Ok so, I'm looking at the storage screen of my Freenas box and for some reason I can't comprehend a missing 2.3 TiB of space. For reference and so that we can all laugh at whatever small mistake I'm making here is my storage:
first the Used and Available of the first row makes sense I have 4 6TB drives in there.
However, if I add up the used space I some how can't get it to total 4.5 TiB. I know media shows up twice at 2.2 TiB each time, do I add them both in? is one a backup copy of the other? am I suppose to have that backup of everything? For the record I'm running Raidz2 and this is the first Freenas box that I have setup.
Right, so you need to zoom out a bit from that screen shot and look at the compression. FreeNas will compress files on the fly too save space, which is likely where the extra file size is coming up, with less actual used on disk.
Raid 5 or ZFS Z1 is not recommended on drives as large as his, as the chance of a second drive dying during the rebuilding process is not small by any means, as the rebuild time for a 6TB drive is ridiculous. This is totally warranted if you care about your data's safety...
Raid Z2 has a minimum of 4 drives, the same as Raid 6. I have no clue where you are getting the 11 drive figure from...
I was thinking Raid Z+2 which was RAID-Z3 in actuality that requires 11 drives lol. But I would still not recommend Raid Z2 without multi disk parity with hot-spare.
Capacity warrants nothing in the sense of a raid, unless you have multiple sources backing that up then I would very much like to read into it.
RAID Z is the ZFS parity raid mechanism, with Z2 and Z3 referring to dual and triple parity, respectively. arity has repercussions both in terms of fault tolerance and performance. Parity generation has a performance cost as ell as resource demand (each parity member requires a disk and a host bus channel) so generally should be ccounted or in terms of sensitivity to downtime (each additional parity bit results in an additional disk that can be lost without ffecting data availability) within the requirements for usable space and available mechanical resources.
Add to that the fact that data is written and read in blocks in binary arrangement, leading to the need to align data to powers of two. Since RAID stripes consist of a number of data blocks for each (set of) parity block(s), optimal RAID arrangements necessarily follow powers of two for maximum efficiency. Since one parity block for 2 drives is very mostly in terms of hardware, optimal RAID configuration therefore follow in 4s- 1 parity disk per 4 data stripes. This provides the most commonly optimal arrangement of aligned stripes with sufficient fault tolerance. So double parity (Z2) would be optimally applied to 10 disk RAIDsets, and triple parity (Z3) for 15 disk RAIDsets.
Above 15 you'd want to split your data into separate raidsets, or create compound (nested) RAID arrangements; on occasion it may makes sense to double up if the application is not sensitive to the greater exposure to fault- eg 9 disks for a Z1, 18 for a Z2, 27 for Z3.
The larger the disk size, the longer it takes to rebuild the array, and more time spent rebuilding an array is more time spent stressing the disks in weird ways. This is why Raid 5 or ZFS Raid Z1 is not recommended for disks of his size, as the time spent rebuilding a 6TB disk can be ridiculous depending on how full the disk was, and all that time spent rebuilding is time stressing disks in hard and weird ways, which can lead to a second disk failing. More time spent rebuilding is more time when the array is in a state where a second drive can fail and yield a total loss of all data.
I cannot believe I did not find this before, awesome. When I've read them I just thought in my head everytime I see a problem with throughput, "Why not just develop a new RAID like method/technology to compensate for the flaws of the current RAID since it is old tech? Or could we transition to SSD's and slowly implement the RAID algorithms towards that instead?
However I did like the knowledge you've given me, thank you.
The hardware problems of Hard Drives need to be mitigated for them to stay relevant, as in 10 years solid state drives will surely phase out hard drives if capacity and speed do not drastically improve. Other than that, Raid alternatives like ZFS can solve some issues. ZFS is the equivalent of ECC technologies in Ram, and its amazing because it prevents Bit Rot. But, in 3 years the capacity of Hard Drives are expected to be so high that Raid Z2 or Raid 6 will not be good enough protection any longer, so Raid Z3 or Raid 7 will have to be used, which is the same concept of Raid 5 and 6, but with three parity drives instead of 2 in raid 6 and 1 in raid 5.
So I posted this right before going to have a baby and haven't been back to see the replies till today.
thecaveman,
when I look at compression it is all relatively low between 1 and 2 % for each of the areas.
I now think I understand why I have 2 media partitions, each taking up 2.2 TB of space. When I originally started this project I had a windows computer that was storing all of our pictures/media files. so I created a CIFS share so that I could move all of this to the Freenas box. After getting everything safely stored on the freenas box, I upgraded to Linux and created a NFS share so that I could access the data.
I still have a CIFS share so that my wife can access all the pictures/media and I'm wondering if Freenas is just duplicating the data between the 2 shares, even though they are suppose to be pointing to the same directory on the Freenas box?
As for Z2 over Z1. I went with the lazy newb setup and if I recall correctly Z2 is what the Freenas software recommended. Of course I now have 4 more 6 TB drives to install so this might need to change as I grow things. I'm not too worried about wasted space, as the top Priority for this is Data Retention. Imagine if you will a nuclear wasteland after a nuke gets drops. If I lose all the baby photos, all the home movies that is what my life will be like after I suffer the wrath.
Which is why I also am backing up to encrypted hard drives that are moved off site. I have 3 3 TB drives that I rotate through a monthly backup, but it would be great to automate this process to a computer stored off site.
