On 21/12/18 16:05, Alexis BRENON wrote: > Ok, so I will try to force assemble my array. > However, smartctl -x reports: > 5 Reallocated_Sector_Ct PO--CK 100 100 036 - 38 > So my disk is probably not as clean as expected, isn't it? Yup. That looks a little worrying. You'll get people say "oh that's not much", and if that doesn't change in a month or two's usage they're probably right, but if you don't monitor those errors and have a long term feel for them it could all go pear-shaped very quickly. I wouldn't trust that drive unless I was confident that figure wasn't changing much. Force-assembling the array ... if you can copy that disk, use the original to force-assemble, because it will force a read error and a raid recovery. If you used the copy, you will get a blank sector and no read error to force a recovery which will likely corrupt your data. (And if things go wrong, it probably won't make much difference whether you're trying to recover from the original or a copy). If you can, when swapping out the WD and the dodgy Seagate, use the --replace syntax for mdadm, it'll reduce stress on the array, and keep redundancy. Cheers, Wol
