CA2007691C - Method and apparatus for pre-process mounting of file systems in a hierarchical file system environment - Google Patents
Method and apparatus for pre-process mounting of file systems in a hierarchical file system environment Download PDFInfo
- Publication number
- CA2007691C CA2007691C CA002007691A CA2007691A CA2007691C CA 2007691 C CA2007691 C CA 2007691C CA 002007691 A CA002007691 A CA 002007691A CA 2007691 A CA2007691 A CA 2007691A CA 2007691 C CA2007691 C CA 2007691C
- Authority
- CA
- Canada
- Prior art keywords
- directory
- file system
- temporary
- file
- temporary directory
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/10—File systems; File servers
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
In a computer system having a hierarchical file structure a file system is provided wherein per-process mounting of file systems can be performed. To provide per-process mounting, a temporary directory is created off the current root directory. The file system, starting at the root directory, is then mounted on the temporary directory using an innovative loop-back file system function which results in the file system being duplicated but having the temporary directory as its root directory. A mount can then be performed on the temporary directory without affecting the original file system. A change root directory command is then executed to change the root directory to the temporary directory whereby execution continues on the file system having the temporary directory as its coot.
Description
1 . jEUE ".D OF Tli~ iNV~NTION:
The field of the invention is computer systems employing hierarchical file systems.
The field of the invention is computer systems employing hierarchical file systems.
2. ART ~ACIfGROU~ID:
Computer systems available today range in cost and sophistication from low cost micro-processor personal computers to sophisticated, high cast mainframe computers. However, the computer systems can be said to include the same basic components: at least one central processing unit which controls the performance of computer operations, input/output units which control the input and output of information within and out of the computer system, and memory or storage space which stores information used by the CPU to perform computer operations.
The CPU is controlled by the operating system which dictates how 2o certain computer operations are to be performed. One portion of the operating system is the file system which determines and controls the file structure in the computer system.
One type of file structure in use is referred to as a hierarchical file structure. In a hierarchical file structure files are organized info a tree structure having a single root node referred to as the "root". One or more nodes may then be connected to the root and one or more nodes may be connected to each node. Every non leaf node (that is, a node which has a Bode connected to it ) of the tree structure is considered to be a directory of files; and the files at the leaf nodes of the tree are either directories, regular files, or special device 82225.P052 1 ~~1~%~~~~
files. The name of any file is identified by a path name which describes how to locate the file in the file system hierarchy. Tha path name is a sequence of component or node names separated by "!" characters. A component name is a sequence of characters that designates a file name uniquely contained in the previous component. A full path name starts with the "/" character which specifies a file that can ba found by starting at the file system root and traversing the file tree following the branches that lead to successive component names of the path names. For example, the path name /usr/srclcmd/one.c represents the portion of a tree structure represented in Fig.
l0 9 and identified as the path through nodes 10, 15, 20, 25 and 30.
There ace numerous commands which manipulate, refer to, or access the file structures and files within the file structure. Such commands include the "mount" and "unmount" commands. The mount command connects a first file system (the term "file system" will hereinafter be used, as it is used in the art, to identify a hierarchical file structure existing in a computer system) to a second file system while the unmount system command disconnects the first file system from the second file system. Once the file system is mounted, the information in that file system is accessed through the directory of the root file 2o system it is mounted upon. This is illustrated in Fig. 2. Root file system 40, has directories or nodes bin 42, src 45 and usr 46. The file system 70 is mounted on directory usr and the node 72 may be accessed using the path name /usr/bin/date.
Further information may ba found on the mount command in most operating systems manuals. More specific information may be found in the context of the UNlX operating system environment in: Mach, J. Tha Des's the Unix Operating Svstem, (Prentice-Hall 1966). As can ba seen from the above-description, file systems era mounted on a directory-by-directory basis;
that is, at least one directory is mounted on a file system comprising at least one s2~s.~o52 2 . -. 2~iC~'~f~~~.
directory. In a convention UNIX operating system, once a file system has been mounted on a specific directory, the inode of that directory may no longer be accessed by processes.
However, as the sophistication of the users and the applications executed on the computer system increase, it has become desirable to have the capability to mount file systems on a per-process basis even though the processes are located within the same directory. Thus, it is advantageous to have several mounts performed within one directory such that different mounts are utilized according to the current process.
82225.~Q52 ~~l t: '~~~~.
--.
n is therefore an object of the present invention to provide a computer system for per-process mounting of fife systems in a hierarchical fife system structure.
