Note: Descriptions are shown in the official language in which they were submitted.
CA 02307212 2000-04-11
WO 99/23562 PCTIUS98I23152
AUTOMATIC BACKUP BASED ON DISK DRIVE CONDITION
5 Field of the Invention
The present invention relates generally to computer systems, and in
particular to automated backup of disk drive data based on the condition of
the
disk drive.
10 Hard disk drives are complex electro-mechanical devices which can
suffer performance degradation or failure due to a single event or a
combination
of events. Some hard disk drive failures happen quickly and without advance
warning. Such unpredictable failures can be caused by static electricity,
handling
damage, or thermal-related solder problems. Other hard disk drive failures
result
15 from the gradual degradation of the drive's ability to perform. Hard disk
drive
failures result in lost data and lost time to a user trying to recover the
lost data.
One way to protect against data loss associated with hard disk drive
failure to use the Self Monitoring, Analysis and Reporting Technology
(S.M.A.R.T.) The failures that result from the degradation of performance are
20 the type of failures that S.M.A.R.T. is designed to predict. S.M.A.R.T.
capable
devices monitor a variety of information internal to the device to assess
reliability and predict an impending device failure. For example, a S.M.A.R.T.
capable drive might monitor the fly height of the head above the magnetic
media.
If the head starts to fly too high or too low, it is likely that the drive
could fail.
25 Other drives may monitor different conditions such soft error rates which
are
errors that occur sporadically and may not appear on successive attempts to
read
data. The monitoring techniques employed by S.M.A.R.T.-capable drives vary
from one manufacturer to another.
When the S.M.A.R.T. capable drive predicts an impending failure, the
30 drive's S.M.A.R.T. capability makes information available through an
interface
to the disk drive. The information may be presented to a user via drivers and
supporting applications. The information reaches an application that can
display
a warning message to a user. The user is responsible for reacting to the
warning
message as desired. Thus, present devices require the user, after a warning is
CA 02307212 2000-04-11
WO 99123562 PCTIUS98/23152
2
given, to back-up vital data and replace suspect devices prior to data loss or
unscheduled down time.
However, a problem results if the user is not able to respond by backing
up the data before the failure occurs. One such a situation arises on
workstations
5 connected to a network if the user does not have the authority or the
ability to
back-up the data and replace the drive. Failure of the hard disk drive results
in
lost data, lost time and in many cases lost money. Further problems may be
caused when computers are constantly left running, such as overnight, when a
user is not normally monitoring the computer. Several times during normal
10 working hours, the user may also be away from a running computer. There is
a
need for addressing disk drive problems when the user is not available. There
is
a further need for enhancing system reliability when a user is not attending
the
system.
15 Backup of data on a personal computer is automatically initiated in
response to selected information provided by disk drive performance
monitoring.
In one embodiment, performance monitoring capabilities in a disk drive provide
information on potential impending failure or performance degradation. The
information is provided to an application such as a tape backup program. The
20 tape backup program initiates a tape backup of data on the disk drive. The
tape
backup is initiated when the information is representative of predefined or
user
defined states of performance or other conditions which indicate an impending
or
possible failure. The predefined states are defined to allow a normal backup
prior to a predicted failure of the disk drive, and to ensure that the disk
drive has
25 sufficient performance to allow optimal data transfer rates during such a
backup.
In one embodiment, the tape backup program augments information
normally provided by the self monitoring functions by indicating that the disk
drive is being backed up at a particular time, and also indicate status of the
backup and completion. If the user is not at the computer system, the tape
30 backup program will automatically begin the backup by ensuring that a
suitable
media, such as a tape is in position in the tape drive. If not, it prompts the
user to
insert a tape.
CA 02307212 2000-04-11
WO 99/23562 PCT/US98/23152
3
The tape backup program allows a user to continue working, and backing
up data real time, such as by use of any writable media, such as tape,
diskette or
zip drive until the potentially failing disk drive can be repaired. The backup
program also allows a user to leave a system unattended, with some assurance
S that potential disk drive failures are likely to be detected and data backed
up
without user intervention.
