Parity Questions - Drive size upgrade vs actual parity check?

[FrozenGamer]

The parity check history calls this a parity check and labels parity as valid.  If i did a parity rebuild to upgrade a drive from 8 to 14 is this effectively a parity check.  Also what would happen to data if a parity was not valid and a rebuild was done of a failed drive or drives? I do have 2 parity drives.  I do understand that if 3 drives failed i would lose the data on failed drives only, which would be kind of a bummer since i assume that the fill drive method has data scattered, ie incomplete collections/albums etc.  Just curious. Thanks in advance.

[itimpi]

3 hours ago, FrozenGamer said:

The parity check history calls this a parity check and labels parity as valid.

What is ‘this’?

 

3 hours ago, FrozenGamer said:

If i did a parity rebuild to upgrade a drive from 8 to 14 is this effectively a parity check. 

No.   It would be a parity sync.    You can only do a parity-check if you already have parity.

 

3 hours ago, FrozenGamer said:

Also what would happen to data if a parity was not valid and a rebuild was done of a failed drive or drives?

The rebuilt drive would have corrupt data on it.

[FrozenGamer]

On 2/11/2023 at 9:49 AM, itimpi said:

What is ‘this’?

Sorry i may not have been very clear - i did a data rebuild of an 8tb drive to 14 (to make more space) - this is what the main page states. "

Last check completed on Mon 06 Feb 2023 01:20:25 AM AKST (seven days ago)
 Duration: 4 days, 13 hours, 22 minutes, 15 seconds. Average speed: 35.6 MB/s
 Finding 0 errors"

 

No.   It would be a parity sync.    You can only do a parity-check if you already have parity.

Yes i did have parity, or at least it was stated that I did and have no reason to believe I didn't. Thanks.

 

On 2/11/2023 at 9:49 AM, itimpi said:

 

 

[trurl]

On 2/11/2023 at 10:26 AM, FrozenGamer said:

assume that the fill drive method has data scattered

Perhaps you mean Highwater (recommended and default) instead of Fillup.

[FrozenGamer]

On 2/13/2023 at 7:47 AM, trurl said:

On 2/11/2023 at 6:26 AM, FrozenGamer said:

assume that the fill drive method has data scattered

Perhaps you mean Highwater (recommended and default) instead of Fillup.

Yes

[FrozenGamer]

So I should mark this as solved, but I am still not clear on the following example (which is my current situation.  Let me know if i have it right. If i have done a previous parity check, a month ago with zero errors, i upgrade an 8 terabyte drive to 14tb in the array, the history of parity checks now says that it has done a parity check (from the data rebuild) with zero errors. 

 

I now have parity unless parity was lost in the 1 month between last parity check and the data rebuild which rebuilt based on inaccurate data? In which case it says i have parity, which i do, but my data may not actually be perfect.. 

Or i have parity, because i just completed a parity check when i did the data rebuild and my data should be considered accurate?

 

I probably shouldn't worry about this, but i have been curious about this since parity checks are really slow on my machine and prefer not to bog the system down for 3 or 4 days until next month.

Thanks for all the replies and help

 

[Kilrah]

A rebuild is not a check. If parity was inconsistent when a rebuild is done then that can't be detected (at least not with one parity drive).

Edited February 18, 2023 by Kilrah

[FrozenGamer]

Thanks everyone! Sorry if my question was not written clearly enough.