growlith Posted January 13, 2020 Share Posted January 13, 2020 My server has a collection of 4tb and 8tb drives. When I opened it up to add another 8tb to start double parity, I moved one of the 4tb drives to keep the 8tb next to each other. When I booted back up, the 4tb would not show up. After trying various drive locations, and then checking to see if the drive was recognized on another computer, I decided to just use the new drive to rebuild the now not working one. The rebuild seemed very strange because it was running at 3.5Gbps which does not seem possible. When I checked back after it was complete, another 4tb drive got an "UDMA CRC error count" of 13 and was disabled (I don't know if this happened during the rebuild or not). Stopping the array and starting it again triggered another rebuild onto the error drive Long story short I now have 2 drives showing "Unmountable: No file system" with the 4tb one maybe having data on it. Currently running the xfs_repair checks on both drives through the GUI with the array in maintenance mode. Is there any way to get the 4tb drive with write errors to be attached to the array long enough to build the missing drive and then I can replace it with another new drive to rebuild that? The timing is unfortunate because I was just thinking about how I should be better protected and was going for double parity and actually pricing out a less powerful server to ship to the parents to back myself up offsite. Lesson learned I guess. tower-diagnostics-20200113-0551.zip Quote Link to comment
JorgeB Posted January 13, 2020 Share Posted January 13, 2020 The syslog doesn't cover start of the problem, but during disk2's rebuild there were multiple disk errors which suggests there was another invalid disk, I assume disk1 since it didn't mount, this means disk2 rebuild will be totally corrupt, same for the rebuild of disk1 after that. Best bet to recover data from both disks would be to get the old disks working and then do a new config. Quote Link to comment
growlith Posted January 14, 2020 Author Share Posted January 14, 2020 Thanks for looking at it for me. Recovering the disk2 data proved to be beyond my ability, but I was able to confirm that I did not lose any important data just a whole bunch of media. I got lucky. All of my docker data is on the cache drive, and I had just moved my personal files in bulk around and they were all placed on the same drive together (a drive which survived). So I just formatted those two disks and created a new config. Things are up and running now and I am in the process of a parity build for a second parity drive. Now I need to let my family know that I lost a lot of stuff from plex that will take a bit to get back and that their plex user history is going to be lost (the plex container is the only one with data on the array apparently even though /mnt/docker/ is cache only). 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.