gadgetman`s post, he wanted to point out something else. If you read any file from your vdev, all the hard drives from that vdev have to be spinning as the file is spread among all drives. This is different from Unraid. Unraid stores one file on one hard drive. This means only one hard drive has to be spinning when you read the file while the other hard drives can stay spinned down.
Reply
jvdb Wrote:Not just FreeNAS. ZFS supports RAID 0,1 vdevs.. You can also add vdev stripes with a single drive to a ZFS pool. So essentially a single volume with all the space of your drives combined (doesn't matter if they're different sizes). The downside is that this method has no parity/mirror, and if a single drive fails you lose everything.

Good point, thanks.

Quote:Nope. Not even that. I could physically yank those drives and still have access to my volume.

You misunderstand this point (again and again). We're talking about idling the drives to lower their wear & tear and lower the machine's power consumption.

@beckstown: thank you Smile that was exactly the point.
Reply
Costs of the unraid software shouldn't be considered as a significant part of the comparison. Even buying the most expensive version is insignificant vs the costs of media and hardware.

IE (actual costs may vary, but will probably be higher)

MB, CPU, Memory, chassis $400
4 2TB HDD $400
6TB of BR rips ~$1350 (assumes 1.8TB usable/drive & 40GB rips & $10 per disk)

That's $2150 vs a $90 license. As the number of drives goes up the significance of the license costs becomes less and less.
Reply
Would anyone be interested in splitting a 2-pack of unRAID Pro? It would be $70 each. PM me or email s((DOT))m((DOT))[email protected]. (remove the ((DOT))s)
Reply
beckstown Wrote:I have read the whole thread and it seems like the same concerns regarding Unraid and ZFS reappear quite often. So I will try to explain how they both work from what I gathered through this thread. I have to point out that my knowledge regarding ZFS nearly exclusively stems from this thread. I will edit the information if somebody points out flaws.

Similarities of how Unraid and ZFS function
1. Unraid and ZFS both work via SOFTWARE. So you don't have to worry with either of them too lose all your data if you change your motherboard.

2. Both protect your data via parity drives. You can lose as many drives as you have parity drives without loosing data.

Differences in how they protect data and hard drives are added

1. Unraid
a) Unraid only offers one parity drive.

b) Data is not spread among several drives. As a consequence, you can access the data directly from the drive without Unraid. Additionally, if you access a file, only one drive has to be spinning. As a negative side effect, the read/write speed is worse than with ZFS.

c) If you lose more than one non-parity drive, you lose the data on these drives. However, you can still access the remaining data (see b).

d) As long as the parity drive is as big as your biggest non-parity drive, you can add any hard drive you want and your storage increases by the whole amount of storage offered by the new hard drive. So expending storage is pretty inexpensive. The data stored on the new drive is protected by the parity drive already in place.

2. ZFS
a) ZFS arrays (combined storage of several had drives) are called vdev.

b) One vdev offers up to 3 parity drives (RAIDZ3= 3 parity drives, RAIDZ2= 2 parity drives, etc.).

c) You store your data in pools which you can assign any name. The data of one pool can be spread among several vdev.

c) Data is spread among several drives. You cannot take out a single drive and read from it. If you read/write data, all drives of the corresponding vdev have to be spinning. Faster than Unraid when it comes to read/write of files.

e) If you lose more non-parity drives in a vdev than you have parity drives, all the data stored in that vdev will be lost.

f) If you create a vdev with drives of different sizes, the smallest drive capacity will be the limit of capacity for the other drives.
Example: If you create a vdev with one 500 GB and three 2 TB drives, you will only be able to use 500 GB per drive.

g) Adding hard drives: You cannot add hard drives to an existing vdev. Instead, you can create a new vdev and add it your pool.
Example: You have already a vdev with RAIDZ3 and six 2 TB hard drives. This means you have 6 TB of storage and another 6 TB for parity. Now you buy 4 new 2 TB drives which you would like to add to your storage. For this, you have to create a new vdev. The question is then if you still want to use RAIDZ3 for the new vdev as that would mean you would only get 2 TB of new storage as 6 TB would be used for parity. Consequently, you either settle for less parity protection or buy more drives. For the example, let us assume you create a RAIDZ2 vdev. So you know have a new vdev with 4 TB of storage and 4 TB of parity. In total you know have 5 disks used for storage and 5 disks used for parity. However, you don't have to lose more 5 disks to lose data. If you lose more than 2 drives on the on the RAIDZ2 vdev, you will lose all of its data. So while there is an additional safety in this setup, it is not as save as 5 parity drives protecting all the data. Additionally, every time you add a new vdev, you have to pay extra for new parity disks and lose Sata ports due to them. This makes expending storage more expensive in relation to Unraid.

Data Protection

1. ZFS is superior regarding data protection when no drives fail because of additional features that prevent bit rot and other issues. This is not debatable.

2. DRIVE FAILURE
Well this is a heated debate throughout this thread and really depends on what the end-user is more afraid of. For an illustrative example, let us assume a data security paranoid individual (seem to be quite a few around here Laugh ) who can either choose between RAIDZ3 vdev from ZFS (3 parity drives) or Unraid. Let us further assume that money is of no issue and no other means of backup are used as they are available for both systems. Then the question boils down to what you are more afraid of:

a) The unlikely event of losing the data of 2 drives in the case of 2 non-parity drives failing with Unraid. You would not lose any data with your RAIDZ3 ZFS vdev in this scenario.
b) The even less likely event of losing the data of 3 drives in the case of 3 non-parity drives failing with Unraid. You would not lose any data with your RAIDZ3 ZFS vdev in this scenario.
c) The very unlikely event of losing all your data of your RAIDZ3 ZFS vdev array if 4 drives fail. You would "only" lose the data of the failed 4 non-parity drives in the case of Unraid.

The risks are seen relatively. I do not know how probable it is that 2,3, or 4 drives fail at the same time. But I think we can all agree that it is more likely for 2 drives to fail at the same time then 4. Furthermore, some people claim that the likeliness of multiple drive failure increases if you buy multiple drives from the same manufacturer and batch as you often do with ZFS. However, all these additional risk considerations are more of a feeling thing than anything else. So far I have not seen any statistical tables exactly quantifying for example the threat of bit rot or multiple drive failures from a bad batch. If anybody reading this is looking for a master thesis topic, data security of media NAS seems interesting Wink
thanks! You made it a little more clear to me Smile

After reading this, I'm sure that if I ever build a NAS (and I will, sure), I'll go with unRaid. I'm a home user, unRaid is OK for me, and has a few advantages over Freenas with ZFS.

I can't consider having a 2 or 3 drive parity setup and I can't consider adding a drive and not getting its full capacity from the very begining.

Also, I'm curious to know how a 2/3 drive parity works... 1 parity drive is easy, you simply count how many bits are '1' over the drives and the parity bit will be '1' or '0' depending if it was even or uneven parity. How does that work when you have 3 parity drives?
Reply
  • 1
  • 9
  • 10
  • 11(current)
  • 12
  • 13
  • 17

Logout Mark Read Team Forum Stats Members Help
FreeNAS versus unRAID as the operating-system for a DIY NAS?0