JorgeB Posted December 2, 2019 Share Posted December 2, 2019 You need to remove -n or nothing will be done, and if asked use -L. Quote Link to comment
Acps Posted December 2, 2019 Author Share Posted December 2, 2019 so just do -v sorry just like to clarify to prevent further data loss. Quote Link to comment
Acps Posted December 2, 2019 Author Share Posted December 2, 2019 Disk1 Quote Phase 1 - find and verify superblock... - block cache size set to 1496680 entries Phase 2 - using internal log - zero log... zero_log: head block 3466131 tail block 3425645 ERROR: The filesystem has valuable metadata changes in a log which needs to be replayed. Mount the filesystem to replay the log, and unmount it before re-running xfs_repair. If you are unable to mount the filesystem, then use the -L option to destroy the log and attempt a repair. Note that destroying the log may cause corruption -- please attempt a mount of the filesystem before doing this. Disk3 Quote Phase 1 - find and verify superblock... - block cache size set to 1504136 entries Phase 2 - using internal log - zero log... zero_log: head block 1594967 tail block 1594963 ERROR: The filesystem has valuable metadata changes in a log which needs to be replayed. Mount the filesystem to replay the log, and unmount it before re-running xfs_repair. If you are unable to mount the filesystem, then use the -L option to destroy the log and attempt a repair. Note that destroying the log may cause corruption -- please attempt a mount of the filesystem before doing this. So both disks are still showing Unmountable: No file system so I do run the -L now correct? Quote Link to comment
JorgeB Posted December 3, 2019 Share Posted December 3, 2019 11 hours ago, Acps said: so I do run the -L now correct? Yes Quote Link to comment
Acps Posted December 3, 2019 Author Share Posted December 3, 2019 Ok so that was a success, Disks 1 and 3 are now mounted with their data recovered. The array is without disk 2 which failed and needs to be replaced, that data is definitely gone, or should i just remove the disk and install my replacement drive? Or should i do a parity check first? Quote Link to comment
JorgeB Posted December 3, 2019 Share Posted December 3, 2019 You can still try ddrescue on old disk2, but don't connect it to the SASLP or there's a high probability the driver will crash. Quote Link to comment
Acps Posted December 3, 2019 Author Share Posted December 3, 2019 Ok i have room on my motherboard controller for a slot. Do i do a parity check or just replace disk2 with new drive and assign to slot 2 and rebuild? Quote Link to comment
JorgeB Posted December 3, 2019 Share Posted December 3, 2019 You can't currently rebuild disk2 since there's no disk assigned, you can add a new disk to that slot and it will be cleared first before it can be used, but since parity is invalid because disk2 wasn't assigned you'll need to sync it, and because many errors are expected it would be faster just to do a new parity sync, so might as well add disk2 at the same time, the procedure would be like this: -stop the array -unassign both parity disks -start array -stop array -assign new disk2 -start array, if disk2 is new it will be unmountable, format it (make sure it's the only unmountable disk) -stop array -re-assign both parity disks -start array to begin parity sync Any doubts ask first. Quote Link to comment
Acps Posted December 30, 2019 Author Share Posted December 30, 2019 so iv found out that a 5tb isnt very common and much more expensive. I can get 8tb dribes for the same price. Now i know the parity has to be the largest disk, but my question is will i need 2 8tb for dual parity, then i could use the 2 5tb drives as data drives till i update them at a later time for 8tb? Quote Link to comment
Frank1940 Posted December 30, 2019 Share Posted December 30, 2019 Unraid requires that each parity drive has to be as-large-as or larger than the largest data drive. So if your largest data drive is 5TB, both parity drives would have be at least 5TB drives. (Either one or both drives could be larger.) This would mean that if 5TB drive was your largest data drive, you could have an 8TB drive and 5TB drive as parity drives and Unraid would be happy. Quote Link to comment
trurl Posted December 30, 2019 Share Posted December 30, 2019 Not sure how I missed all this excitement. I'm curious though about the outcome. Why was there nearly a month after johnnie's last instruction and OPs next post? Did you get your array going again? I assume disk2 data was lost. User error is the most common cause of data loss. Quote Link to comment
Acps Posted December 30, 2019 Author Share Posted December 30, 2019 Yea i was able to recover everything but disk 2 with jonnys help. It was definately my fault, i was clicking stuff i didnt understand. Its been a few weeks, cause after i got rest of the array back up i turned off my server for the holidays and was out of town. I was about to order just another 5tb but 8th are the same price right now at 150$ so was gonna upgrade. My case setup can olny ohld 2 parity and 6 data disks, so if i want to add room now i need bigger disks without making a bunch of other changes. I do need to replace my raid controller card tho, ive been told it has alot of issues with unraid so i might do that to avoid future issues too. Quote Link to comment
trurl Posted December 30, 2019 Share Posted December 30, 2019 On the subject of backups. My server has 28TB capacity. I have 4 x 2TB disks formatted NTFS so I can easily read them on Windows. After my monthly parity check, and my monthly desktop image, I mount the "oldest" disk in the rotation with Unassigned Devices to backup those things I consider important and/or irreplaceable. It all fits onto a single 2TB disk. I store the 2 most recent offsite. The rest of my server contents I don't worry about, it isn't important enough and can be probably be replaced if I care to. 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.