Can't resilver, degraded and online in the same Raidz2

[mactsk]

Hi, i had several faulted drives and temporary put some standard hdd's, now i am swapping the standard drives with WD NAS units.
After successfully resilvering some units i am stuck with this situation that seems strange to me


(see attached image)


If i make the ada 3 offline and force it to connect i get this error, and the drive reconnects

Code:

Request Method: POST
Request URL: http://192.168.5.4/storage/zpool-raid6/disk/replace/6782683768401974622/
Software Version: FreeNAS-11.0-U4 (54848d13b)
Exception Type: MiddlewareError
Exception Value:
[MiddlewareError: b'Disk replacement failed: "cannot open \'gptid/11a9080f-de9a-11e7-b90e-6805ca2bf565\': no such GEOM provider, must be a full path or shorthand device name, "']
Exception Location: ./freenasUI/middleware/notifier.py in zfs_replace_disk, line 1142
Server time: Mon, 11 Dec 2017 18:38:29 +0100
Traceback

 

Request information
GET
No GET data

POST
Variable Value
__all__ ''
force 'on'
replace_disk 'ada3'
__form_id 'form_ZFSDiskReplacementForm'
FILES
No FILES data

COOKIES
Variable Value
csrftoken 'mlDDNhYdNmETwzTbQSOvINuJpcSMdhCcVtj7kxKmsKsLmXY70XsZzxLJxzwdBk4D'
fntreeSaveStateCookie 'root%2Croot%2F262%2Croot%2F262%2F273%2Croot%2F18%2Croot%2F1%2Croot%2F1%2F12'
sessionid 'dr5tqd4o3bm3hsiojekd6z3459eddgze'
META
Variable Value

The data is readable, but it's a lot of stuff and i can't just copy it to another place.

 

[rs225]

Do you have ECC RAM, or have run a memory test on this machine before? You are showing top-level checksum errors. If you can't find the cause, it means the pool is going to have to be abandoned. You can not complete resilver with top-level checksum errors. Don't off-line any more drives.

What are the drive types and controller? What kind of drive bays?