RAIDZ volume reporting the wrong size.

[mtomsovic]

Hi, new to FreeNAS but have been scouring the forums and web for the last week or so since the hardware has been on order trying to learn and prepare.

I completely understand that there is a 2gb per-drive allocation for swap space. With that I would expect the size difference to be negligible, however, I'm about 1TB off.

4x2tb drives

Code:

[root@freenas] ~# zfs list
NAME       USED  AVAIL  REFER  MOUNTPOINT
Volume_1   698K  5.19T   209K  /mnt/Volume_1

Code:

[root@freenas] ~# df -h
Filesystem             Size    Used   Avail Capacity  Mounted on
/dev/ufs/FreeNASs1a    926M    378M    473M    44%    /
devfs                  1.0k    1.0k      0B   100%    /dev
/dev/md0               4.6M    3.2M      1M    76%    /etc
/dev/md1               823k    2.0k    756k     0%    /mnt
/dev/md2               149M     10M    126M     8%    /var
/dev/ufs/FreeNASs4      19M    499k     17M     3%    /data
/dev/md3               1.9G    584k    1.7G     0%    /var/tmp/.cache
Volume_1               5.2T    209k    5.2T     0%    /mnt/Volume_1

I'm expecting 4 drives - 1 drive = 3 drives * 2tb == 6TB of usable space.

There were a lot of posts on this in the forums but they all seemed to be over minimal differences in space and related to the 2GB swap allocation.

A point in the right direction would be greatly appreciated. I'm holding off on copying my data back to the NAS until after I figure this out, which makes it inaccessible, but will save me grief if I need to fix something I did wrong and reformat.

Thanks much!

P.S. I'm using the latest available release version

Code:

FreeBSD 8.3-RELEASE-p4 (FREENAS.amd64)

 

[joeschmuck]

That is about correct for a four drive RAIDZ1 using 2TB drives. That is what I had + 8GB because I didn't use a swap partition on my drives.

It has to do with if you lost a drive, how do you create the data lost. Well it resides on all the drives so overall you loose a bit of space for that extra data. You should read up on RAIDZ1 and how the data is striped across the drives and the error recovery data how it works.

Either way, ~5.19TB sounds right.

 

[mtomsovic]

Interesting.. I will definitely read up on it some more. I didn't realize RAIDZ took up that much more parity space than RAID5. 1.5 drives vs 1 drive is more than I was expecting.

Thanks for your time!

 

[joeschmuck]

Yea, I don't know all the specifics behind it, I gave up trying to figure it out but I think some of it has to do with ZFS itself.

 

[Brosif_My_Nif]

I couldn't figure out why my space seemed low recently, then realized I had a few snapshots eating up a couple TB. Something to keep in the back of your mind when space goes missing.

 

[paleoN]

I didn't realize RAIDZ took up that much more parity space than RAID5.

It doesn't.

This question get asked all the time. Not sure what joeschmuck was thinking, but you don't get 2TB useable out of a 2TB drive; try ≈ 1.8TB.

Code:

(Total drives - # of parity drives) * size of smallest drive = total space.

(4-1) * 1.8TB = 5.4TB

Just about what you have. The additional .2TB or so is due to pool creation overhead.

 

[joeschmuck]

Did I state you get 2TB out of a 2TB drive somewhere? What I said was the size he had was consistent with what everyone else gets and gave a simple explanation of what I think is going on, why you don't get the full 6TB of data storage. Of course you didn't explain why the size was 1.8TB. I suspect it's from formatting and parity data but that is why I said he should read up on it because I wasn't sure.

 

[paleoN]

Did I state you get 2TB out of a 2TB drive somewhere?

No, you did not. I read more into it than you wrote.

Of course you didn't explain why the size was 1.8TB.

Marketing BS. Marketing uses 1000 bytes vs binary 1024 bytes. I thought you had posted as much some time ago. Clearly, faulty memory on my part.