removing part of a zpool

[iTris666]

Hi, instead of replacing a device, I managed to add the new disk to a zpool and I try to get thins back to normal. spool won't let me detach, offline or remove the device. Could you help me ? Here is my zpool status :

NAME STATE READ WRITE CKSUM

vol DEGRADED 0 0 0

raidz1-0 DEGRADED 0 0 0

17707987528179553980 UNAVAIL 0 0 0 was /dev/gptid/87d0afc6-52f0-11e3-a1b9-bc5ff4cef133

gptid/b7232bec-a797-11e6-a30e-bc5ff4cef133 ONLINE 0 0 0

gptid/88924114-52f0-11e3-a1b9-bc5ff4cef133 ONLINE 0 0 0

gptid/88f6cc5a-52f0-11e3-a1b9-bc5ff4cef133 ONLINE 0 0 0

gptid/bc4502c6-6ed7-11e7-b20c-0010b5fb1eab ONLINE 0 0 0


I want to remove gptid/bc4502c6-6ed7-11e7-b20c-0010b5fb1eab and use it to replace the missing disk in the raidz1.

 

[BigDave]

Is the drive gptid/bc4502c6-6ed7-11e7-b20c-0010b5fb1eab part of the RAIDz1 pool,
or was it added as a spare device?

Please follow the Forum Rules for posting (link is found at the top of the page).

We need more information prior to helping you with this, the output you have
posted thus far is a start.

Please re-post the output from the following command via the CLI zpool status poolnamehere
Please do not edit the output prior to posting :)

 

[danb35]

I want to remove gptid/bc4502c6-6ed7-11e7-b20c-0010b5fb1eab and use it to replace the missing disk in the raidz1.

You can't. You added it as a separate striped disk to your pool, blowing through a warning screen with bold red text, and switching into manual mode, in order to do so. Back up your data, destroy your pool, and recreate it with the correct disks. And next time check the manual's instructions on replacing disks.

 

[danb35]

What part of this screen:

made you think, "sure, I should switch to Manual Setup so I can bypass this"?

 

[iTris666]

FYI I run corral and on it it is a drag and drop and a click on the apply button. No need to take me for a fool.

I bought a 10 TB drive that will let me backup everything and I will destroy and rebuilt the pool. More than 48 hours to rsync the 9 TB of data I have.

 

[danb35]

I run corral

There's your problem--Corral is dead and unsupported.

Edit: Among the reasons that it's dead is that its GUI was so half-baked--letting you do dangerous thing (like what you did) without warning, while making it impossible to do ordinary things (like replacing a failed disk).

 

[SweetAndLow]

No need to take me for a fool.

Corral has been dead for months and should be migrated away from.

Mod note: Edited to remove unnecessary comments.

- Ericloewe

 

[Ericloewe]

More than 48 hours to rsync the 9 TB of data I have.

You don't need rsync. Use zfs send | recv and it should be much faster.

 

[cyberjock]

You don't need rsync. Use zfs send | recv and it should be much faster.

+ 10000 on this idea. Use replication as it will go as fast as the disks can. rsync is likely the post painful way possible to move large quantities of data like you are doing. ;)

 

[Stux]

+ 10000 on this idea. Use replication as it will go as fast as the disks can. rsync is likely the post painful way possible to move large quantities of data like you are doing. ;)

FreeNAS needs a feature for one-shot local replications...

 

[danb35]

FreeNAS needs a feature for one-shot local replications...

Yes, and for recurrent local replications as well.

 

[Ericloewe]

FreeNAS needs a feature for one-shot local replications...

It's in the works for the replication rewrite.