US20170364294A1 - Method and system for automatically preserving persistent storage - Google Patents

Abstract

Computer-based methods, techniques, and systems for automatically protecting a storage device from unwanted alterations are provided. Example embodiments provide a Disk Access Redirection System, which includes a Redirection Driver, an Available Space Table (“AST”), a Protected Space Redirection Table (“PSRT”), and optionally an Unprotected Space Table (“UST”). The Redirection Driver is installed and registered with the computer operating system so that it can intercept storage device access requests (such as a disk read/write). When a storage access request for a read or write is sent, the request is intercepted by the Redirection Driver, transparent to the code that invokes the storage access request. The Redirection Driver uses the AST, PSRT, and optionally the UST, to allocate available storage space for redirected write requests, redirect write requests for protected areas of the storage device, and redirect read requests when the read request specifies a storage location that has been previously redirected.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS
• This application is a continuation of U.S. patent application Ser. No. 13/030,872, filed Feb. 18, 2011, entitled “METHOD AND SYSTEM FOR AUTOMATICALLY PRESERVING PERSISTENT STORAGE” which is a continuation of U.S. patent application Ser. No. 12/427,643, filed Apr. 21, 2009, entitled “METHOD AND SYSTEM FOR AUTOMATICALLY PRESERVING PERSISTENT STORAGE”, which is a continuation of U.S. patent application Ser. No. 09/923,727, filed Aug. 6, 2001, entitled “METHOD AND SYSTEM FOR AUTOMATICALLY PRESERVING PERSISTENT STORAGE”, which claims the benefit of U.S. Provisional Patent Application No. 60/223,829, filed Aug. 8, 2000, entitled “METHOD AND SYSTEM FOR IMPLEMENTING TEMPORARY MODIFICATIONS TO A COMPUTER SYSTEM”, all of which are incorporated herein by reference in their entireties.
TECHNICAL FIELD
• The present disclosure relates to methods, techniques, and systems for protecting a computing system, and, in particular, to methods, techniques, and systems for providing a storage redirection driver that protects the storage devices of a computing system from alteration.
BACKGROUND
• Often times it is desirable to insure that a computer workstation remains unaltered, even though it is being used by a multitude of users in potentially varied and unknown ways. For example, in public environments such as schools, libraries, and other community facilities, access to a group of computer systems is highly desired, for example, to conduct research, use common resources of a company, or to learn and try new programs. Typically, these environments can be characterized as having “uncontrolled access,” because the moment by moment usage of the computer systems is not monitored.
• In such situations, it is highly desirable to insure that users can utilize the functionality of the computer system but not be allowed to “corrupt” the persistent storage of the system, so that follow-on users will find the system in a pristine state. One mechanism for preventing such corruption is to make a copy of the pristine state of the computer system desired and to restore the copy prior to shutdown of the system by each user. One difficulty with such an approach is that it is often difficult in operating system environments to capture the hardware on/off switch, and so, when a user cycles power without a proper operating system shutdown, the system is unable to successfully restore the pristine state.
• Some systems have used special hardware cards to perform the copying. For example, one such system copies the pristine state of the disk storage device into a special disk partition before use access is permitted. Data from the special disk partition is then copied back (restored) upon computer system shut down.
• In database environments, other techniques have been used. Typically, for example, the database changes (in opposite order) are “undone” to the storage device to return it to a pristine state. This technique also suffers from failure in the cycle power situation.
SUMMARY
• Embodiments described herein provide methods, techniques, and systems for automatically preserving an original state of a computing system upon rebooting. Example embodiments provide a Disk Access Redirection System (the “DARS”) to allow all or portions of a storage device to be protected from modification. The DARS can protect such storage devices as disk drives, and other persistent and semi-persistent storage devices. The DARS reads data from and writes data to a redirected data area (a redirected space) when a storage access request is received that would otherwise alter the state of an area of the storage device that has been designated as protected. When the computing system is shut down, the redirected data area is discarded; thus, when the computing system is rebooted, the original state of the protected portions if the storage device are automatically restored, without the need to copy any information from a backup area.
• In one example embodiment, the DARS comprises a Redirection Driver, and several redirection tables, including an Available Space Table (“AST”), a Protected Space Redirection Table (“PSRT”), and, when unprotected areas can be designated, an Unprotected Space Table (“UST”). The AST indicates available space of a storage device and is used to allocate space as redirected space. The PSRT indicates the mappings of protected locations on the storage device to locations in the redirected space. The UST indicates unprotected locations of the storage device. In some embodiments, one or more entire storage devices can be designated as protected. In other embodiments, portions of a storage device can be designated as protected and other portions designated as unprotected. The DARS differentiates between protected and unprotected storage areas, redirecting storage write requests to a redirected data area when a storage area has been designated as protected.
• In one example embodiment, the protection level for the DARS is configurable. For example, the protection level may be configured as all storage is protected, portions of storage (at the device level or within a device) are protected, or no storage is protected. Variations of these combinations are also configurable.
• In some embodiments, when a portion of storage is indicated as unprotected, it is written to directly by the appropriate storage driver. In other embodiments, the Redirection Driver redirects storage access requests to unprotected areas, and integrates them into the unprotected portion of storage upon computer system shut down, or upon reboot.
• In one example embodiment, the Redirection Driver translates an original location in a storage access request to a redirected storage location and forwards a revised storage access request to the storage driver to perform the storage access. In another embodiment, the Redirection Driver performs the actual access itself.
• In yet another example embodiment, the Redirection Driver can redirect data at different driver access levels. For example, the Redirection Drive can direct data at a file level, a cluster level, a logical sector level, or a physical sector level. Combinations of redirecting data at different levels are also provided. In addition, a Redirection Driver that operates using virtual clusters (or any data abstraction that can be implemented as any size object for storing data) can be used with the DARS. In this embodiment, available space objects are provided that implement the mapping between virtual clusters and the actual size of the space in the redirected area. Virtual cluster embodiments provide extensibility across operating systems and different storage devices.
• In one example embodiment that supports a layered driver architecture, the Redirection Driver is inserted into a chain of native operating system drivers and registered with the operating system. When a storage access request is made, the appropriate driver in the chain is invoked by the operating system. In this manner, the Redirection Driver intercepts storage access requests. In some such embodiments, the layer at which the Redirection Driver is inserted into the chain of drivers is based upon the level of storage access being redirected.
• In some embodiments, the DARS optionally performs functions upon computer system shutdown. In one such embodiment, the DARS optionally saves the AST, PSRT, and UST tables so that the redirected data can be restored upon computer system reboot. In another embodiment, redirected data that corresponds to unprotected areas of storage is also saved.
