January 8, 20206 yr Community Expert 3 minutes ago, johnnie.black said: Do you mean on the emulated disk or the actual disk mounted outside the array? Just now, TheTechnoPilot said: On the actual disk when I put it back into the array via new configuration (trusted parity). So neither the emulated disk nor the actual disk mounted outside the array?
January 8, 20206 yr Author 1 minute ago, trurl said: There were 8 hours between the post you quoted and your post. The post you quoted suggested some possible approaches. You don't mention if you actually did anything during those eight hours. Did you just examine the drive and discover you might be missing something, or did you do more than that? Those 8 hrs I slept, I followed the new configuration, trusted parity instruction as the only change from having a functional full array to an unfunctional one last night was a reboot and accidental array start while seemingly missing a disk (not using ControlR to start the array again since I didn't pickup on the issue in its interface).
January 8, 20206 yr Author 1 minute ago, trurl said: So neither the emulated disk nor the actual disk mounted outside the array? No, however, considering the available free space listing of the disk outside the array (before I added it back), it now seems wrong when I think back as it had 2.9TB free when it really should only be at about 1.3TB and this is the same when mounted in the array. Honestly with all that I have done in the last couple days though, having yet done a full parity check, I worry the existing parity is probably already wrong at this point. This is also why I preferred to bet on the original driving being intact.
January 8, 20206 yr Community Expert It would be very weird, if not extremely unlikely, that if you used the old disk there's data missing, assuming it mounted directly without running xfs_repair.
January 8, 20206 yr Community Expert Earlier in the thread you mentioned having a backup of the disk. So that seems like the way forward.
January 8, 20206 yr Author 22 minutes ago, johnnie.black said: It would be very weird, if not extremely unlikely, that if you used the old disk there's data missing, assuming it mounted directly without running xfs_repair. That's what I'm thinking, it seems very odd for the data to be missing and while admittedly I only looked in the share for it (so perhaps not being recognized for being part of that share), the total size of used space on the disk doesn't support unraid thinking it is there in the native disk file system either though... I'm wondering if potentially the filesystem on that drive got damaged and in essence lost the pointing to that folder and no longer considers it used space. That's the only explanation I can imagine for loosing essentially one whole folder from what I can tell. I'm tempted to run a filesystem check on the drive when I next continue my trouble shooting to see if it can repair perhaps such corruption to the directory structure (though I don't want it to go as aggressive as I saw on disk2 if I can help it because the files were so strune about, I totally recovered from backup only and deleted the recovered files). For @trurl who asked, unfortunately I've lost the ability to recover this directory from backup it seems when I checked this morning, because in my migration to unraid, my old build has been offline for too long to recover this directory from my Backblaze backup. 😞 I have taken the array off-line for now and will probably shut-down the server (unless any further diagnostics first would be of help), and then wait till a pack of new locking SATA cables I ordered an hour ago arrive on Friday. I think that's a good first step before I go any further working with the system to eliminate that one area of potential failure (these ones I just used admittedly are what came with the SATA controller I installed an are pretty generic, no-name, non-locking).
January 8, 20206 yr Community Expert Are you sure the missing data isn't on another disk? Easy to accidentally move something to some place you didn't intend, such as under another folder for example.
January 8, 20206 yr Author 3 minutes ago, trurl said: Are you sure the missing data isn't on another disk? Easy to accidentally move something to some place you didn't intend, such as under another folder for example. Honestly not at that folder size since my server should be honestly at just over 30TB (77%) space utilization and was at shut-down, now it is only showing about 28.5TB (74%). So while I wish that was the case, I don't suspect it being a possible explanation unfortunately... Oh and also even if somehow someone (no authorized users though besides myself on the current network) deleted the files on the share, I am also using recyclebin with manual emptying only, so it should still be taking up the space on the array. Edited January 8, 20206 yr by TheTechnoPilot Added details about recyclebin
January 12, 20206 yr Author On 1/8/2020 at 12:18 PM, trurl said: Are you sure the missing data isn't on another disk? Easy to accidentally move something to some place you didn't intend, such as under another folder for example. OMG, @trurl & @johnnie.black I figured out what an idiot I am! I accidentally somehow managed to put the wrong 8TB Barracuda drive into my case and therefore told the system to add what actually was my old Disk2 to the array as Disk4! Okay so that I don't bork everything, please confirm my understanding that at this point is that I should be able to go back through and use New Configuration (parity valid) option to correct this, putting the real Disk4 back into the array and then once that is done, due a full parity check (with write corrections to disk) to get myself back up and running correct? I'm still amazed at my basic stupidity and how I somehow managed to put the wrong drive into the external enclosure... Edited January 12, 20206 yr by TheTechnoPilot
January 13, 20206 yr Community Expert You can do that but if there are many sync errors it could be faster to just do a full parity sync, it depends on how much data changed on the other disk.
January 13, 20206 yr Author 2 hours ago, johnnie.black said: You can do that but if there are many sync errors it could be faster to just do a full parity sync, it depends on how much data changed on the other disk. By that you just mean using new configuration but not flagging it as parity valid, correct?
January 16, 20206 yr Author Thank you so much @johnnie.black and @trurl for all your help! I am back up and running with all my data fully parity synced and now just working on getting it all backed-up through a VM back into the cloud! While initially this all made me feel like my array was fragile, I now know that it is even more robust then I realized and have learned so much for going forward! Thanks again for stepping in and making that the case!
Archived
This topic is now archived and is closed to further replies.