I am now running the dd you suggested on the whole disk, which will take a couple hours.
That'll just highlight any other duff sectors that might be after the one that triggers the SMART test failure.
This has now completed.
Output:
# dd if=/dev/sdb of=/dev/null bs=1M conv=noerror
dd: error reading '/dev/sdb': Input/output error
858+1 records in
858+1 records out
900374528 bytes (900 MB, 859 MiB) copied, 17,6413 s, 51,0 MB/s
dd: error reading '/dev/sdb': Input/output error
864+2 records in
864+2 records out
906690560 bytes (907 MB, 865 MiB) copied, 29,761 s, 30,5 MB/s
dd: error reading '/dev/sdb': Input/output error
864+3 records in
864+3 records out
907038720 bytes (907 MB, 865 MiB) copied, 37,393 s, 24,3 MB/s
dd: error reading '/dev/sdb': Input/output error
875+4 records in
875+4 records out
918679552 bytes (919 MB, 876 MiB) copied, 51,737 s, 17,8 MB/s
dd: error reading '/dev/sdb': Input/output error
6026+5 records in
6026+5 records out
6320365568 bytes (6,3 GB, 5,9 GiB) copied, 147,069 s, 43,0 MB/s
dd: error reading '/dev/sdb': Input/output error
7300+6 records in
7300+6 records out
7656632320 bytes (7,7 GB, 7,1 GiB) copied, 179,381 s, 42,7 MB/s
dd: error reading '/dev/sdb': Input/output error
7301+7 records in
7301+7 records out
7657721856 bytes (7,7 GB, 7,1 GiB) copied, 186,602 s, 41,0 MB/s
dd: error reading '/dev/sdb': Input/output error
7301+8 records in
7301+8 records out
7658487808 bytes (7,7 GB, 7,1 GiB) copied, 193,757 s, 39,5 MB/s
dd: error reading '/dev/sdb': Input/output error
7303+9 records in
7303+9 records out
7660621824 bytes (7,7 GB, 7,1 GiB) copied, 201,901 s, 37,9 MB/s
2861578+12 records in
2861578+12 records out
3000586416128 bytes (3,0 TB, 2,7 TiB) copied, 35095,2 s, 85,5 MB/s
smartctl now shows for that device:
197 Current_Pending_Sector 0x0032 200 200 000 Old_age Always - 12
