Replacing drive in mirror with smaller one

[zimon]

I have a boot zfspool with 3 mirrored devices. 2 16GB usb drives and one old 128GB SSD.

Unfortunately the SSD just died so I wanted to replace it with another 16GB USB drive (the exact same type of the other already mirrored 2)
But when I try to replace it (in the WebGUI) I get an error that the drive is too small.
Do I need a drive wich is 128GB big? Even if the others are only 16GB?

If so, can I remove the (already removed) SSD from the pool for good and add another 16GB usb mirror?

thanks, hope someone can help me

 

[Ericloewe]

When did you first install FreeNAS? The EFI system partition size was increased a few months ago, with the practical result that you can't use disks of the same size - they'd need to be 100 MB (or however much) larger. Of course, you can do the whole thing manually, but that's a pain in the ass.

 

[zimon]

ah ok now I understand you, I guess.

So normally replacing the 128GB disk with a 16GB flash drive should work since the mirror is only 16GB right?

Well I have FreeNAS-9.10.2-U1 (86c7ef5) I did not update a long time ago (I think) so that should actually work, or am I mistaken?

Can't I just kick out the already removed disk so that I just have a mirror with 2 disks? And then I could just add another 16GB drive right?

 

[Ericloewe]

Hmm, 9.10 is much older than the change I mentioned. Can you post the exact error you get?

 

[zimon]

I tried two different usb flash drives where one was completely unused and new.

Code:

Nov 20 21:55:00 theVault GEOM: da2: the primary GPT table is corrupt or invalid.

Nov 20 21:55:00 theVault GEOM: da2: using the secondary instead -- recovery strongly advised.

Nov 20 21:55:01 theVault notifier: 32+0 records in

Nov 20 21:55:01 theVault notifier: 32+0 records out

Nov 20 21:55:01 theVault notifier: 33554432 bytes transferred in 3.322055 secs (10100504 bytes/sec)

Nov 20 21:55:05 theVault notifier: dd: /dev/da2: short write on character device

Nov 20 21:55:05 theVault notifier: dd: /dev/da2: end of device

Nov 20 21:55:05 theVault notifier: 33+0 records in

Nov 20 21:55:05 theVault notifier: 32+1 records out

Nov 20 21:55:05 theVault notifier: 33595392 bytes transferred in 3.933975 secs (8539808 bytes/sec)

Nov 20 21:55:12 theVault ZFS: vdev state changed, pool_guid=15607185754263842137 vdev_guid=2704911025820488020

Nov 20 21:55:14 theVault manage.py: [middleware.exceptions:37] [MiddlewareError: Failed to attach disk: cannot replace 17602617589208774497 with da2p2: device is too small

]

Nov 20 21:55:15 theVault manage.py: [rollbar:1265] Got unexpected status code from Rollbar api: 403

Response:

{

  "err": 1,

  "message": "access token not found: caf06383cba14d5893c4f4d0a40c33a9"

}

Nov 20 21:55:15 theVault manage.py: [rollbar:1097] Exception while posting item ApiError(u'access token not found: caf06383cba14d5893c4f4d0a40c33a9',)

Traceback (most recent call last):

  File "/usr/local/lib/python2.7/site-packages/rollbar/__init__.py", line 1095, in _send_payload

	_post_api('item/', payload, access_token=access_token)

  File "/usr/local/lib/python2.7/site-packages/rollbar/__init__.py", line 1138, in _post_api

	return _parse_response(path, SETTINGS['access_token'], payload, resp)

  File "/usr/local/lib/python2.7/site-packages/rollbar/__init__.py", line 1274, in _parse_response

	raise ApiError(json_data.get('message') or 'Unknown error')

ApiError: access token not found: caf06383cba14d5893c4f4d0a40c33a9

Nov 20 21:55:42 theVault manage.py: [freeadmin.navtree:577] Couldn't retrieve http://192.168.188.25/plugins/resilio/6/_s/treemenu: timed out

 

[zimon]

any help is highly appreciated!

Can't I even just remove the "already removed" drive from the pool so that I have a working pool with 2 mirrored drives? That should be possible right?

 

[Ericloewe]

Can't I even just remove the "already removed" drive from the pool so that I have a working pool with 2 mirrored drives? That should be possible right?

Yeah.

 

[zimon]

So then I just could remove it, and attach one new drive to the pool via the WebGUI, correct?


The detachment of the old (already removed) drive is only possible via CLI right? I could not find anything in the WebGUI

 

[Ericloewe]

I haven't really tried it, but the option might still be there, if the disk is still a part of the pool.