On 12/6/18 11:06 AM, Brad Campbell wrote: > On 6/12/18 10:33 pm, Niklas Hambüchen wrote: >> On 2018-12-04 01:27, Brad Campbell wrote: > Puzzling also as to why md didn't re-write that sector when it found a > read error. I have it do that from time to time on RAID-6. Hmmm. I just went back through the thread and I don't see your partition tables posted anywhere. Is there any chance this array predates 4k sector drives and has been upgraded to 4k physical/ 512B logical drives? With a non-aligned partition? That would cause MD's sector re-writes to fail, as the drive would have to turn them into read-modify-write cycles across two actual sectors. BTW - your smartctl gists are gone, so I can't go back and look at the drive details. In the future, paste such text files right into your email (in non-wrapping plaintext mode) so the archives will be complete. This list accepts messages up to about 100k. Phil
