Help, Pool unhealthy state, Can't isolate the issue

[Nyarlathotep]

I migrated from 11.3u5 to 12.0u1. I am sitting on a error message in the gui. I ran zpool status -v and found that my first manual snapshot on my media dataset after my upgrade contained a error. Since I had automated snapshots completed already I simply deleted the snapshot and ran zpool clear. This didn't clear the error and i'm unsure what is going on now.

My pool is showing green checkmarks but "unhealrthy" my zpool status -v is showing a nice 0 0 0 read write chksum, but some kind of error is present. Not sure if this is a data migration issue or something elses. I"m only running conveyance short SMART tests weekly so I don't have any data. In hindsite maybe i should be running long tests monthly.?

zpool clear didn't remove the current displayed error

 

[Samuel Tai]

Try a scrub. The error appears to be in a single metadata block somewhere in your pool. The last scrub wasn't able to clear the error, but a second might.

 

[Nyarlathotep]

Thanks, I did do a scrub and that didn't appear to clear the error. Is this a situation where I can safely run zpool -nf or zpool clear?

 

[Samuel Tai]

I think so, but zpool clear typically only resets the fault counters for the READ/WRITE/CKSUM counters in the zpool status, so I don't think this will help your situation.

Try zpool status -sv to view a SLOW column, which may indicate the drive where the bad block might be.

 

[Nyarlathotep]

The double scrub appears to have corrected the problem. Just took a longer time to finish the scrub than I thought it would.