Procedure to move from striped mirror to single mirror

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

crazyj

Member
Nov 19, 2015
75
2
8
49
I've currently 4x 500GB drives in a striped mirror. I'm wanting to move these over to 2x 4TB disks since I'm running short on room. What's the procedure to accomplish this? I have extra slots, so all disks reside in the server currently. Is it possible to mirror to the 4TB, then remove the 500GB disks?
 
  • Like
Reactions: T_Minus

ttabbal

Active Member
Mar 10, 2016
747
207
43
47
Create a new pool using the 4TB mirror. zfs send/recv to copy the data. Unmount the old pool and pull the drives as backups or use them for whatever your plan is.

Note that it needs to be a new pool. If you add them to the existing pool, you will have to reformat everything to undo it. You can rename the pool later if desired.
 

crazyj

Member
Nov 19, 2015
75
2
8
49
And is this handled anywhere in the GUI, or am I needing to do command line?
 

ttabbal

Active Member
Mar 10, 2016
747
207
43
47
You should ask @gea that. I haven't used napp-it in a while so I'm not sure. I would think you could at least do the pool management stuff from the GUI, but I'm a CLI guy myself.
 

gea

Well-Known Member
Dec 31, 2010
3,159
1,195
113
DE
Can all be done in napp-it
Create a new pool from a mirror, then create a replication job to transfer filesystems from your raid-10 to the new mirror. Then export both pools and import the mirror under the former poolname of the raid-10.

If you want to preserve the exact same pool layout:
As a replication always creates a filesystem below another filesystem you cannot replicate the whole pool but must for example replicate single filesystems ex oldpool/filesystem -> newpool to get newpool/filesystem. If you replicate the whole pool recursively you get newpool/oldpool/filesystem.
 

crazyj

Member
Nov 19, 2015
75
2
8
49
Is the end result the same? I did a send/receive and haven’t double checked data but I assume it all went to plan.