In still further embodiments, other forms of nonvolatile storage devices
are used as a backup device, such as another disk drive, or a writable CD ROM.
In one variation, the disk drive is backed up via a network connection to a
server
10 or other device having suitable storage capabilities.
Figure 1 is a block diagram of a computer system employing the present
invention.
Figure 2 is a block diagram of functional modules used in one
15 embodiment of the present invention.
Figure 3 is a flowchart depicting steps followed by the functional modules
in Figure 2 to detect a potential failure condition and initiate a
backup of the data in the potentially failing device.
Figure 4 is a flowchart depicting steps followed to determine if a backup
20 is required based on prior backup history.
In the following detailed description, reference is made to the
accompanying drawings which form a part hereof, and in which is shown by
way of illustration specific embodiments in which the invention may be
25 practiced. These embodiments are described in sufficient detail to enable
those
skilled in the art to practice the invention, and it is to be understood that
other
embodiments may be utilized and that structural changes may be made without
departing from the scope of the present invention. Therefore, the following
detailed description is not to be taken in a limiting sense, and the scope of
the
30 present invention is defined by the appended claims.
A block diagram of a computer system 100 in Figure 1 will be described
with respect to the present invention. Further details of software modules
CA 02307212 2000-04-11
WO 99/23562 PCT/US98/23152
4
implementing the invention will be described with reference to Figure 2, and
flowcharts depicting details of the process implemented by the modules and
computer system will be described in Figures 3 and 4.
Computer system 100 in one embodiment is a typical personal computer
5 and comprises a processor 110 coupled to a memory 112 and system controller
114. The system controller is also coupled to the processor 110 and both the
processor 110 and system controller 114 can access data in memory 112. The
system controller 114 is also coupled to a host bus 116. Host bus 116 is also
coupled to a plurality of peripheral devices comprising a disk drive 118, a
tape
10 drive 120, PCI device interface 122, a graphics controller 124 which is
further
coupled to a display device 126, and a keyboard/mouse controller i 28 which in
turn is coupled to a keyboard 130. All of these elements operate together in a
well known manner, with software residing in memory 112 such as RAM,
BIOS, DRAM or other memory being executed in processor 110. System
15 controller 114 provides an interface to the peripheral devices, allowing
data
transfers between the peripheral devices and to and from memory 112 without
data having to first be routed through processor 110.
Some of the programs that processor 110 executes include an operating
system, application programs, peripheral device drivers and other modules or
20 programs. In Figure 2, a block diagram wherein the blocks represent program
modules and devices shows blocks involved in detecting potential failures in
disk drive device 118, permitting backup of data on disk drive 118 onto tape
drive 120. Predictive failure analysis functionality is provided on many disk
drives that are available on the market today from disk drive vendors
including
25 IBM Corporation, Western Digital Corporation, Seagate and Quantum to name
a few. One industry standard for predictive failure analysis functionality is
referred to as Self Monitoring, Analysis and Reporting Technology
(S.M.A.R.T.) as indicated in block form at 210.
Information regarding the operational characteristics of the disk drive
30 118 are provided at registers which are then polled by BIOS/Driver 212 and
provided to an application agent 214. Application agent 214 provides messages
to a user regarding the status of the disk drive 118 and initiates a tape
backup of
CA 02307212 2000-04-11
WO 99123562 PCTIUS98l23152
data on the disk drive if it is determined that a failure of the disk drive is
likely
to occur within a set time. Application agent 214 first ensures that proper
media 216 is available for use by the tape drive 120, and if not, will prompt
a
user to insert suitable media such as a tape. Application agent 214 then
invokes
5 operating system services 220 to start a backup program 222 which can be the
same program as normally used to backup the disk drive 118. Backup program
222 initiates the backup, and data from the disk drive is transferred to the
tape
as represented by a bus 218, such as a PCI bus. It should be noted that backup
program 222 can be used to cause backup to any suitable storage device,
10 whether local or remote via network. Application agent 214 serves as a
router
between the bios 212 and the operating system.
