| 09:28 | Ke> | total bytes scrubbed: 4.02TiB with 1 errors |
| 09:29 | vpeter> | Still not good? |
| 09:31 | Ke> | well there was one error, not sure where |
| 09:31 | Ke> | but it's very good in the sense that it has been able to read 4TiB (I think actually might be 8TiB) without crashes |
| 09:32 | Ke> | my array is composed of inexpensive leftover disks past warranty dates |
| 09:32 | Ke> | well I have one disk that cost more than 300€ |
| 09:32 | vpeter> | But error is error. Can't be sure everything is ok. |
| 09:33 | Ke> | yup, though that issue is likely not the same as the rcu issue that was very frequent |
| 09:35 | Ke> | vpeter: also this is btrfs raid5 which has write holes, so certain amount of IO errors are possible without software malfunction |
| 09:37 | Ke> | btrfs concept of IO error afaik includes checksum errors |
| 09:37 | Ke> | at least userland seens EIO |
| 09:37 | Ke> | -n |