• In one embodiment, the DARS saves the AST and PSRT tables at a desired frequency to the hard disk, so that redirected data can be restored upon a system reboot.
BRIEF DESCRIPTION OF THE DRAWINGS
• FIG. 1 is a graphic illustration of an example use of a Disk Access Redirection System (DARS) in a library.
• FIG. 2 is an example block diagram of disk access redirection performed by the Disk Access Redirection System.
• FIG. 3 is a block diagram of the components of an example embodiment of a Disk Access Redirection System.
• FIG. 4 is an overview flow diagram of the example operation of the Disk Access Redirection System to achieve storage preservation according to the techniques described herein.
• FIG. 5 is an example flow diagram of the steps performed by the Disk Access Redirection System when the Redirection Driver intercepts a disk access request.
• FIG. 6 is an example block diagram of a general purpose computer system for practicing embodiments of a Disk Access Redirection System.
• FIG. 7 is an example block diagram of a typical layout of a physical disk storage device in a computer system.
• FIG. 8 is an example block diagram of a disk drive partitioned into logical drives (partitions).
• FIG. 9 is an example block diagram of a layered I/O driver architecture of a typical operating system with an inserted Redirection Driver.
• FIG. 10 is an example block diagram illustrating how the Disk Access Redirection System tables are used by the Redirection Driver to redirect disk accesses.
• FIG. 11 is an example block diagram of a disk drive with a File Allocation Table arranged according to the FAT 16 architecture.
• FIG. 12 is an example block diagram of a disk drive with a File Allocation Table arranged according to the FAT 32 architecture.
• FIG. 13 is an example block diagram of a disk drive arranged according to the New Technology File System (“NTFS”) architecture.
• FIG. 14 is an example flow diagram of the steps performed by the Disk Access Redirection System when the computer system boots up.
• FIG. 15 is an example flow diagram of the steps performed by the Disk Access Redirection System when a new device is mounted by the computer system.
• FIG. 16 is an example flow diagram of the steps performed by the Redirection Driver when a read request is intercepted.
• FIG. 17 is an example flow diagram of the steps performed by the Redirection Driver when a write request is intercepted.
• FIG. 18 is an example flow diagram of the steps performed by the Redirection Driver when a system shutdown request is intercepted.
DETAILED DESCRIPTION
• Embodiments described herein provide methods, techniques, and systems for preserving an original state of a computing system upon rebooting. Common operating systems and application programs rely upon storing data to and retrieving data from persistent storage devices such as hard disks and other types of memory. Example embodiments provide a Disk Access Redirection System (the “DARS”) to allow all or portions of such storage to be protected from modification. Using the DARS, operating systems, application programs, and other code read and write data on a computer system storage device in a way that automatically allows the original state of protected portions of the storage device to be re-established when the computer system is rebooted. Specifically, the DARS evaluates storage access requests and automatically directs or redirects the flow of information so that, transparent to the user, the original state of the protected storage device is maintained and the changes to protected portions of the storage device are discarded when the computer system is rebooted. In this manner, security of the storage device is insured—even from a user who powers off the machine using a power switch without properly shutting down the system. Although primarily discussed below with reference to disk storage and access, especially hard disks, the techniques described herein are also applicable to other types of persistent and semi-persistent storage devices, including such devices as: CDROMS, Flash Memory, floppy disks, and other types of removable media storage devices.
• The Disk Access Redirection System is useful in a multitude of situations, including those in which it is desirable to allow operators (users) of the computer system to actually use the system to produce useful output, but where the risk of leaving the system in an inoperable state is too high. For example, the DARS may be useful in a teaching situation where it is highly desirable that the computer system isn't truly modified after each teaching session and is returned to its initial state before each session. Or, for example, the DARS could be used in a library where a variety of persons with different levels of experience may need to use computer systems to locate particular pieces of literature, perform database searches, or edit documents (public access word processing, for example). In these situations, it is desired that the machine can be easily (and speedily) returned to some known, operable state.
• FIG. 1 is a graphic illustration of an example use of a Disk Access Redirection System (DARS) in a library. Alibrarian101 operates and maintains acomputer system103 for the benefit ofvarious users102. Thecomputer system103 includes a storage device104 (for example, a hard disk) withavailable space108. Theusers102 may want to use thecomputer system103, for example, to access an online card catalog, access the Internet, or perhaps even to download and execute programs on thecomputer system103. Without the use of the DARS, theusers102 may store files on thehard disk104 that may quickly exhaust theavailable space108 or may execute malicious code, for example, that installs viruses on thehard disk104. The DARS allows thelibrarian101 to easily restore the original state of the disk, which preferably contains sufficientavailable space108 and does not contain malicious code, by simply rebooting the system.
• Specifically, thelibrarian101 first installs and activates the DARS on thecomputer system103. The DARS identifies space on the disk that is to be protected, protectedspace106, and space that is available,available space108, (e.g., space currently unused by any programs). Optionally, if only a portion of the disk is to be protected,unprotected space105 may also be identified. As needed, the DARS allocatesavailable space108 to a “redirected”space107. When theusers102 execute programs that request data to be written to the protectedspace106 on the disk, the DARS automatically intercepts the write request and redirects the write to memory in the redirectedspace107 on the disk. When theusers102 execute programs that request data to be read from the disk, the DARS intercepts the read request and, automatically determines from which location the read should be performed. When thelibrarian101 or one of theusers102 shuts down thecomputer system103, data stored in the redirectedspace107 is discarded, rendering any apparent changes to the protectedspace106 “lost” from the user's perspective and “ignored” from the system's perspective; when thecomputer system103 is rebooted, the protectedspace106 is unchanged from the state it occupied prior to installation of the DARS. In one embodiment, the DARS supports the ability to protect some but not all portions of the disk. In this case, when thecomputer system103 is rebooted, modifications written to theunprotected space105 remain.
• FIG. 2 is an example block diagram of disk access redirection performed by the Disk Access Redirection System. In summary, the DARS operates by intercepting disk access requests before the appropriate device driver handles the disk access requests. Adisk access request201 is formed, for example, by an application program or an operating system function, and includes indicators that identify the type of request and the memory address to which the access is being requested. TheDARS202 intercepts thedisk access request201 before it is sent to theappropriate disk driver203. Thedisk driver203 then reads from or writes to the designated area ondisk drive204.
• In a typical operating system, storage devices are accessed by device drivers that are organized according to a layered architecture. For example, in the Windows 9X or NT systems, developed by Microsoft Corporation, application requests to access a file are passed to one or more high level drivers, which process and eventually forward the request into one or more requests to a low level driver that communicates with a device using physical address indicators. TheDARS202 can be implemented to intercept disk access requests at several of these levels. Which driver(s) is (are) intercepted by theDARS202 depends upon whether redirection is occurring on the file, cluster, or sector level, or according to some other logical drive division scheme implemented by the computer system or by theDARS202.