Analysis block 210 monitors a range of attributes and sends attribute and
threshold information to application agent 214 via registers. In normal
operation, analysis block 210 then decides if an alert is warranted, and sends
that message to the system, along with the attribute and threshold
information.
The attribute and threshold level implementation varies with each disk drive
vendor, and are based on historical failure analysis of data, collected from
information stored in disk drives that have failed. Attribute individualism is
important because drive architectures vary from model to model. Attributes and
20 thresholds that detect failure for one model may not be functional for
another
model.
Predictable failures are characterized by degradation of an attribute over
time, before the disc drive fails. This creates a situation where attributes
can be
monitored, making it possible for predictive failure analysis. Many mechanical
25 failures are typically considered predictable, such as the degradation of
head
flying height, which would indicate a potential head crash. Certain electronic
failures may show degradation before failing, but more commonly, mechanical
problems are gradual and predictable.
Though attributes are drive-specific, a variety of typical characteristics
30 can be identified: head flying height, data throughput performance, spin-up
time, re-allocated sector count, seek error rate, seek time performance, spin
try
recount, and drive calibration retry count to name a few. Others may be used
in
CA 02307212 2000-04-11
WO 99/23562 PCTIUS98/23152
6
various disk drives dependent upon the design and historical failure
information.
There are currently two S.M.A.R.T. specifications which are being
implemented in disk drives. S.M.A.R.T. emerged for the ATA/IDE
5 environment when SFF-8035 was placed in the public domain. SCSI drives
incorporate a different industry standard specification, as defined in the
ANSI-SCSI Informational Exception Control (IEC) document X3T10/94-190.
The S.M.A.R.T. system technology of attributes and thresholds is
similar in ATA/IDE and SCSI environments, but the reporting of information
differs. In an ATA/IDE environment, software on the host interprets the alarm
signal from the drive generated by the "report status" command of S.M.A.R.T.
Application agent 214 polls the drive on a regular basis to check the status
of
this command, and if it signals imminent failure, sends an alarm to the end
user
or system administrator. Application agent 214 evaluates the attributes and
15 alarms reported, in addition to the "report status" command from the
S.M.A.R.T. analysis block 210.
Generally speaking, SCSI drives with reliability prediction capability
only communicate a reliability condition as either good or failing. In a SCSI
environment, the failure decision occurs at the disc drive as represented at
20 analysis block 210, which notifies the user, and initiates tape backup. The
SCSI
specification provides for a sense bit to be flagged if the disc drive
determines
that a reliability issue exists.
APIs are provided to set ATA registers in ATA/iDE disk drives
supporting S.M.A.R.T. via BIOS/DRIVER 212 which is a BIOS or driver which
25 is capable of sending S.M.A.R.T. commands to and receiving S.M.A.R.T. data
from the ATA interface registers. Application agent 214, such as a backup
program is provided on top of the BIOS or driver to allow a user to control
the
S.M.A.R.T. device and monitor the status of that device. Some subcommands
and their respective codes include ENABLE/DISABLE ATTRIBUTE
30 AUTOSAVE - code D2h, ENABLE S.M.A.R.T. OPERATIONS - code DBh,
ENABLE S.M.A.R.T. OPERATIONS - code D9h, and RETURN S.M.A.R.T.
CA 02307212 2000-04-11
WO 99/23562 PGTNS98/23152
7
STATUS - code DAh. The RETURN S.M.A.R.T. STATUS subcommand is
used to retrieve status information from one or more ATA registers.
In Figure 3, steps taken to monitor the status of the disk drive 118 and
respond are detailed. These steps may be implemented entirely in a device
5 driver, BIOS or an application program, or spread therebetween. Most
implementations will provide for status polling in a driver or BIOS, with
other
steps implemented in an application program written in any number of high
level languages such as C++. At 310, the drive registers or bit is polled. A
polling interval can be user defined or preset. A shorter time will provide a
better chance of recovering if a failure is quick to develop, but it should be
recognized that there are some modes of failure that are currently not
predictable. The interval time should be selected to ensure significant system
resources are not consumed by the polling and further processing activity
associated with each poll.
