[OpenIndiana-discuss] zpool errors a bit confusing

[oi.mail-list at gmx.com]

On 4/5/2014 9:25 AM, Harry Putnam wrote:
> "oi.mail-list at gmx.com"  <oi.mail-list at gmx.com>  writes:
>
> [...]
>
>>> >>Can anyone else report having seen something like this?
>>> >>
>>> >>Can the fact of my setting all four of my current pools to be
>>> >>mountpointed under /p  have any bearing here?
>>> >>
>> >          You can try  doing this to see if it is able to  correct the
>> >problem:
>> >http://docs.oracle.com/cd/E19253-01/819-5461/gbbwl/index.html
> Looking at the examples shown there as types of errors one might get
> using `zpool status', I notice NONE of the examples show the
> corruption being listed as a directory.  All examples show actual end
> of the line filenames, accept those where the error says the pool is
> FAULTED.
>
> Have you ever seen a 'zpool status' output that showed only directory
> names as the corrupted 'files'?
>
> Is that unusual?
>
> I have no experience on which to draw.
>
>> >          My opinion is this  problem most likely occur because it was
>> >lacking redundancy.
>> >          and scrub was not often run.
> Can you specify what you actually mean by 'lacking redundancy'.
>
> And for the record, one of the pools showing the posted output was
> only 2 days old, so no... there had not been major scrubbing.....
>
> Even the oldest one showing problems was less than a week old.
      A fast summary,  ZFS redundacy; basically is when bad data is 
found by zfs  it fetches a clean copy of the data from another 
disk(mirrored) and repairs it with the correct data,etc.

         If you have not already read this, I suggest reading "Oracle 
Solaris Administration ZFS File system"
http://docs.oracle.com/cd/E23824_01/html/821-1448/gbscy.html#scrolltoc

        I'm reading it myself and there are some things that are still 
not clear:-)