• Once the request is intercepted, theDARS202 evaluates the type of the request and determines whether the requested original address or a redirected address should be accessed. TheDARS202 then forwards thedisk access request201 with the original address or a redirected address to theappropriate drivers203. Thedisk drivers203 then handle thedisk access request201, accessing as indicated, the protectedspace206, theunprotected space205, or the redirectedspace207. In an alternative embodiment, theDARS202 may handle the disk access request itself rather than forwarding the request to native operating system disk drivers. Other combinations, such as forwarding the request when the original address is used and handling it otherwise, are also contemplated.
• FIG. 3 is a block diagram of the components of an example embodiment of a Disk Access Redirection System. The example embodiment of a DiskAccess Redirection System301 comprises aRedirection Driver302, an Available Space Table (AST)303, a Protected Space Redirection Table (PSRT)304, and, optionally, an Unprotected Space Table (UST)305. TheAST303,PSRT304, andUST305 may be stored in any type of storage, including for example, volatile memory. TheRedirection Driver302 intercepts the disk access request, determines whether or not the requested memory address is to be redirected, and forwards the disk access request to the next appropriate Input/Output (“I/O”) driver. (See, for example,FIG. 2.) TheAST303 is used by theDARS301 to map available memory for use as the redirected space (e.g.,available space208 and redirectedspace207 inFIG. 2). ThePSRT304 is used by theDARS301 to manage the mappings of protected space to redirected space (e.g., protectedspace206 inFIG. 2). TheUST305 is used by theDARS301 to map the unprotected space so that it is either written to directly or redirected to space that is preserved across boot sessions. TheUST305 is optional in that there may be implementations where all storage space is either protected or available, and none of the storage space is allowed to be unprotected (alterable).
• FIG. 4 is an overview flow diagram of the example operation of the Disk Access Redirection System to achieve storage preservation according to the techniques described herein. Instep401, a user turns on the computer system. Instep402, the computer system loads the Redirection Driver. For example, in a computer system running the Microsoft Windows 9X Operating System, the redirection driver is stored in a special folder, for example the “windows\system\iosubsys” folder to be automatically loaded when the operating system boots up. In addition, the driver indicates to the operating system at which location in the layered architecture it desires to be installed. In another example, a computer system running the Microsoft Windows NT or 2000 Operating System, the redirection driver is registered with the operating system during an installation procedure. When the computer is booted, the redirection driver is loaded by the operating system. The specific steps necessary to install, register, initialize and/or load the redirection driver may vary depending on requirements of the operating system or other system components. These system dependent modifications are contemplated and are to be included within the scope of the techniques described.
• Instep403, the computer system determines the desired protection level. For example, the computer system may allow a user, for example, a system administrator or other operator, to access a password protected user interface, through which the user specifies a desired level of system protection. The user interface may be implemented, for example, as a portion of the DARS redirection driver (which grabs control of the system right after it boots up) or as a separate part of the operating system installed as an initial start-up program. Example protection levels may include no protection, partial protection, and full protection. These protection levels correspond to the amount of security desired, where full protection insures no alteration of the persistent storage devices. Instep404, the system examines the determined protection level, and if no protection is desired, the DARS is effectively terminated in that the Redirection Driver is not initialized to intercept requests; otherwise, the system continues instep405.
• Instep405, if the DARS determines that the computer system is to be partially or fully protected, it registers the Redirection Driver to intercept disk access requests at the appropriate level as discussed with reference toFIG. 2. For the purposes herein, “registered” implies any actions (including none) that are required by a particular operating system to make the Redirection Driver known to the operating system so that the Redirection Driver can intercept access requests. In some cases, “registered” means that the Redirection Driver calls a particular operating system function to become known. In other cases, such as within Windows 9X, the driver is automatically registered to intercept requests at a location in the driver chain according to data stored in the Redirection Driver file at bootup time provided the driver is stored in the “windows\system\iosubsys” directory. In systems running the Windows NT operating system, data stored in the operating system registry when the driver is installed is used to register the driver at the correct level each time the system is booted.
• Instep406, as disk access requests are received, the Redirection Driver intercepts the disk access requests. The steps performed by the DARS when the Redirection Driver intercepts disk access requests are discussed below in detail, with respect toFIG. 5. Typically, at some later point in time, instep409, the computer system shuts down. Optionally, steps407 and408 are performed before the computer system shuts down. Instep407 the Redirection Driver intercepts the system shutdown request. Instep408, based on the determined protection level, the DARS may either save modifications that were made or discard them.
• FIG. 5 is an example flow diagram of the steps performed by the Disk Access Redirection System when the Redirection Driver intercepts a disk access request. As discussed above, with respect toFIG. 4, the computer system determines the protection level for the system. If no protection level is desired, the Redirection Driver is not initialized (or registered to intercept access requests), so has no effect on disk accesses. If at least partial protection of the computer system is desired, then the Redirection Driver is registered at the appropriate level in the system, for each storage device to be protected, and the following steps are performed when the Redirection Driver intercepts a disk access request.
• Specifically, instep501, the Redirection Driver determines whether the access requested is for a read or a write. If the request is for a read, the Redirection Driver continues instep502, else, it continues instep504. Instep502, the Redirection Driver determines whether or not the read is being requested against memory in the protected space (e.g., protectedspace206 inFIG. 2). If the read is being requested against memory in the protected space (no alteration is allowed), the Redirection Driver continues instep503, else it continues instep508. Instep503, the Redirection Driver determines whether or not the requested memory address in the protected space already has been redirected to redirected space (e.g., redirected space207), and, if so, continues instep507, else continues instep508. Instep507, the Redirection Driver determines the redirected address. Instep508, the Redirection Driver forwards the read request with the original or redirected address as determined to the next appropriate I/O driver.