To provide per-process mounting, a temporary directory is created off the current root directory. The file system, starting at the root directory, is then mounted on the temporary directory using an innovative loop-lo back file system function which results in the file system being virtually duplicated having the temporary directory as the root directory. A mount of the file system containing process specific files can than be performed on the temporary directory without affecting the original file system. A change root directory command is then executed to change the root directory to the temporary directory and execution then continued with respect to the mounted file system having the temporary directory as its root.
2~~( f~'7~91 The objects, features and advantages of the system of the present invention will be apparent from the following description in which:
FtG. 1 illustrates a hierarchical file structure employed in computer systems.
FIG. 2 illustrates the system command mount which ~s employed to in prior art operating systems.
FIG. 3 is a block diagram of the kernel in the UNIX operating system employed in the preferred embodiment of the system of the present invention.
FIG. 4 is a flow chart describing the preferred embodiment of the process of the present invention.
FIG. 5 is a block diagram illustrative of the preferred embodiment of the system of the present invention.
82225.P052 Through the system of the present invention per-process mounting may be achieved on a computer system employing a hierarchical file structure thereby adding increased flexibility and capabilities to the computer syste m.
~Dne operating system which uses a hierarchical file structure is the UNIX~ (UNIX is a registered trademark of AT&T) operating system.
io Although the preferred embodiment is described in a UNIX operating system environment, it is evident to one skilled in the art that the system of the present invention may be applied to other computer systems employing an operating system which utilizes a hierarchical file structure.
i5 In a UNIX-based computer system, the operating system interacts directly with the hardware providing common services to programs and isolating the software and users from the hardware idiosyncrasies of a particular computer system. Viewing the system as a set of layers, the operating system is commonly called the "system kernel" or just "kernel", 2o emphasizing its isolation from the user programs. because the programs are independent of the underlying hardware, it is easy to move the programs between UNIX systems running on different hardware. The UNIX file system, located within the kernel, organizes and controls access to the file structures.
25 The kernel is depicted generally in dig. 3. dig. 3 shows ~ levels:
user/programs 74, kernel 75 and hardware 76. The system call module represents the border or interface between the user or user programs and the kernel. The system calls are divided into those system calls that interface with the file system 77 and those that interact with the process control system 76.
3o The process control system is responsible for process synchronization, 82225.P052 6 ~(jCt'7fi~~.
interprocess communication, memory management and process scheduling.
The file system communicates with the hardware control module &0 which interfaces to the computer system hardware and manages files, allocates fifes space, administers free space, controls access to files and retrieves data for users and programs.
The system will herein be described in reference to Figs. 4 and 5.
At block 100, Fig. 4, a temporary directory is created. This directory is preferably created off the root directory so as to eliminate any subsequent 2o confusion with the following process steps and to clearly distinguish the temporary sub-directory from other sub-directories of the same root. The directory identification as temporary is used because it is preferred that after the process, for which the file system was created, has completed execution, the temporary directory and nodes connected to it are removed such that the 25 file system is restored to its previous structure. However, the temporary directory does not have to be removed and can remain indefinitely in the file system structure. The first step described in Fig. 4 block 100 is illustrated by Figs. 5a and 5b. Fig. 5a shows the root directory 200, sub-directory usr 205 and sub-directory of usr identified as src 210. Fig. 5b illustrates the file system 2o after the step 100 of Fig. 4. The temporary directory tmp 215, has been created having a sub-directory off of tmp identified as too 220.
At block 105, Fig. 4, the file system comprising directories 200, 205, 210, 215 and 220 are mounted on the directory too 220, whereby the 25 directory too 220 reflects the contents of the original root directory "/"
200. This is illustrated in Fig. 5c, wherein the file system shown in Fig. 5b is duplicated by mounting the system an directory too 220, resulting in the structure, illustrated by the rectangular shaped nodes, foo 220 having sub-directories tmp 225, foo 230, usr 235 and src 240. This process is achieved using an 3o innovative technique referred to as loop-back mounting in which the file 82225.P052 ~~0.6'~~:91.
structure or file system is duplicated off a different directory or different file system by adding a second pointer which dictates the location of the directory in the file system. This process, results in the "virtual" placement of the directory in two locations. ft should be noted that the apparent duplication of the directories does not result in the actual duplication of the directaries;
only one set of directories exist but the directories appear to be concurrently connected to two different nodes.
