Error detection/correction and fault detection/recovery – Data processing system error or fault handling – Reliability and availability
Reexamination Certificate
2000-04-17
2003-08-12
Baderman, Scott (Department: 2184)
Error detection/correction and fault detection/recovery
Data processing system error or fault handling
Reliability and availability
C714S006130
Reexamination Certificate
active
06606714
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to disk drives for computer systems. More particularly, the present invention relates to using a defect list to perform uninterrupted writing of non-relocated data blocks and caching of relocated data blocks for deferred writing.
2. Description of the Prior Art
Disk drives typically employ some form of defect management for mapping defective sectors in data tracks recorded on a disk to spare disk locations. As illustrated in
FIG. 1A
, a data track
4
may comprise a predetermined number of disk locations, where a predetermined number of the disk locations are spare disk locations. The example data track
4
of
FIG. 1A
comprises twelve disk locations, the last three of which are spare disk locations. The disk will also typically include a plurality of spare tracks comprising only spare disk locations as illustrated by the spare track
6
shown in FIG.
1
A.
The disk locations in a track are identified by a physical block address (PBA), for example, a consecutive sequence of numbers. A host computer connected to the disk drive will provide a logical block address (LBA) to the disk drive in connection with a request to write or read data. The disk drive translates the LBA into the corresponding PBA which facilitates defect management in a manner that is transparent to the host computer.
One known method for defect management is to “push down” LBAs past defective disk locations that are discovered during manufacturing of the disk drive. When a defective disk location is discovered during normal operation (i.e., a “grown defect”), the LBA associated with the defective disk location is mapped to a spare disk location in one of the spare tracks. This defect management technique is illustrated in
FIG. 1A
which shows a data track
4
comprising a defective disk location discovered during manufacturing, and three defective disk locations discovered during normal operation. The defective disk location discovered during manufacturing has a PBA of
2
and therefore LBA
2
is assigned to PBA
3
. The remaining LBA's are pushed down one disk location such that the last LBA (LBA
8
) is assigned to the first spare disk location in the track having a PBA of 9. During normal operation, disk locations with PBAs of
4
,
7
and
8
became defective and therefore the corresponding LBAs
3
,
6
and
7
were mapped to spare disk locations of a spare track
6
. The data blocks stored in the spare disk locations of the spare track
6
are referred to as “Relocated Data Blocks” (RDB), and the data blocks stored in the sectors of the data track
4
are referred to as “Non-Relocated Data Blocks” (NRDB).
The spare tracks are typically banded together on a reserved area of the disk, for example, at the inner diameter of the disk. This is illustrated in
FIG. 1B
which shows a disk drive
1
comprising a disk
2
including the data track
4
and the spare track
6
of
FIG. 1A
, wherein the spare track
6
is located at an inner diameter of the disk
2
.
The disk drive
1
of
FIG. 1B
comprises a head
8
connected to an actuator arm
9
which is actuated radially over the disk
2
, and disk control system
20
for executing the steps of a program to perform read and write operations. During a write operation, the disk drive
1
receives from a host a write command to write data blocks to the disk
2
, typically in a consecutive sequence of LBAs. For example, the disk drive
1
may receive a request from the host to write LBAs
0
-
8
in FIG.
1
A. The corresponding data blocks received from the host are stored sequentially in a cache
16
using a cache data structure
17
. The disk control system
20
also employs a defect list
18
which stores information to facilitate translating the LBAs into the corresponding PBAs and then controls the head
8
to seek to the appropriate tracks. The data blocks are retrieved from the cache
16
using the cache data structure
17
, and then written to the disk
2
via a write channel circuit
12
once the head is positioned over the target sector.
For example, when writing the LBA sequence
0
-
8
of
FIG. 1A
, the disk control system
20
positions the head
8
over data track
4
, sequentially retrieves NRDB
0
and NRDB
1
, from the cache
16
using the cache data structure
17
, and then sequentially writes NRDB
0
and NRDB
1
to PBA
0
and PBA
1
via the write channel
12
. Because LBA
2
has been pushed down to PBA
3
(as mapped by the defect list
18
), the disk control systems
20
skips over PBA
2
and writes NRDB
2
to PBA
3
. Because LBA
3
is mapped to PBA
0
of spare track
6
, the disk control system
20
performs a seek operation to position the head
8
over spare track
6
in order to write RDB
3
. The disk control system
20
then performs another seek operation to reposition the head
8
over data track
4
in order to write NRDB
4
and NRDB
5
to PBA
5
and PBA
6
. Because LBA
6
and LBA
7
have been mapped to PBA
1
and PBA
2
of spare track
6
, the disk control system
20
performs another seek operation to spare track
6
in order to write RDB
6
and RDB
7
. Finally, the disk control system
20
performs yet another seek operation to reposition the head
8
over data track
4
in order to write NRDB
8
(to the first spare disk location of data track
4
).
Seeking to a spare track during write operations in order to write relocated data blocks can increase significantly the access time of the disk drive
1
. If the disk drive
1
cannot keep up with the data transfer rate of the host, the cache
16
will eventually fill up with data. The disk control system
20
will then signal the host to stop transferring data until more room is available in the cache
16
. Thus, the write efficiency of the disk drive
1
is limited by the prior art technique of seeking to spare tracks during write operations to write relocated data blocks.
There is, therefore, a need to improve write efficiency in a disk drive, particularly with respect to writing relocated data blocks to spare tracks.
SUMMARY OF THE INVENTION
The present invention may be regarded as a disk drive for receiving a write command from a host computer, the write command comprising a plurality of logical block addresses (LBAs) and a plurality of data blocks. The plurality of data blocks comprises at least one non-relocated data block (NRDB) and at least one relocated data block (RDB), and at least one of the LBAs corresponds to the NRDB and at least one of the LBAs corresponds to the RDB. The disk drive comprises a disk comprising a plurality of first disk locations each having a defect, a plurality of second disk locations for storing NRDBs, and a plurality of spare disk locations for storing RDBs. A defect list is used to map LBAs from the plurality of first disk locations to the plurality of spare disk locations. A cache is logically partitioned into a first set of memory locations managed using a first cache data structure, and a second set of memory locations managed using a second cache data structure. A disk control system uses the first cache data structure to store the NRDB in a first memory location selected from the first set of memory locations. The disk control system evaluates the defect list using the LBAs to identify the RDB and using the second cache data structure stores the RDB in a second memory location selected from the second set of memory locations. During a first time interval established for writing NRDBs, the disk control system uses the first cache data structure to write the NRDB to one of the second disk locations. During a second time interval established for writing RDBs, the disk control system uses the second cache data structure to write the RDB to one of the spare disk locations.
The present invention may also be regarded as a method of performing a write operation in a disk drive in response to a write command received from a host computer, the write command comprising a plurality of logical block addresses and a plurality of data blocks. The plurality
Baderman Scott
Damiano Anne L.
Shara, Esq. Milad G.
Sheerin, Esq. Howard H.
Western Digital Technologies Inc.
LandOfFree
Disk drive using defect list to perform uninterrupted... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Disk drive using defect list to perform uninterrupted..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Disk drive using defect list to perform uninterrupted... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3111009