• When it is determined instep501 that the requested disk access is for a write, then, instep504, the Redirection Driver determines whether or not the write is being requested to memory in the protected space (the write is “unallowed”), and, if so, continues instep505, else, continues instep508 to forward the write request to the next appropriate I/O driver. Instep505, the Redirection Driver determines whether or not the requested memory address has been previously redirected (for example, due to a prior write request), and, if so, continues insteps507 and508 as described, else continues instep506. Instep506, the Redirection Driver allocates an available memory address to be used for redirected space and maps the requested memory address to an address in the redirected space. The Redirection Driver then continues instep508 as described above. In an alternate embodiment, instep508, the Redirection Driver handles the disk access requests rather than forwarding them to a native operating system I/O driver.
• FIG. 6 is an example block diagram of a general purpose computer system for practicing example embodiments of a Disk Access Redirection System. Thecomputer system601 contains a central processing unit (CPU)602, input/output devices603, includingstorage device613, adisplay device612, and a computer memory (memory)604. The DiskAccess Redirection System611, comprisingRedirection Driver605, Available Space Table608, Protected Space Redirection Table609, and Unprotected Space Table610 preferably resides in memory with theoperating system606 andother programs607. Thestorage device613 contains anunprotected space614, a protectedspace615, a redirectedspace616, and anavailable space617. Thestorage device613 may be implemented in any of various configurations including, for example, one or more physical or virtual disk drives located on one computer or located on multiple computers connected through a network.
• Various arrangements of this computer system and its components are possible and contemplated by the methods, techniques, and systems described. For example, the various tables of the Disk Access Redirection System may reside in separate memories or span across several memories or be non-contiguous. Well-known techniques for handling such data structures and memory management can be used. In addition, the Redirection Driver, before being installed into memory, may be remotely located and accessed for use via a network when desired. Various other modifications to the storage organization and the location of the other parts of the computer system are also contemplated. In addition, in the example flow diagrams described, different orderings of the steps and different divisions of the steps are likewise contemplated.
• As previously mentioned, theRedirection Driver605 may be implemented to intercept I/O (input/output) requests at various levels in the storage driver architecture of theoperating system606 depending upon what level of abstraction of storage is being redirected. For example, theRedirection Driver605 may be implemented to redirect data access at the file level or at the physical sector address level, or at some other level. Additionally, multiple Redirection Drivers may be used, for example, one Redirection Driver for each storage device type, or for different levels of access.
• FIG. 7 is an example block diagram of a typical layout of a physical disk storage device in a computer system. As shown on the left side ofFIG. 7, adisk storage device701 is made up of a series ofsectors730. Each sector may be identified by (1) a head number (e.g.,720), which defines the disk surface where the sector resides; (2) a cylinder number (e.g.,740), which defines the track where the sector resides; and (3) a sector number (e.g.,730) within the associated track. These identifying attributes may then be used to map a unique number to each physical sector. The right side ofFIG. 7 shows the samedisk storage device701, but using acontiguous sector description702. For example, the first sector is assigned thenumber 00, the second is assigned thenumber 01, and so on. This allows each sector to be accessed through a simple indexing scheme, 0, 1, 2 . . . (n−1), where n is the total number of sectors on the physical disk. Eachsector702 comprises a fixed number ofbytes703. For example,Sector 00 contains 8 bytes, labeled Byte 0-Byte 7.
• In addition, a physical disk may be partitioned into one or more logical drives.FIG. 8 is an example block diagram of a disk drive partitioned into logical drives (partitions). The disk drive is shown divided into 3 logical partitions:802,803, and804. Within a partition, the sectors are referred to as logical sectors. For example,partition802 contains physical sectors812-818 which are mapped to logical sectors 0-6, andpartition803 contains physical sectors820-823 mapped to logical sectors 0-3. Logical sectors are accessed by asimple indexing scheme 0, 1, 2, . . . (n−1) relative to the partition, where n is the number of logical sectors in the partition. The size in bytes of a logical sector is not necessarily the same as a physical sector, but may be an integral number of physical sectors. For example,partition804 contains physical sectors840-845 mapped to logical sectors 0-2, where each logical sector is equal in size to two physical sectors. Althoughlogical partitions802,803, and804 are shown mapped to contiguous physical sectors, in some systems, logical partitions may be mapped to non-contiguous physical sectors.
• User or application information and other data are typically stored in files. A file can be stored on a logical drive in one or more logical sectors, which may or may not be contiguous. Different operating systems use different schemes to keep track of where a file is stored on a logical drive and to keep track of which sectors a file occupies. Often times some sort of “table,” “database,” or other appropriate data structure, is used to locate the file on the logical drive and to track the sectors allocated to the file. In some systems, to reduce the size of the database required to describe the files on a disk storage device, the abstraction of a cluster is used. A cluster is an integral number of sequential sectors. For example, in the Windows 9X operating system, a cluster can be up to 64 logical sectors. The way that an operating system organizes files into clusters and the way it keeps track of this organization in a “database” is typically referred to as the file system. The database used by the file system must also be stored on the disk storage device. This database can be stored on the logical drive as a sequential set of logical sectors that is fixed in size and location, or it can be stored as a special file where access to it is achieved through a smaller fixed set of data.
• In any operating system and file system, user and application data starts out in the form of a file. The operating system maintains a database of the clusters that are used to store the file data. Any read or write of a file or portion of a file ends up being a read or write of one or more physical sectors. Modern operating systems are typically coded in such a way that the file system reads and writes to a disk storage device through a layering of drivers as briefly described above. These drivers are defined such that drivers at the top communicate at the file level, while drivers at the bottom communicate at the physical sector, level of the disk storage device. At some point in the layered input/output system, a driver can be inserted to intercept all the reads and writes at the level it is inserted, for example, at either the logical sector level or the physical sector level.
• FIG. 9 is an example block diagram of a layered I/O driver architecture of a typical operating system with an inserted Redirection Driver. The layered I/O driver architecture901 comprisesfile level drivers902, logicalsector level drivers903, and physicalsector level drivers904. InFIG. 9, theRedirection Driver920 implements logical sector redirection, so is shown inserted before the other drivers that handlelogical sectors903. When a request for file access comes in, afile level driver902 translates the request into a read or a write request of the appropriate logical sectors and forwards the translated request down the driver chain. TheRedirection Driver920, which has been inserted into the driver chain beforelogical sector drivers903, receives the translated request and processes the request according to the described techniques. TheRedirection Driver920 can be similarly inserted at other points (or at multiple points) in the driver chain, depending upon what disk access abstraction has been implemented in the driver.
