Register for the iXsystems Community to get an ad-free experience and exclusive discounts in our eBay Store.

2x individual vs stripe for temporary data migration

ShimadaRiku

Newbie
Joined
Aug 28, 2015
Messages
92
Rebuilding my 4x8TB pool into a 6x8TB which contains 7TB of data.

I have 2x spare 4TB drives to use as temp storage.

Should I
  • stripe them together into a single volume for ease of transfer
  • make individual drive volumes splitting the data
Either case there will be zero redundancy. Have no backup, not important/critical, plex media content can be re-downloaded, but would take a long time... Chances are low the temp 4TB drives will completely fail during migration copying, but it exist. Just concerned of corruption while transferring so I know which data to re-download.

If there is a more advisable route let me know.
 

Chris Moore

Super Moderator
Moderator
Joined
May 2, 2015
Messages
9,551
Either case there will be zero redundancy. Have no backup, not important/critical, plex media content can be re-downloaded, but would take a long time... Chances are low the temp 4TB drives will completely fail during migration copying, but it exist. Just concerned of corruption while transferring so I know which data to re-download.
If data is corrupted, ZFS should be able to tell you.
If there is a more advisable route let me know.
The only better option would be a third 4TB drive, so you can make a RAIDz1 for some redundancy. Still not a great option.
 

ShimadaRiku

Newbie
Joined
Aug 28, 2015
Messages
92
Code:
[root@freenas] ~# zpool list
NAME           SIZE  ALLOC   FREE  EXPANDSZ   FRAG    CAP  DEDUP  HEALTH  ALTROOT
HDD1          3.62T   483G  3.15T         -     1%    12%  1.00x  ONLINE  /mnt
HDD2          43.5T  15.9T  27.6T         -    21%    36%  1.00x  ONLINE  /mnt
freenas-boot  15.9G   664M  15.2G         -      -     4%  1.00x  ONLINE  -


Rebuilt my HDD2 from 4x8TB to 6x8TB but it's FRAG is 36%

Was assuming snapshot send/receive datasets would write everything sequentially thus reducing Fragmentation of free space.

Also didn't know changing dataset default recordsize from 128K to 1024K & sync=disabled would greatly improve disk performance for large media files.
 
Last edited:
Top