Jan 8 12:28:54 sopa kernel: [ 13.552491] BTRFS info (device dm-0): disk space caching is enabled Jan 8 12:28:54 sopa kernel: [ 13.803006] BTRFS info (device dm-0): bdev /dev/mapper/sopa-data errs: wr 420, rd 44, flush 0, corrupt 0, gen 0 Any idea what these are and if they're recent or old? This is the original device with all the data and there are read and write errors. Write errors mean writes have been dropped. It's usually an indicator of a device problem. Jan 8 12:59:40 sopa kernel: [ 229.560937] Alternate GPT is invalid, using primary GPT. Jan 8 12:59:40 sopa kernel: [ 229.560954] sdb: sdb1 sdb2 sdb3 This suggests something overwrote the backup GPT on /dev/sdb. This probably isn't related to the problem you're having, because it comes before you add /dev/sdb3 to the Btrfs volume, and doesn't look like Btrfs has written anything to /dev/sdb3 at this point. But whatever did this should be tracked down, and in the meantime the backup GPT should be fixed. Most anything will fix it these days, I send to use gdisk. Jan 10 08:16:12 sopa kernel: [156022.305629] BTRFS info (device dm-0): relocating block group 2051006267392 flags data Jan 10 08:16:23 sopa kernel: [156033.373144] BTRFS info (device dm-0): found 75 extents Last line repeats hundreds of times, over the course of hours. Even if it's highly fragmented, I wouldn't expect it takes this long to read 75 extents. The differential is reasonable in some cases if these are small 4K file, but others, it's suspicious: Jan 10 12:10:14 sopa kernel: [170063.613418] BTRFS info (device dm-0): found 75 extents Jan 10 12:10:44 sopa kernel: [170093.813300] BTRFS info (device dm-0): found 75 extents But no errors, for many hours. I can't really make much out of the task list, whether Btrfs is the cause or a victim. --- Chris Murphy