• In example embodiments of the Disk Access Redirection System, the Available Space Table608, Protected Space Redirection Table609, and Unprotected Space Table610 may be organized at the logical sector level, at the cluster level, at some other level of data abstraction, or a combination of data abstraction levels. When organized at the sector level, the Protected Space Redirection Table609 tracks all sectors that can be redirected, and each entry corresponds to one logical sector. For example, if an entry for a particular sector is zero, then no redirection has occurred. In contrast, if an entry is non-zero, then redirection has occurred, and an indicator of the redirected sector is stored in that table entry. (The actual entry may contain a variety of data, such as a logical sector address in Redirected Space, or a pointer to an entry in theAST608, which is allocated to the Redirected Space.) When organized instead at the cluster level, the Protected Space Redirection Table609 tracks all clusters that can be redirected, and each entry in the table corresponds to one cluster. For example, if an entry for a particular cluster is zero, then no redirection has occurred. In contrast, if an entry is non-zero, then redirection has occurred, and an indicator of the number of the redirected cluster (or other reference) is stored in the table entry.
• The basis of redirection may be clusters, sectors, or any other level of data abstraction, and is based on the organization of the DARS. In some embodiments, the use of clusters will reduce the memory size of the DARS. One embodiment organizes the DARS tables as groups of clusters or sectors instead of as individual clusters or sectors. When using groups of clusters, a table entry consists of a cluster number and an extent (i.e., a number of sequential clusters). Because operating systems attempt to keep reads and writes of related data in sequential clusters, the first write to a cluster will usually be a write to a sequence of clusters. For example, if a write operation calls for data to be stored inclusters 5, 6, 7, 8, 9, and 10, this information may be stored in a table where each element in the table consists of 3 pieces of information: (1) the original starting cluster number, (2) the number of sequential clusters, and (3) the number of the first redirected cluster. In this example, storing the redirection information in the Protected Space Redirection Table609 requires only three numbers, rather than the six required when redirection information is referred to as individual clusters. As the write blocks get larger, the savings in memory become significant. No matter how large the write becomes, only three numbers are required to store all the redirection information. The other DARS tables can be similarly implemented in terms of logical sectors, clusters, or a combination of both.
• FIG. 10 is an example block diagram illustrating how the Disk Access Redirection System tables are used by the Redirection Driver to redirect disk accesses. As described earlier, disk storage comprises protectedspace1001, which cannot be altered;unprotected space1004, which can be altered; redirectedspace1002, which stores attempted alterations to the protectedspace1001; andavailable space1003, which is currently unused space. The tables of the DARS (tables1020,1030, and1040) are used to manage the mappings between the protectedspace1001, redirectedspace1002,available space1003, andunprotected space1004. The Available Space Table1030 (“AST”) maps the available space1003 (shown hatched); the Protected Space Redirection Table1020 (“PSRT”) maps the protectedspace1001 to the redirectedspace1002; and the Unprotected Space Table1040 (“UST”) maps/records the unprotected space1004 (shown hatched).
• In one embodiment, theAST1030 stores a series of records, each record indicating a portion of memory that is available for redirection use. The AST may be organized as a simple list of available cluster numbers, or it may be organized as a table of items, each of which contain a starting cluster number and an extent (i.e., number of sequential empty clusters). In the embodiment shown, each AST record comprises anavailable address1031 that indicates the beginning of an available portion of memory and anextent1032 that indicates the length of the available portion of memory (e.g., the first shown record indicates two clusters of storage with addresses A01 through A02). As empty clusters are used for redirection, they are removed from the AST or marked as not available. A write operation of “n” sequential clusters generates a request for “n” sequential empty clusters for redirection. The AST services this request. If it is not possible to find “n” sequential clusters in the AST, the request is preferably broken up into smaller units.
• TheUST1040 is similar to theAST1030, storing a series of records, each comprising abeginning address1041 and an extent1042 (e.g., the first shown record indicates two clusters of storage with addresses U01 through U02). The records in the UST indicate portions ofunprotected space1004, which can be altered. In an alternative embodiment, theunprotected space1004 is mapped via theUST1040 to redirectedspace1002 and saved upon computer shut down. This embodiment has the disadvantage of potentially losing data written to theunprotected space1004 when the user simply powers down the machine using the hardware power switch, without gracefully shutting down the operating system and allowing the redirectedspace1002 to be saved.
• ThePSRT1020 maps portions of protectedspace1001 to redirectedspace1002. Each record in thePSRT1020 comprises a protectedaddress1021 that indicates the beginning of a protected portion of storage, anextent1022 that indicates the length of the protected portion of storage, and a redirectedaddress1023 that indicates the beginning of a portion of memory to which all reads and writes to the indicated protected address of that record will be redirected. For example, in the first record, addresses P01 through P03 of storage in the protected space1001 (as shown by an extent of 3 clusters) is currently redirected to addresses R02 through R04 in the redirectedspace1002. Although described with respect to clusters, a similar organization and description can be used to implement the tables and the space layout using other abstractions, such as at the file level or sector level.
• In an example embodiment shown, the AST, UST, and PSRT (e.g.,AST1030,UST1040, andPSRT1020 inFIG. 10) are stored in a volatile memory of the computer system. In alternate embodiments, these tables may be written to persistent storage and deleted prior to the computer system being shut down. In another alternate embodiment, the determined system protection level may indicate whether the tables are stored in volatile memory or persistent storage. For example, one level of protection may be defined which maintains the AST, PSRT, and UST in persistent storage across several system reboots. This may be beneficial on a system that is being used to test new software. When testing is complete, or at various stages during the testing process, the system protection level could be modified to a level that causes the DARS to store the AST, PSRT, and UST in volatile memory or to delete them from persistent memory on shutdown. Various protection levels may be defined, each requiring possible variations in implementation of the DARS system tables. These variations may include the specific data structures of the tables in addition to the type of memory used to store the tables. Well-known techniques for varying the data structures and for indicating mappings can be used. All of these variations are contemplated and are intended to fall within the scope of the described techniques, methods, and systems.
• In some operating systems, certain portions of protected memory need to be redirected prior to allowing a user access to the system. In particular, in operating systems whose file systems store data on persistent storage, it is necessary to redirect the file system data itself to properly protect these systems. For example, in the Microsoft DOS and Windows™ operating systems, there are a number of different file systems, which store data on a system hard drive. These systems use a File Allocation Table (FAT) to track and allocate clusters to files. In particular, FAT file systems use a table of “next” cluster numbers. To work correctly in these environments, the DARS is implemented to handle both the FAT 16 system and the FAT 32 system. In FAT 16, the FAT is a table of 16 bit cluster numbers. In FAT 32, the table consists of 32 bit cluster numbers.
• FIG. 11 is an example block diagram of a disk drive with a File Allocation Table arranged according to the FAT 16 architecture. InFIG. 11, alogical drive1101 is organized as: (1)Boot record sector1102, (2) FAT sector1103, (3) Root directory sector1104, and (4)Data clusters1105. The first threesections1102,1103, and1104 are fixed in size. The data section starts at a fixed point on the drive and is viewed by the operating system as a sequence of clusters.
