Hello Wolfgang, thanks for your reply! On 2018-12-03 11:49, Wolfgang Denk wrote: > Did you notice this part of the error logs: > > sd 2:0:0:0: [sdb] tag#2 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE > sd 2:0:0:0: [sdb] tag#2 Sense Key : Medium Error [current] [descriptor] > sd 2:0:0:0: [sdb] tag#2 Add. Sense: Unrecovered read error - auto reallocate failed > -----------------------------------------------------------------^^^^^^^^^^^^^^^^^^^^^^ > > This typically happens if the disk has already a large number of > sectors reallocated and is running out of spare sectors. This makes sense, but oddly, the reallocation count for the disks is 0 (see below). > Do you watch your disks for critical SMART entries like > Reallocated_Sector_Ct, Current_Pending_Sector and > Offline_Uncorrectable on a regulkar base? > > What are these numbers for your sdb drive? Yes, I watch for those, they are all shown as 0 in smartctl. The full output can be seen here for both disks: https://gist.github.com/nh2/db886f3afbbb4b186aa5088ca2782c06#file-smartcl-a-output-txt I also included `smartctl -xall` output, in case it gives more insights. Note the failed "Short offline" self checks are from when I ran `smartctl -t` after mdadm eventually sent me an email about the failed write. Niklas
