Hi guys,

 

I'm facing a similar issue to the OP of this issue (https://lime-technology.com/forum/index.php?topic=39980.0), in that after a swap disable, the parity check is finding hundreds of millions of errors in the parity disk area beyond what was the original parity disk size (4-8tb, and I have no data disk bigger than 4tb). It is estimated that it will take 13 days to complete the exercise!

 

I did a swap-disable due to a dead drive and a replacement that was larger than parity... I replaced the 4tb parity with an 8tb drive. Note however that I did do a preclear on the 8tb first, so the comments in that post shouldn't apply, because it should already be zeroed...?

Fyi I did the swap disable on unraid 5.0.5... And all of my drives are still on reiserfs. And I also did have a problem when parity swapping initially, it got to 99% and stopped, so I had to rerun it in safe mode... But that should not have changed any of the data in the 4-8tb range.

 

If I assume that there is no data in the 4-8tb range, and that there are no errors in the core data range, am I better off stopping this parity check process and just rebuilding parity from scratch? Or do I just wait the two weeks for this prices to complete?

 

Fyi I posted my original swap disable issue here: https://lime-technology.com/forum/index.php?topic=45998

There’s been more than one report of this happening recently, another example, from today:

 

https://lime-technology.com/forum/index.php?topic=46090.msg441080#msg441080

 

Maybe something LT should take a look at.

 

And this reminds me that I just recently did this procedure and did not yet do a parity check, will report if I find the same issue.

 

Thanks Johnnie, I ended up stopping the process and rebuilding the parity drive from scratch (2 days to do this rather than the estimated 10-12 to 'fix parity'). I then did a final parity check and all good, no errors...

 

Not sure if you've run swap disable process yourself yet to see if you have entered the same issue? I'd be interested to know if this is something we need to log with LT, or if it may have been related to my swap disable done on an older version...

There’s been some reports of this happening, but it’s certainly not a general issue, I just did 2 swaps recently and parity check finished with 0 sync errors, maybe only happens in some configuration/circumstances, on the other hand, I don’t see how this could be user error, because the rebuild only starts after a successful parity copy, best chance of finding out what’s happening is that someone with this issue saves the logs.