At step 910 Fig. 4 the file system, for example the file system that io is desired to be process-specific, is mounted on the temporary file system created. Referring to Figs. 5d and 5e the fife system shown in Fig. 5d comprising directories bin and etc are mounted on the file system, specifically the src sub-directory 240. The resulting structure is represented by the trapezoidal shaped nodes shown in Fig. 5e. At step 115, Fig. 4, the root Zs directory is changed to be ltmp/foo node 220 of Fig. 5e. (ln the tJNlX
operating system the change root directory command is CHROOT). As a result, the file system visible to the user or process is as shown in Fig. 5f; since it is visible only to the process, it is termed "per-process. The stnrcture comprises the root node, which in Fig. 5e was directory foo 220 which now appears as the root 20 260, having sub-directories tmp 265, and fao 270 as well as sub directories usr 2$0, 75 src 280, bin 290 and etc 300.
Thus, the process may be executed within the file structure of Fig.
5f without affecting the ability of other processes executed in the original file 25 system (depicted in Fig. 5a).
it can be seen that the steps set forth in the flowchart of Fig. 4 may be repeated, for example by creating a temporary directory tmpl having a sub-directory foo, and mounting the directories containing the files to be executed 3o with respect to another process. Furthermore the process described may again 82225.PO52 $
be repeated for a second process, third process or more processes such that each process is mounted on a separate file system thereby emulating or providing per-process mounting.
The loopback mounting technique may be easily accomplished in one implementation of the UNIX operating sytem. This implementation includes a virtual file system in the kernel by which it is able to support multiple file . system types such as multiple versions of UNIX and non-UNIX systems such as ~SDOS. Such a virtual file system is described in y ode ~ ~, Architecture for to ~ Iti i~i~ SX~Qm Tyroes in Sun U~ix, S. R. Kieiman, published by Sun ~llicrosystems, inc., fufountain Viaw, California.
As described in that paper, the conventional file system independent inode of the UNIX is renamed vnode (virtuaP node). All file m manipulation is done with a vnode object. File systems are manipulated through the virtual file system. Each virtual file system~includes an analog to the mount table entry of the conventional UNiX system; this specifies the file system type, the directory which is the mount point, generic flags (e.g., read only), and a pointer to file system type specific data. This is linked to a list of 2o mounted file systems so that file system dependent data may be provided.
This virtual file system allows loopback mounts to be made.
To make a loopback mount, the user executes the IUfOUNT () system call. fn the command, the user specifies (1 ) that this call is to be a 2s ioopback mount, (2) the directory which is to be mounted, and (3) the directory it is to be mounted on. As pointed out, the vnode structure in the kernel contains a pointer to the virtual file system specific data. This data inciud~s for a loopback file, among other things, a pointer to the "real vnode" for the.file.
Thus, if directory A is loopback mounted onto directory ~, then in the kernel, 3o the vnode for directory ~ is a loopback vnode whose real vnode is the vnode s222s.POS2 2~1t:' o ~~''~, for directory A. For example, in Fig. 5(e)-(f), where the root directory "/"
is mounted ~o the directory foo 220, the vnode for foo 220 has a pointer to the root "/" directory as its real directory.
When the kernel receives a filesystem request for directory foo220, then the kernel executes the file system independent lookup code before invoking the file system dependent lookup code for the virtual file system that the vnode is in. Since directory foo's vnode is in the loopback file system, the loopback lookup operation is performed. This operation calls the lookup to operation for the real vnode, the vnode of root directory "/". in this way, the contents of directory foo are the same as the contents of the root directory "/"
since the same files are found in either directory. Thus, the additional pointer which causes directory foo to point to the root directory "/" is the real vnode pointer in directory foo's loopback vnode.
While the invention has been described in conjunction with a preferred embodiment, it is evident that numerous alternatives modifications, variations and uses will be apparent to those skilled in the art in light of the foregoing description.
82225.P052 ~ d
Computer systems available today range in cost and sophistication from low cost micro-processor personal computers to sophisticated, high cast mainframe computers. However, the computer systems can be said to include the same basic components: at least one central processing unit which controls the performance of computer operations, input/output units which control the input and output of information within and out of the computer system, and memory or storage space which stores information used by the CPU to perform computer operations.
The CPU is controlled by the operating system which dictates how 2o certain computer operations are to be performed. One portion of the operating system is the file system which determines and controls the file structure in the computer system.
