Rescued Raid... Now What?

[mneimeyer]

I think I've rescued my mdadm based raid... I'm copying everything off now (so I don't have to pull it all down from my cloud backup).

Is there anything I can do to see what got lost/is damaged? Either from damaged parts in the raid itself OR damaged parts of the file system? (If anything? Everything I've copied off so far seems healthy... )

Thanks!

P.S. If it matters I probably did something stupid along the way... Friday I got a notice of SMART read errors on drive 1 of a six drive RAID5 array. When I logged in to remove that drive from the array (to replace it) I noticed that drive 2 had dropped out for some unknown reason in the past. I added back drive 2 and when it was done re-syncing I had drive 1 dropped out and drive 2 was now a spare. A second forced assemble failed. Assuming that SMART read errors were the problem I used the replacement drive and used ddrescue to copy the bad disk 1 onto that (ddrescue claimed only 114k unrecoverable). When I plugged back in the new drive 1 and tried to force assemble I got bad superblock errors. Following the steps here https://raid.wiki.kernel.org/index.php/RAID_Recovery I did a recreate/assume-clean instead of an assemble (leaving out drive 2) and I now have an apparently working array with 5 out of 6 drives.

When I'm done copying I need to add back in drive 2 (but not as a spare) and check the file system... Fundamentally, I'm wondering if there is any good way to find out "the errors were in files X, Y and Z" so that I can pull down clean copies of JUST those files? Again, sparing me the need to pull down the entire backed up set.

[eccerr0r]

If you ever make assumptions like "assume clean" all bets are off. I think you may have to either use your backup for consistency or at least use checksums to tell what was inconsistent but "assumed" consistent.

I despise internet backup systems for disk failures...
_________________
Intel Core i7 2700K/Radeon R7 250/24GB DDR3/256GB SSD
What am I supposed watching?