Ivan Wang wrote:
> Ugh, sorry, a bit late coming to this thread,
> Is this issue reproducible when one uses pkg image-update exclusively?
> It sounds like the failed mount is because /opt is not empty before zfs 
> actually mounts BE /opt dataset, then that figures. I would be worried if 
> there's possibility that issue still can happen with only ``pkg 
> image-update'' to manage /opt.

This issue is not specific to pkg image-update it is caused initially by
mounting a BE, activating that BE but not unmounting the BE before booting
to it.

-evan

> 
> Thanks
> Ivan.
> 
>> It's likely that the mountpoint got messed up while
>> these mounts were failing 
>> due to /opt not being empty. It's possible that
>> repeated mounting and unmounting 
>> of the BE could have caused this. To fix the issue
>> try the following steps:
>>
>> if rpool/ROOT/opensolaris-8/opt is not currently
>> mounted:
>> zfs set mountpoint=/opt rpool/ROOT/opensolaris-8/opt
>> zfs mount rpool/ROOT/opensolaris-8/opt
>>
>> if rpool/ROOT/opensolaris-8/opt is mounted you'll
>> need to unmount it before 
>> cleaning up the mountpoint property using the
>> commands listed above.
>>
>> -evan
>>

_______________________________________________
indiana-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/indiana-discuss

Reply via email to