One type of file structure in use is referred to as a hierarchical file structure. In a hierarchical file structure files are organized info a tree structure having a single root node referred to as the "root". One or more nodes may then be connected to the root and one or more nodes may be connected to each node. Every non leaf node (that is, a node which has a Bode connected to it ) of the tree structure is considered to be a directory of files; and the files at the leaf nodes of the tree are either directories, regular files, or special device 82225.P052 1 ~~1~%~~~~
files. The name of any file is identified by a path name which describes how to locate the file in the file system hierarchy. Tha path name is a sequence of component or node names separated by "!" characters. A component name is a sequence of characters that designates a file name uniquely contained in the previous component. A full path name starts with the "/" character which specifies a file that can ba found by starting at the file system root and traversing the file tree following the branches that lead to successive component names of the path names. For example, the path name /usr/srclcmd/one.c represents the portion of a tree structure represented in Fig.
l0 9 and identified as the path through nodes 10, 15, 20, 25 and 30.
There ace numerous commands which manipulate, refer to, or access the file structures and files within the file structure. Such commands include the "mount" and "unmount" commands. The mount command connects a first file system (the term "file system" will hereinafter be used, as it is used in the art, to identify a hierarchical file structure existing in a computer system) to a second file system while the unmount system command disconnects the first file system from the second file system. Once the file system is mounted, the information in that file system is accessed through the directory of the root file 2o system it is mounted upon. This is illustrated in Fig. 2. Root file system 40, has directories or nodes bin 42, src 45 and usr 46. The file system 70 is mounted on directory usr and the node 72 may be accessed using the path name /usr/bin/date.
Further information may ba found on the mount command in most operating systems manuals. More specific information may be found in the context of the UNlX operating system environment in: Mach, J. Tha Des's the Unix Operating Svstem, (Prentice-Hall 1966). As can ba seen from the above-description, file systems era mounted on a directory-by-directory basis;
that is, at least one directory is mounted on a file system comprising at least one s2~s.~o52 2 . -. 2~iC~'~f~~~.
directory. In a convention UNIX operating system, once a file system has been mounted on a specific directory, the inode of that directory may no longer be accessed by processes.
However, as the sophistication of the users and the applications executed on the computer system increase, it has become desirable to have the capability to mount file systems on a per-process basis even though the processes are located within the same directory. Thus, it is advantageous to have several mounts performed within one directory such that different mounts are utilized according to the current process.
82225.~Q52 ~~l t: '~~~~.
--.
n is therefore an object of the present invention to provide a computer system for per-process mounting of fife systems in a hierarchical fife system structure.
To provide per-process mounting, a temporary directory is created off the current root directory. The file system, starting at the root directory, is then mounted on the temporary directory using an innovative loop-lo back file system function which results in the file system being virtually duplicated having the temporary directory as the root directory. A mount of the file system containing process specific files can than be performed on the temporary directory without affecting the original file system. A change root directory command is then executed to change the root directory to the temporary directory and execution then continued with respect to the mounted file system having the temporary directory as its root.
2~~( f~'7~91 The objects, features and advantages of the system of the present invention will be apparent from the following description in which:
FtG. 1 illustrates a hierarchical file structure employed in computer systems.
FIG. 2 illustrates the system command mount which ~s employed to in prior art operating systems.
FIG. 3 is a block diagram of the kernel in the UNIX operating system employed in the preferred embodiment of the system of the present invention.
FIG. 4 is a flow chart describing the preferred embodiment of the process of the present invention.
FIG. 5 is a block diagram illustrative of the preferred embodiment of the system of the present invention.
82225.P052 Through the system of the present invention per-process mounting may be achieved on a computer system employing a hierarchical file structure thereby adding increased flexibility and capabilities to the computer syste m.
~Dne operating system which uses a hierarchical file structure is the UNIX~ (UNIX is a registered trademark of AT&T) operating system.
io Although the preferred embodiment is described in a UNIX operating system environment, it is evident to one skilled in the art that the system of the present invention may be applied to other computer systems employing an operating system which utilizes a hierarchical file structure.
i5 In a UNIX-based computer system, the operating system interacts directly with the hardware providing common services to programs and isolating the software and users from the hardware idiosyncrasies of a particular computer system. Viewing the system as a set of layers, the operating system is commonly called the "system kernel" or just "kernel", 2o emphasizing its isolation from the user programs. because the programs are independent of the underlying hardware, it is easy to move the programs between UNIX systems running on different hardware. The UNIX file system, located within the kernel, organizes and controls access to the file structures.