• FIG. 12 is an example block diagram of a disk drive with a File Allocation Table arranged according to the FAT 32 architecture. Alogical drive1201 is organized as: (1)Boot record sector1202, (2)FAT sector1203, and (3)Data clusters1204. The first twosections1202 and1203 are fixed in size. The data section starts at a fixed location and is viewed by the operating system as a sequence of clusters.
• FIG. 13 is an example block diagram of a disk drive arranged according to the New Technology File System (“NTFS”) architecture. The entire logical drive is organized using cluster segmentation, and all data is organized as files within the cluster framework. For example,cluster1301 contains the Boot Sector.
• The DARS is implemented to redirect all portions of the disk drive, including those used by the operating system, as shown inFIGS. 11, 12, and 13. This feature allows a random hardware “power off” to leave the system secure. One method of implementing the DARS in FAT 16 and FAT 32 is to define two sets of the AST and PSRT redirection tables, which are organized to different data size abstractions. According to this embodiment, the redirection tables are sector-based for that part of the disk drive that precedes the data section and cluster-based for the data part (the remainder) of the disk drive. Because all available space on a drive will be in the data section of the drive, sectors from the pre-data part of the drive are redirected to the cluster organized part of the drive. Thus, several computations are necessary in this embodiment to translate sectors to clusters and vice versa.
• To determine in which cluster a logical sector lies, the DARS in this example embodiment uses the following formula:
• 
  Cluster=((LSector−D0)/SPC)+2;  (1)
• where,
  • D0=logical sector number where the data section starts
  • SPC=# sectors per cluster.
    The +2 results from a specific nuance in that the FAT 16/32 architecture labels the first cluster in the “data” section asCluster 2.
• To calculate the first sector of a cluster, the DARS uses the formula:
• 
  LSector=((Cluster−2)*SPC)+D0.  (2)
• To calculate the offset within a cluster for a given logical sector, the DARS uses the formula:
• 
  Offset=(LSector−DO)SPC.  (3)
• • Where the “%” is a modulus operator that gives the integer remainder from a division.
• Using the Microsoft New Technology File System “NTFS”, the operating system views the entire disk drive as a sequence of clusters, and all space on the drive is allocated in clusters. (SeeFIG. 13.) Conversion from clusters to sectors or sectors to clusters is required because the file system usually handles cluster numbers, but reads and writes are usually handled as sectors. For example, queries for available space will usually be obtained in terms of clusters. Therefore, the following example formulas can be used to translate between logical sectors and clusters:
• 
  LSector=Cluster*SPC  (4)
• 
  Cluster=LSector/SPC  (5)
• For purposes of managing the redirection information generically, the DARS can be implemented using a concept of virtual clusters. In this implementation, the DARS only deals with redirection at the cluster level and is designed to be independent of the file system and the operating system.
• Real cluster numbers are translated to virtual clusters using the formula:
• 
  VCluster=RCluster+Offset;  (6)
• where,
  • Offset=(C0/SPC)+A;
  • C0=the logical sector number of Real Cluster zero.
  • SPC=the number of sectors per cluster
  • A=(C0% SPC)? 1:0; Where the “?” is an operator, indicating that if the expression (C0% SPC) evaluates to “true” (any non-zero value), then A=1. If the expression (C0 SPC) evaluates to “false” (zero), then A=0.
• The virtual cluster number that contains a logical sector is given by:
• 
  VCluster=(LSector−S0)/SPC  (7)
• where,
  • SO=(C0% SPC) is the offset of sector zero within Virtual cluster zero
• The first sector of a VCluster is:
• 
  LSector=(VCluster*SPC)+S0;  (8)
• To calculate the offset within a virtual cluster for a given logical sector, the DARS uses the formula:
• 
  Offset=(LSector−S0)SPC  (9)
• In an embodiment that uses virtual clusters, the virtual clusters don't have to be redirected to physical clusters on the same drive. Clusters may be redirected to any place where storage is available, for example, on other drives, on network drives etc. The virtual cluster concept only requires an object of “available space” to which clusters may be redirected, and thus has the advantage of flexibility for a variety of operating systems. Virtual clusters may also be implemented to include virtual sectors or any other “virtual” data abstraction level. In systems that support object-oriented programming, the virtual cluster can be implemented as a class, whose implementation is specified by a particular instance of the Redirection Driver.
• In some embodiments, a protection level is supported that allows the specification of some unprotected space. (For example,unprotected space1004 inFIG. 10.) Unprotected space can be implemented by allowing some set of sectors to be read and written into their original locations. Storage that is designated as unprotected is persistent through a reboot. By making the unprotected space appear to the operating system as a drive (which is designated to not be redirected), files can be stored in this space that will be persistent through a reboot. Specifically, because the Redirection Driver can be implemented to intercept reads and writes at the file level, files thus can be redirected into unprotected space—a special drive—instead of redirected space. The information in these files will then be persistent through a reboot. Redirection at the file level may be implemented such that the file appears to the user as if it is still in its original location.
• In an alternative embodiment, an option is presented to a user to save the modifications attempted to protected space. For example, an option can be presented upon computer system shut down to save redirected data in this fashion. The DARS tables contain the information necessary to identify all modifications to the protected space. Thus, by transferring the information to the appropriate file allocation tables, the DARS can make the changes that have been made in the redirected space permanent in the protected space. For example, a user may wish to protect the data on a computer system while running a new software application in order to prevent unexpected data modifications. After running the software application and verifying that no unwanted modifications were made, the changes that were made in the redirected space can be made permanent in the protected space. If unexpected data modifications were made, the user could simply re-boot the computer system, restoring the original data.
• In an example embodiment of the DARS, implemented on a computer system running the Microsoft Windows 9X operating system, the Redirection Driver is implemented with the standard driver entry points that allow the operating system to communicate with any driver. The operating system supports the implementation of a Virtual Device Driver (a V×D), which can be inserted into the driver chain, so that the standard entry points of the driver are invoked by the operating system when the associated event is triggered.
• FIGS. 14-18 are example flow diagrams describing the steps performed by the DARS Redirection Driver at each of five standard driver entry points. A Redirection Driver may be implemented with additional entry points or other entry points, and that the five entry points described with relation toFIGS. 14-18 are intended only to show an example of an example embodiment.
