zfs discuss - Apr 2008 - Replacing a drive, resilvering results in checksum errors?

This is probably a FAQ but I have been unable to turn up the answer in searches,
thanks for your patience.

I have a zfs testbed set up with 3x 200 GB SATA drives in raidz. I pulled a
drive (ad4) and replaced it to experience the rebuild procedure. After
scrubbing/resilvering, I get the status report below. These are my questions...

- when is it safe to use zpool clear? the array seems to be happy, but... 
- why are checksum errors reported on a drive which should be freshly
resilvered?

Many thanks. 

[0123][root at ztest:/home/matt]# zpool status
  pool: pond
 state: ONLINE
status: One or more devices has experienced an unrecoverable error.  An
        attempt was made to correct the error.  Applications are unaffected.
action: Determine if the device needs to be replaced, and clear the errors
        using ''zpool clear'' or replace the device with
''zpool replace''.
   see: http://www.sun.com/msg/ZFS-8000-9P
 scrub: scrub completed with 0 errors on Fri Apr 25 01:16:51 2008
config:

        NAME        STATE     READ WRITE CKSUM
        pond        ONLINE       0     0     0
          raidz1    ONLINE       0     0     0
            ad4     ONLINE       0     0 44.68
            ad6     ONLINE       0     0     0
            ad8     ONLINE       0     0     0

errors: No known data errors
 
 
This message posted from opensolaris.org

Hi Matt,
if you can remember the status information and are the only  
administrator, you can just doa zpool clear and re-scrub your pool  
regularly.
zpool clear does no more or less than resetting the status and the  
counters.
(A reboot or export does the same thing, at least no Mac OS X.)
The point is that all errors have been corrected. All errors are  
accumulated until you do a clear.
If you want to know how well your disks work, the easy way is to do a  
zpool clear followed by a scrub.
If you continue to see checksum errors, you might want to conside  
replacing your disk,
ralf

Am 25.04.2008 um 16:42 schrieb zfs-discuss-request at opensolaris.org:
> [zfs-discuss] Replacing a drive,	resilvering results in
> 	checksum errors?
> To: zfs-discuss at opensolaris.org
> Message-ID: <25424431.1209112324546.JavaMail.Twebapp at oss-app1>
> Content-Type: text/plain; charset=UTF-8
>
> This is probably a FAQ but I have been unable to turn up the answer  
> in searches, thanks for your patience.
>
> I have a zfs testbed set up with 3x 200 GB SATA drives in raidz. I  
> pulled a drive (ad4) and replaced it to experience the rebuild  
> procedure. After scrubbing/resilvering, I get the status report  
> below. These are my questions...
>
> - when is it safe to use zpool clear? the array seems to be happy,  
> but...
> - why are checksum errors reported on a drive which should be  
> freshly resilvered?
>
> Many thanks.
>
> [0123][root at ztest:/home/matt]# zpool status
>  pool: pond
> state: ONLINE
> status: One or more devices has experienced an unrecoverable error.   
> An
>        attempt was made to correct the error.  Applications are  
> unaffected.
> action: Determine if the device needs to be replaced, and clear the  
> errors
>        using ''zpool clear'' or replace the device with
''zpool replace''.
>   see: http://www.sun.com/msg/ZFS-8000-9P
> scrub: scrub completed with 0 errors on Fri Apr 25 01:16:51 2008
> config:
>
>        NAME        STATE     READ WRITE CKSUM
>        pond        ONLINE       0     0     0
>          raidz1    ONLINE       0     0     0
>            ad4     ONLINE       0     0 44.68
>            ad6     ONLINE       0     0     0
>            ad8     ONLINE       0     0     0
>
> errors: No known data errors