25 The kernel is depicted generally in dig. 3. dig. 3 shows ~ levels:
user/programs 74, kernel 75 and hardware 76. The system call module represents the border or interface between the user or user programs and the kernel. The system calls are divided into those system calls that interface with the file system 77 and those that interact with the process control system 76.
3o The process control system is responsible for process synchronization, 82225.P052 6 ~(jCt'7fi~~.
interprocess communication, memory management and process scheduling.
The file system communicates with the hardware control module &0 which interfaces to the computer system hardware and manages files, allocates fifes space, administers free space, controls access to files and retrieves data for users and programs.
The system will herein be described in reference to Figs. 4 and 5.
At block 100, Fig. 4, a temporary directory is created. This directory is preferably created off the root directory so as to eliminate any subsequent 2o confusion with the following process steps and to clearly distinguish the temporary sub-directory from other sub-directories of the same root. The directory identification as temporary is used because it is preferred that after the process, for which the file system was created, has completed execution, the temporary directory and nodes connected to it are removed such that the 25 file system is restored to its previous structure. However, the temporary directory does not have to be removed and can remain indefinitely in the file system structure. The first step described in Fig. 4 block 100 is illustrated by Figs. 5a and 5b. Fig. 5a shows the root directory 200, sub-directory usr 205 and sub-directory of usr identified as src 210. Fig. 5b illustrates the file system 2o after the step 100 of Fig. 4. The temporary directory tmp 215, has been created having a sub-directory off of tmp identified as too 220.
At block 105, Fig. 4, the file system comprising directories 200, 205, 210, 215 and 220 are mounted on the directory too 220, whereby the 25 directory too 220 reflects the contents of the original root directory "/"
200. This is illustrated in Fig. 5c, wherein the file system shown in Fig. 5b is duplicated by mounting the system an directory too 220, resulting in the structure, illustrated by the rectangular shaped nodes, foo 220 having sub-directories tmp 225, foo 230, usr 235 and src 240. This process is achieved using an 3o innovative technique referred to as loop-back mounting in which the file 82225.P052 ~~0.6'~~:91.
structure or file system is duplicated off a different directory or different file system by adding a second pointer which dictates the location of the directory in the file system. This process, results in the "virtual" placement of the directory in two locations. ft should be noted that the apparent duplication of the directories does not result in the actual duplication of the directaries;
only one set of directories exist but the directories appear to be concurrently connected to two different nodes.
At step 910 Fig. 4 the file system, for example the file system that io is desired to be process-specific, is mounted on the temporary file system created. Referring to Figs. 5d and 5e the fife system shown in Fig. 5d comprising directories bin and etc are mounted on the file system, specifically the src sub-directory 240. The resulting structure is represented by the trapezoidal shaped nodes shown in Fig. 5e. At step 115, Fig. 4, the root Zs directory is changed to be ltmp/foo node 220 of Fig. 5e. (ln the tJNlX
operating system the change root directory command is CHROOT). As a result, the file system visible to the user or process is as shown in Fig. 5f; since it is visible only to the process, it is termed "per-process. The stnrcture comprises the root node, which in Fig. 5e was directory foo 220 which now appears as the root 20 260, having sub-directories tmp 265, and fao 270 as well as sub directories usr 2$0, 75 src 280, bin 290 and etc 300.
Thus, the process may be executed within the file structure of Fig.
5f without affecting the ability of other processes executed in the original file 25 system (depicted in Fig. 5a).
it can be seen that the steps set forth in the flowchart of Fig. 4 may be repeated, for example by creating a temporary directory tmpl having a sub-directory foo, and mounting the directories containing the files to be executed 3o with respect to another process. Furthermore the process described may again 82225.PO52 $
be repeated for a second process, third process or more processes such that each process is mounted on a separate file system thereby emulating or providing per-process mounting.
The loopback mounting technique may be easily accomplished in one implementation of the UNIX operating sytem. This implementation includes a virtual file system in the kernel by which it is able to support multiple file . system types such as multiple versions of UNIX and non-UNIX systems such as ~SDOS. Such a virtual file system is described in y ode ~ ~, Architecture for to ~ Iti i~i~ SX~Qm Tyroes in Sun U~ix, S. R. Kieiman, published by Sun ~llicrosystems, inc., fufountain Viaw, California.