At 312, the register value or values which comprise information
regarding the status of the disk drive and attributes such as those listed
previously are received and compared with predefined or user defined values.
In one embodiment, only the status of the disk drive, which in the case of
SCSI
devices is a single bit indicating potential failure condition. If a potential
failure
20 condition is either received or deduced from the attributes at 320,
messages
indicating such a failure condition being eminent are provided to the user or
a
system administrator at 322. If no failure condition is detected, control is
returned to polling at 310.
Following detection of a potential failure condition, tape backup is
attempted starting at 324, where the tape drive is checked for suitable media
such as a tape cartridge. If no media is detected, the user is prompted to
insert
such media at 328 and a wait state is entered at 330 until such media is
detected
as present. Following the detection of media at 324, a normal tape backup
operation is begun at 336. Such operations are well known in the art and in
the
30 past have been user initiated or periodically performed during normal
operation.
Status of the backup operation via messaging facilities is provided to the
user as
indicated at 338 either before or during the tape backup operation. When the
CA 02307212 2000-04-11
WO 99123562 PCTIUS98/23152
8
tape backup is completed at 344, an indication of the completion is provided
to
the user prior to end 346.
Users interact with application agent 214 via function provided in the
flowchart of Figure 4. At block 410, the user is provided an interface via
5 command, graphical user interface, menu driven interface, voice or other
constructs to enable or disable the automatic tape backup feature. At 412, the
user is permitted to edit the backup criteria via similar interface. This
allows a
user to attempt to ensure that the data throughput of the disk drive is still
sufficient to provide data fast enough to keep the tape drive operating in a
streaming mode. If the data transfer rate is too slow, the tape device may
only
be able to write one block at a time and then try to resynchronize the tape to
write the next block of data after stopping and rewinding following the first
block if the second block is not immediately available. Buffering techniques
can be useful in ensuring that the tape drive operates in a streaming mode,
but
15 may not suffice if the performance of the disk drive has deteriorated too
far.
The enable/disable and editing criteria interfaces may be combined into
a single screen, which may also be combined with normal control of disk drive
functions, such as via a control panel as is commonly used in personal
computer
operating environments or operating systems.
At 418, previous backup information which has been stored is
interrogated and if the drive has been recently backed up as determined at
422,
the backup feature is disabled for a selected period of time. Following this
time, which is user definable but defaulted to approximately 24 hours, the
backup feature is enabled at 430. The user may also set values at 412 to
25 indicate that the backup feature should not be automatically enabled. If
the disk
drive has been recently backed up at 422, control is returned at 432. The
functions provided by blocks 418, 422 and 430 may also be performed on a
periodic basis, which again can be user definable at 412.
CONCLUSION
30 A system for providing automatic backup of disk drive data upon
detection of potential future failure of the disk drive has been described. It
is to
be understood that the above description is intended to be illustrative, and
not
CA 02307212 2000-04-11
WO 99/23562 PCT/US98I23152
9
restrictive. Many other embodiments will be apparent to those of skill in the
art
upon reviewing the above description. While the invention makes use of the
predictive failure analysis capabilities described in S.M.A.R.T., other
predictive
failure analysis capabilities may also be used to provide an automated backup
function. Such capabilities can also easily be integrated into other types of
devices which store data and whose potential failure can be predicted, such as
CD ROM devices and other devices which may not yet even be invented.
Further, while tape drives have been specified in the embodiments described as
the backup device, other devices may also be used, such as semiconductor
memory devices, or even other disk drives on the same computer system or on a
server or other networked computer or storage facility. Many of the functions
provided by BIOS or the application can be provided by software, hardware or
firmware as is well known to those skilled in the art, and the location of the
provider of the functions is also a matter of well known design choice.
Further,
the present invention could be incorporated with other computer systems, such
as a portable computers, servers, midrange computers or other computers.