• FIG. 14 is an example flow diagram of the steps performed by the Disk Access Redirection System when the computer system boots up. Instep1401, the computer system or a portion of the DARS determines the protection level for the computer system, as described with reference toFIG. 4. Instep1402, the DARS examines the determined protection level, evaluating whether or not any protection is to be implemented for the computer system. If no protection is to be performed for the computer system, the Redirection Driver is not registered (not inserted into the driver chain) and will not intercept any disk access requests. If full or partial protection is designated, then the DARS continues instep1405, to create an Available Space Table to map all of the available space for the appropriate disk drives. Instep1406, the DARS creates an Unprotected Space Table in embodiments that support partial protection. Instep1407, the DARS creates a Protected Space Redirection Table. Instep1408, the DARS maps the available storage space in the Available Space Table to initialize the AST. Instep1409, the DARS maps unprotected space in the Unprotected Space Table to initialize the UST. For each storage device that is part of the computer system, steps1408 and1409 are performed. In one embodiment, a single set of DARS tables is used. In an alternative embodiment, a separate set of DARS tables is provided for each storage device in the system. Available space may be determined by any of a number of methods, for example, by querying the appropriate level device driver, by reading the FAT tables directly, by consulting a file containing addresses of such areas, accessing an access control list maintained by the operating system, or querying the user. The addresses that correspond to the unprotected space and protected space are determined from the input received after determining the desired protection level (step1401). Instep1410, the system registers the Redirection Driver with the operating system as required by that system. For example, in systems running a Windows 9x operating system, any driver stored in the appropriate directory is loaded and put in the correct location to intercept access requests when the system boots. In systems running other operating systems, a different registration process may be required.
• In an alternate example embodiment, a user may specify a computer system protection level that allows modifications to be maintained across several system reboots. Instep1403, if the determined protection level indicates that the Available Space Table, Protected Space Redirection Table, and, optionally, the Unprotected Space Table are to be retrieved from a previous session, then the DARS proceeds to step1404. Instep1404, the DARS loads the existing tables, preferably from persistent memory into volatile memory, and then proceeds to step1410 where the Redirection Driver is registered to intercept disk access requests, as described above.
• FIG. 15 is an example flow diagram of the steps performed by the Disk Access Redirection System when a new device is mounted by the computer system. Instep1501, the DARS determines whether or not the newly mounted device is a storage device, and, if not, it terminates, else it continues instep1502.Steps1502 and1503 are similar tosteps1408 and1409, respectively, described above. Instep1502, the DARS maps the available space in the Available Space Table. Instep1503, the DARS maps any designated unprotected space in the Unprotected Space Table. In one embodiment, a user interface may need to be presented to query the user as to whether it is desired to protect this new device (for example, when partial protection has been indicated). Instep1504, when the operating system so requires, the Redirection Driver is registered to intercept read and write requests to the new storage device.
• FIG. 16 is an example flow diagram of the steps performed by the Redirection Driver when a read request is intercepted. Instep1601, the driver determines whether the requested address is in the UST, indicating unprotected space that may be freely written to and read from. If so, then the driver proceeds to step1604 to read from the requested address, else, it proceeds to step1602. Instep1602, the driver determines whether or not the requested address is mapped in the PSRT, indicating that the protected space already has been redirected and the redirected address needs to be read from instead of the original address. If the requested address has been redirected, then the driver continues instep1603, else, it continues instep1604. Instep1603, the driver sets the address to the redirected address already mapped in the PSRT. Instep1604, the system forwards the read request, with either the original or redirected address, to the next I/O driver in the driver chain to perform the read.
• FIG. 17 is an example flow diagram of the steps performed by the Redirection Driver when a write request is intercepted. Instep1701, the driver determines whether or not the requested address is in the UST, indicating unprotected space that may be freely written to. If so, then the driver continues instep1707 to write to the requested address, else it continues instep1702. Instep1702, the driver determines whether or not the requested address is mapped in the PSRT, indicating that the protected space already has been redirected. If the requested address has not yet been redirected, then the driver continues instep1703, else it continues instep1706 to retrieve the redirected address. Instep1703, the driver identifies and allocates available space from the AST as required to satisfy the request. The original write request is progressively broken up into smaller requests, if a large enough space is unavailable to satisfy the request. Instep1704, the driver writes a new record to the PSRT, mapping the requested protected address to the available address allocated from the AST instep1703. Instep1705, the driver removes the allocated space from the AST, because, once the available space is used as redirected space, it is no longer available. Instep1706, the driver sets the address requested in the disk access request to the redirected address as mapped in the PSRT. Instep1707, the driver forwards the write request with the original unprotected or redirected address to a lower level disk driver to perform the write.
• As described above, in an alternate embodiment of the DARS, protection levels are implemented that allow modifications to the redirected space to remain persistent across one or more system reboots. In one such embodiment, the standard driver entry point for system shutdown is implemented in the Redirection Driver.FIG. 18 is an example flow diagram of the steps performed by the Redirection Driver when a system shutdown request is intercepted. Instep1801, the driver determines whether or not the protection level in effect for the computer system indicates saving the DARS tables. If so, then the driver continues instep1802, else, the on_shutdown routine terminates. Instep1802, the driver writes the Available Space Table to persistent storage. Instep1803, the driver writes the Unprotected Space Table to persistent storage. Instep1804, the driver writes the Protected Space Redirection Table to persistent storage. Persistent storage for these purposes may also be a logical drive in the file system that has been designated as unprotected storage, as described above. Other steps may be performed on computer system shutdown in this routine, for example, to store unprotected areas if these had been redirected to redirected space instead of written to directly by the appropriate driver.
• All of the above U.S. patents, U.S. patent application publications, U.S. patent applications, foreign patents, foreign patent applications and non-patent publications referred to in this specification and/or listed in the Application Data Sheet, including but not limited to U.S. patent application Ser. No. 13/030,872, entitled “METHOD AND SYSTEM FOR AUTOMATICALLY PRESERVING PERSISTENT STORAGE,” filed Feb. 18, 2011; U.S. patent application Ser. No. 12/427,643, entitled “METHOD AND SYSTEM FOR AUTOMATICALLY PRESERVING PERSISTENT STORAGE,” filed Apr. 21, 2009 and issued as U.S. Pat. No. 7,917,717 on Mar. 29, 2011; U.S. patent application Ser. No. 09/923,727, entitled “METHOD AND SYSTEM FOR AUTOMATICALLY PRESERVING PERSISTENT STORAGE,” filed Aug. 6, 2001 and issued as U.S. Pat. No. 7,539,828 on May 26, 2009; and U.S. Provisional Patent Application No. 60/223,829, entitled “METHOD AND SYSTEM FOR IMPLEMENTING TEMPORARY MODIFICATIONS TO A COMPUTER SYSTEM,” filed Aug. 8, 2000, are incorporated herein by reference, in their entireties.
• Although specific embodiments have been described herein for illustrative purposes, various modifications may be made without deviating from the spirit and scope of the present disclosure. Equivalent methods, structures, processes, steps, and other modifications may be made. For example, the teachings provided herein can be applied to any systems with associated persistent data storage, for example, a personal computer system with a hard disk drive, or a networked server system with remote data repositories. In addition, the teachings may be applied to other types of systems where driver-like code may be implemented to redirect the flow of data through the system. These and other changes may be made to the present disclosure in light of the above-detailed description.