As described in that paper, the conventional file system independent inode of the UNIX is renamed vnode (virtuaP node). All file m manipulation is done with a vnode object. File systems are manipulated through the virtual file system. Each virtual file system~includes an analog to the mount table entry of the conventional UNiX system; this specifies the file system type, the directory which is the mount point, generic flags (e.g., read only), and a pointer to file system type specific data. This is linked to a list of 2o mounted file systems so that file system dependent data may be provided.
This virtual file system allows loopback mounts to be made.
To make a loopback mount, the user executes the IUfOUNT () system call. fn the command, the user specifies (1 ) that this call is to be a 2s ioopback mount, (2) the directory which is to be mounted, and (3) the directory it is to be mounted on. As pointed out, the vnode structure in the kernel contains a pointer to the virtual file system specific data. This data inciud~s for a loopback file, among other things, a pointer to the "real vnode" for the.file.
Thus, if directory A is loopback mounted onto directory ~, then in the kernel, 3o the vnode for directory ~ is a loopback vnode whose real vnode is the vnode s222s.POS2 2~1t:' o ~~''~, for directory A. For example, in Fig. 5(e)-(f), where the root directory "/"
is mounted ~o the directory foo 220, the vnode for foo 220 has a pointer to the root "/" directory as its real directory.
When the kernel receives a filesystem request for directory foo220, then the kernel executes the file system independent lookup code before invoking the file system dependent lookup code for the virtual file system that the vnode is in. Since directory foo's vnode is in the loopback file system, the loopback lookup operation is performed. This operation calls the lookup to operation for the real vnode, the vnode of root directory "/". in this way, the contents of directory foo are the same as the contents of the root directory "/"
since the same files are found in either directory. Thus, the additional pointer which causes directory foo to point to the root directory "/" is the real vnode pointer in directory foo's loopback vnode.
While the invention has been described in conjunction with a preferred embodiment, it is evident that numerous alternatives modifications, variations and uses will be apparent to those skilled in the art in light of the foregoing description.
82225.P052 ~ d
Claims (14)
1. In a computer system comprising a central processing unit (CPU), an input/output unit and memory, said CPU being controlled by an operating system, said operating system comprising a file system module, wherein the file system module organizes and controls access to files in the computer system, said file system module organizing files into a file system having a hierarchical file structure comprising a root directory and one or more sub-directories, said file system module further comprising:
means for creating a temporary directory connected to the root directory;
means for performing a loopback mount of the file system on the temporary directory wherein said file system is virtually duplicated having the temporary directory as its root directory;
means for mounting files on the temporary directory;
means for changing the root directory to the temporary directory;
whereby the file system which appears to the process is a virtual duplication of the original file system with a special mount visible only to that process.
means for creating a temporary directory connected to the root directory;
means for performing a loopback mount of the file system on the temporary directory wherein said file system is virtually duplicated having the temporary directory as its root directory;
means for mounting files on the temporary directory;
means for changing the root directory to the temporary directory;
whereby the file system which appears to the process is a virtual duplication of the original file system with a special mount visible only to that process.
2. The system of claim 1 wherein the means for performing a loopback mount of the file system on the temporary directory comprises means for creating a virtual file system; and means for creating a pointer from the temporary directory to the root directory whereby the two directories contain the same information to the process.
3. The system of claim 1 wherein the operating system is the UNIX operating system.
4. The system of claim 3 wherein the means for creating a temporary directory connected to the root directory comprises the execution of the command "make directory" by the file system module.
5. The system of claim 3 wherein the means for performing a loopback mount of the file system on the temporary directory comprises means for creating a virtual file system; and means for creating a pointer from the temporary directory to the root directory whereby the two directories contain the same information to the process.
6. The system of claim 3 wherein the means for mounting files containing process data on the file system mounted on the temporary directory comprises the execution of the "mount" command by the file system module.
7. The system of claim 3 wherein the means for changing the root directory to the temporary directory comprises the execution of the "change root directory" ("CHROOT") command by the file system module.
