jimbojones21 Posted January 5, 2017 Share Posted January 5, 2017 I recently had to replace my failing 2TB hard drive with a 4TB hard drive. I had some issues with unraid freezing and shares becoming inaccesible, so I decided to update unraid, and also change the formatting of the new drive from reiserf to xfs. I think I messed up that part, as I didn't back up the data on that drive before switching the file system. After switching the file system, I had to reformat the drive to initialize it. Now the drive is empty and I'm missing the files that we're on it. Question: Did my parity drive get updated after I reformatted? I'm wondering if I can recover the HD contents from the parity drive? Quote Link to comment
John_M Posted January 5, 2017 Share Posted January 5, 2017 Formatting an unRAID data disk has the same effect as formatting any hard disk - it writes a new empty filesystem onto the disk. So yes, in short, you've lost everything on that disk. Answer: Your parity drive got updated at the same time as you were reformatting the data drive. No you certainly can't recover the lost data from the parity drive. You might have two possible chances of recovering some of your data. 1. If your old failing drive can still be read you might be able to recover some files from it. 2. You don't say what exactly you did between deciding to replace the drive and finding you'd lost your files. If you rebuilt the contents onto the new drive before your attempt at converting the file system, then you might be able to run the ReiserFS repair tools on it and recover some or even most of your files. Reformatting a disk doesn't actually delete everything on it, just certain key areas, so most of the files will still be there but inaccessible. The ReiserFS repair tools are very good at repairing even very badly corrupted disks, so you might be in with a chance. Quote Link to comment
garycase Posted January 5, 2017 Share Posted January 5, 2017 The simple answer is Yes, you've lost your data -- hopefully you have good backups and can simply restore the missing data from those. In the absence of backups, your best hope is that the old 2TB drive is readable ... at least enough to run reiserfsck against it and recover most, if not all, of your files. Alternatively, as John M noted, you MAY be able to run reiserfsck against the new, reformatted 4TB drive and it MAY find some of the old Reiser files. But the original 2TB drive is your best hope if you don't have backups. Note that NO RAID system eliminates the need to back up your data. It always amazes me that folks consider their data important enough to build a fault-tolerant server; but don't bother to backup that data. Quote Link to comment
jimbojones21 Posted January 6, 2017 Author Share Posted January 6, 2017 Thanks for the suggestions. I'm usually careful about doing stuff with unraid, but I got a little too clicky with my mouse and cost me. I'm pretty sure my old drive is still readable, so I'll try to recover the files from that. Not sure why, but it looks like unraid is doing a parity sync/data rebuild now. It says it will take 15 hours to complete. In the meantime my 4TB drive is showing an orange triangle "device contents emulated". I assume it will stay that way until the check is completed? Also probably not a good idea to try to dump new files to the array until it's completed? As for backups, I do store extra copies of my really important stuff. I can live without most of the content on my unraid server. It's more of a hassle to get everything back (or even know what files I lost!) Quote Link to comment
JorgeB Posted January 6, 2017 Share Posted January 6, 2017 Although the disk is empty it still needs to be rebuilt, let it finish before starting to copy the data back. Quote Link to comment
jimbojones21 Posted January 7, 2017 Author Share Posted January 7, 2017 So I was able to recover my files from my old 2TB drive, but unfortunately I'm still running into a problem with my server locking up. For example, as I transfer files to the server, Windows will give me an error "An unexepcted network error occured". If I go to the unraid web interface, it will not be responsive. I had to hard reboot and the server ran fine for a bit, but again locked up after starting the transfer again. I managed to get the diagnostics. Is the problem related to the new hard drive? Any ideas? tower-diagnostics-20170107-0927.zip Quote Link to comment
trurl Posted January 7, 2017 Share Posted January 7, 2017 SMART looks OK. syslog is just after a reboot so nothing to go on. You might try putting Fix Common Problems in Troubleshooting mode. Quote Link to comment
jimbojones21 Posted January 12, 2017 Author Share Posted January 12, 2017 The server keeps freezing up, so I installed fix common problem in troubleshooting mode. I'm attaching my syslog.txt. Would it also be helpful to view the diagnostic zip files? Looks like there is one generated every hour, but the last one was last night so I'll post that. I can also post the others if that's helpful. syslog.zip tower-diagnostics-20170111-1931.zip Quote Link to comment
Squid Posted January 12, 2017 Share Posted January 12, 2017 Processes keep getting killed due to Out Of Memory issues. Looks like you've only got 2GB of memory and are trying to run Plex within that. Quote Link to comment
jimbojones21 Posted January 12, 2017 Author Share Posted January 12, 2017 Yes that's true, but I've been running plex for many years on 2GB ram without issue. Quote Link to comment
trurl Posted January 12, 2017 Share Posted January 12, 2017 Yes that's true, but I've been running plex for many years on 2GB ram without issue. Was that on v5? Quote Link to comment
Squid Posted January 12, 2017 Share Posted January 12, 2017 Yes that's true, but I've been running plex for many years on 2GB ram without issue. Didn't really catch whether you were running Plex as a docker or a plugin, but docker does have some overhead vs a straight plugin like you would have been using had you been on v5 Also, given identical source code, the 64bit compiled executable will by definition be larger and consume more memory. Quote Link to comment
jimbojones21 Posted January 12, 2017 Author Share Posted January 12, 2017 v5 and also v6. I upgrade to v6.1.9 around March of last year and also started using the docker for plex. I uploaded to v6.2.4 only after I started having issues. Quote Link to comment
Squid Posted January 12, 2017 Share Posted January 12, 2017 Other rule of thumb (not a given however) is that later versions of unRaid (and the kernel) will always inherently consume more memory for their own purposes than prior versions. 2GB on unRaid is suitable for a basic NAS. 4GB is the fundamental minimum for running any thing else on the server. 8GB would be the recommended. Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.