Claims (20)

1.-83. (canceled)

84. A method in a computer system for automatically protecting data, the computer system having a storage area on one or more storage devices, comprising:

determining storage protection level for at least one storage device;

when the protection level indicates that some level of storage protection is desired,

designating a first storage area on the least one storage device as redirected space;

designating a second storage area on the at least one storage device as unprotected space and generating an unprotected space table that stores addresses of the at least one storage device in the unprotected space;

designating a third storage area on the at least one storage device as available space and generate an available space table that stores addresses of the at least one storage device in the available space;

designating a fourth storage area on the at least one storage device as protected space and generating a protected space redirection table that maps addresses of the at least one storage device in the protected space to addresses in the redirected space; and

registering a software redirection driver as an input/output driver and loading the registered redirection driver in a volatile memory of the computer system, the software redirection driver configured to intercept attempted accesses to location in the protected space and redirect the intercepted accesses to the redirected space.

85. The method ofclaim 84 wherein the software redirection driver is further configured to intercept, from requesting code that is external to the software redirection driver, a request to modify a location in the protected space of the at least one storage device and instead redirect the intercepted request to modify a location in the redirected space.

86. The method ofclaim 84 wherein the method is performed when the computer system is performing a bootup operation.

87. The method ofclaim 84 wherein the method is performed where the computer system is performing a mount of a new storage device.

88. The method ofclaim 84, further comprising:

responsive to an indication to shutdown or reboot the computer system, disregarding data in the redirected space.

89. The method ofclaim 84, further comprising:

responsive to an indication to shutdown or reboot the computer system, saving data in the redirected space to the unprotected space.

90. The method ofclaim 84, further comprising:

mapping all the available space into the available space table; and

mapping all the unprotected space in the unprotected space table.

91. The method ofclaim 84 wherein the determining protection level for the at least one storage device is determined in response to user input.

92. The method ofclaim 84 wherein the determining protection level for the at least one storage device determines a level of protection that is one of: no protection, partial protection, of full protection.

93. A computer system for automatically protecting data, comprising:

a first storage area on at least one storage device designating redirected space;

a second storage area on the at least one storage device designating unprotected space;

a third storage area on the at least one storage device designating available space;

a fourth storage area on the at least one storage device designating protected space;

an unprotected space table that stores addresses of the at least one storage device in the unprotected space;

an available space table that stores addresses of the at least one storage device in the available space;

a protected space redirection table that maps addresses of the at least one storage device in the protected space to addresses in the redirected space; and

a software redirection driver, wherein the redirection driver is registered as an input/output driver and loaded into a volatile memory of the computer system, the software redirection driver configured to intercept attempted accesses to location in the protected space and redirect the intercepted accesses to the redirected space.

94. The system ofclaim 93 wherein the software redirection driver is further configured to intercept, from requesting code that is external to the software redirection driver, a request to modify a location in the protected space of the at least one storage device and instead redirect the intercepted request to modify a location in the redirected space.

95. The system ofclaim 93 wherein allocations of the storage device into the first, second, third, and forth storage areas are performed upon boot up.

96. The system ofclaim 93 wherein allocations of the storage device into the first, second, third, and forth storage areas are performed upon mount of a new storage device.

97. The system ofclaim 93 wherein responsive to an indication to shutdown or reboot, the computer system disregards data in the redirected space.

98. The system ofclaim 93 wherein responsive to an indication to shutdown or reboot, the computer system saves data in the redirected space to persistent storage.

99. The system ofclaim 93, further comprising:

an available space table that maps the available space; and

the unprotected space table that maps the unprotected space.

100. The system ofclaim 93, further comprising:

a protection level configured to designate protection of the at least one storage device, wherein the protection level is at least one of unprotected, partially protected, or fully protected.

101. The system ofclaim 93 wherein storage device is arranged according to an architecture that provides files within a cluster framework that provides read and write accesses according to sector number within a cluster of the cluster framework.

102. The system ofclaim 93 wherein the storage device is arranged according to an architecture that provides virtual clusters.

Images