Alternate-sector remapping
Alternate-sector remapping redirects a logical block from its original physical sector to a reserved replacement location.
The host address can remain unchanged while its physical backing changes. In the cited SCSI manual Rev. J, REASSIGN BLOCKS replaces the backing of listed LBAs but does not preserve their contents; the manual recommends recovering those bytes first, then writing them back. This statement is revision-specific, not a rule for every automatic reallocation path. [1]
Failure boundary: in the same revision, a spare shortage ends the command with CHECK CONDITION and may leave only part of the list reassigned. The first unprocessed LBA is reported when known; FFFFFFFFh can mean that this information is unavailable or that every listed defect was reassigned. The sentinel alone therefore proves neither completion nor rollback. [1]
Relations
Limitations
- Changing the mapping does not itself reconstruct unreadable data; the cited command explicitly separates reassignment from restoring the affected contents. [1]
- A remapping count alone does not determine access latency; the 1994 model needed layout and rotational-position information and omitted individual spared-sector effects. [2]
Citations
- Seagate SCSI Interface Product Manual, 77738479 Rev. J (April 1999), section 5.27, pp. 176-177
- Ruemmler and Wilkes, An introduction to disk drive modeling (1994), Data layout and Conclusions, author PDF pp. 5 and 15
Sources & caveats
- Source review, not a destructive command recipe or device experiment. Spare placement, trigger policy and latency are implementation-dependent.