Just so that I have a clear understanding of this scenario, it sounds like when JB said:

 

"the best option here would likely have been to re-enable the disk, instead of rebuilding, in some cases when a disk gets dropped filesystem might get some corruption on the emulated disk."

 

that he did not mean to use the "parity is already valid" function (as I first thought) but rather the FS repair should have been done first, and then the re-enable and rebuild procedure should have followed. Assuming of course that the preferred option referred to by johnathanm, to rebuild to a different disk, wasn't on the table. Hopefully I have this right now, it does seem to make more sense...

Edited March 2, 20206 yr by Dissones4U

2 hours ago, Dissones4U said:

Just so that I have a clear understanding of this scenario, it sounds like when JB said:

 

"the best option here would likely have been to re-enable the disk, instead of rebuilding, in some cases when a disk gets dropped filesystem might get some corruption on the emulated disk."

It depends on the situation, IIRC in this case the emulated disk was unmountable, and the user rebuilt on top of the old disk without fixing the filesystem first, this isn't recommended, first make sure the filesystem can be repaired and the repaired emulated disk looks to have the correct data, only if that looks OK then go ahead and rebuild on top, if the filesystem isn't repairable or there are a lot of lost+found files, then it's worth seeing if the dropped disk mounts and/or has a less corrupt filesystem, like mentioned any writes made after the disk got disabled will be lost, but this only applies to a mounted filesystem.

5 hours ago, johnnie.black said:

first make sure the filesystem can be repaired and the repaired emulated disk looks to have the correct data

 

This definitely adds some clarity too, thank you JB. I've been trying to understand file system corruption better because I've noticed that there have been several users who have encountered it lately and most of them seem to mishandle the recovery. I think that part of my confusion was not understanding emulated disk corruption vs physical disk corruption, but I think I'm starting to get it. The key being emulation has to be verified before anything is done and then knowing that the physical disk is probably bad but the smart data will help to determine whether or not it can be salvaged. I think I'll prepare a new disk that is specifically for rebuilding disabled / emulated disks as soon as I can afford it. Then, best case scenario, the old disk can be reused once it's been thoroughly tested. All of you guys were great in clarifying this, thank you.