February 22, 201610 yr Hopefully reiserfsck will let you recover at least some of your data, since I gather you don't have backups. You made TWO significant mistakes ... (1) You tried to rebuild a failed drive onto the same bad disk drive ... so, of course, it failed again as soon as you tried to use it. You should, of course, have got a NEW drive and done the rebuild onto that. (2) As you now understand was a mistake, you formatted the drive as XFS ... which wipes out the file system on the drive => and updates parity to reflect those changes. So if you rebuild a drive now, it will simply have that nice new EMPTY XFS file system on it. Parity does NOT care about the file system -- in fact it has no idea what the file system is; it simply updates the parity information to reflect whatever's written to the disk ... whether that's new headers for a format or updated data being stored on the disk. As you also know now, reiserfsck MAY be able to recover the old data, since formatting only writes the file system, so most of the old data is still on the disk. I gather you're now running reiserfsck on the emulated disk. That's fine => just let it complete, and when it's done REPLACE the disk and let it rebuild. This will be a Reiser disk with your recovered data on it. Do NOT format that disk !! You can, if you want to convert it to XFS, first copy ALL of the data you have recovered to another location [elsewhere on your array or to another PC ... but be CERTAIN you don't do anything that will result in the "user share copy bug" (if you don't understand that, ASK)] Once there is NO data on the disk that you need, you can simply Stop the array; change the file system for that disk to XFS, and then let it format the drive.
February 22, 201610 yr Community Expert ...since this was a mistake i have formatted back to reiserfs and runned... I hope you mean you changed the file system back to reiser, no need to format anything, although if you did format it again it probably won’t make much difference in the recovery.
March 21, 201610 yr Author i have updated news. the 5gb drive when received it seemed used already opened broken shield so went for replace had to wait few days. what i have recovered was partial but with kodi library i was able to identify the files. is going to be a slow recover from now on. I have also turned almost all my drives to xfs 3 more to go. all seems good but unraid 6 problems also seem not to end. To format the drives to xfs i copied the files first on another pc with no issues. now that i try to write them back i take copy stall(timeouts), meaning copy process starts but it stalls for no apparend reason. i will post here tower diagnostics if anyone can help what is going on(if not i will make new post for more to see) tower-diagnostics-20160321-2003.zip
March 21, 201610 yr Community Expert Diagnostics were taken after a reboot, so there are no errors on the syslog, but this disk does not look good: Device Model: WDC WD10EACS-65D6B0 Serial Number: WD-WCAU42094115 197 Current_Pending_Sector 0x0032 198 192 000 Old_age Always - 372 198 Offline_Uncorrectable 0x0030 199 194 000 Old_age Offline - 317
March 21, 201610 yr Community Expert Just noticed that the disk is not part of the array, post new diagnostics after trying to copy something and getting timeouts.
March 21, 201610 yr Author Just noticed that the disk is not part of the array, post new diagnostics after trying to copy something and getting timeouts. ok tried to copy 120gb of big movies files direct to disk8 from another pc on the network when the copy reached 22gb it timeouts ,stalls i attached new diagnostics file tower-diagnostics-20160321-2119.zip
March 21, 201610 yr Author tried the same files to copy to disk2 now that is stills reiserfs(disk8 is xfs), copy went till 31gb this time before it stops.......no idea what is going...system log seems ok
March 21, 201610 yr Community Expert There's still nothing on the logs, do you have another pc you can copy from? Ideally using a different switch and network cable.
Archived
This topic is now archived and is closed to further replies.