8. In a computer system comprising a central processing unit (CPU), an input/output unit and memory, said CPU being controlled by an operating system, said operating system comprising a file system module, wherein the file system module organizes and controls access to files in the computer system, said file system module organizing files into a file system having a hierarchical file structure comprising a root directory and one or more sub-directories, a method for virtually duplicating file systems to provide per-process mounting comprising the steps of:
creating a temporary directory connected to the root directory;
performing a loopback mount of the file system on the temporary directory wherein said file system is virtually duplicated having the temporary directory as its root directory;
mounting files on the temporary directory;
changing the root directory to the temporary directory;
whereby the file system which appears to the process is a virtual duplication of the original file system and mounts made by this process in the file system are only visible to this process or any process rooted to the same temporary directory thereby providing per-process mounting.
creating a temporary directory connected to the root directory;
performing a loopback mount of the file system on the temporary directory wherein said file system is virtually duplicated having the temporary directory as its root directory;
mounting files on the temporary directory;
changing the root directory to the temporary directory;
whereby the file system which appears to the process is a virtual duplication of the original file system and mounts made by this process in the file system are only visible to this process or any process rooted to the same temporary directory thereby providing per-process mounting.
9. The method of claim 8 wherein the step of performing a loopback mount of the file system on the temporary directory comprises the steps of creating a virtual file system; and creating a pointer from the temporary directory to the root directory whereby the two directories contain the same information to the process.
10. The method of claim 9 wherein the operating system is the UNIX operating system.
11. The method of claim 10 wherein the step of creating a temporary directory connected to the root directory comprises executing the command "make directory"
12. The method of claim 10 wherein the step for performing a loopback mount of the file system on the temporary directory comprises the steps of creating a virtual file system; and creating a pointer from the temporary directory to the root directory whereby the two directories contain the same information to the process.
13. The method of claim 10 wherein the step of mounting files containing process data on the file system mounted on the temporary directory comprises executing the "mount" command.
14. The system of claim 10 wherein the step of changing the root directory to the temporary directory comprises executing the "change root directory" ("CHROOT") command.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US31572489A | 1989-02-24 | 1989-02-24 | |
| US315,724 | 1989-02-24 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CA2007691A1 CA2007691A1 (en) | 1990-08-24 |
| CA2007691C true CA2007691C (en) | 2000-03-14 |
Family
ID=23225770
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA002007691A Expired - Fee Related CA2007691C (en) | 1989-02-24 | 1990-01-12 | Method and apparatus for pre-process mounting of file systems in a hierarchical file system environment |
Country Status (5)
| Country | Link |
|---|---|
| CA (1) | CA2007691C (en) |
| FR (1) | FR2643734B1 (en) |
| GB (1) | GB2228599B (en) |
| HK (1) | HK101193A (en) |
| SG (1) | SG68493G (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5813028A (en) * | 1993-10-12 | 1998-09-22 | Texas Instruments Incorporated | Cache read miss request invalidation prevention method |
| US5537570A (en) * | 1993-10-12 | 1996-07-16 | Texas Instruments Incorporated | Cache with a tag duplicate fault avoidance system and method |
| US5835934A (en) * | 1993-10-12 | 1998-11-10 | Texas Instruments Incorporated | Method and apparatus of low power cache operation with a tag hit enablement |
| US7062660B2 (en) | 2001-08-06 | 2006-06-13 | International Business Machines Corporation | Method and apparatus for controlling the performance of a file system mount operation by a user lacking superuser authority |
| US7763741B2 (en) | 2003-11-12 | 2010-07-27 | The Trustees Of Columbia University In The City Of New York | Separation of ginkgolides and bilobalide from G. biloba |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0664536B2 (en) * | 1986-01-17 | 1994-08-22 | インタ−ナショナル ビジネス マシ−ンズ コ−ポレ−ション | Control method of virtual terminal subsystem |
| US4916608A (en) * | 1986-05-30 | 1990-04-10 | International Business Machines Corporation | Provision of virtual storage resources to an operating system control program |
| EP0278314B1 (en) * | 1987-02-13 | 1994-08-10 | International Business Machines Corporation | Single system image |
-
1989
- 1989-11-01 GB GB8924640A patent/GB2228599B/en not_active Expired - Fee Related
- 1989-12-14 FR FR898916547A patent/FR2643734B1/en not_active Expired - Fee Related
-
1990
- 1990-01-12 CA CA002007691A patent/CA2007691C/en not_active Expired - Fee Related
-
1993
- 1993-05-26 SG SG684/93A patent/SG68493G/en unknown
- 1993-09-30 HK HK1011/93A patent/HK101193A/en not_active IP Right Cessation
Also Published As
| Publication number | Publication date |
|---|---|
| GB2228599A (en) | 1990-08-29 |
| GB8924640D0 (en) | 1989-12-20 |
| GB2228599B (en) | 1993-03-17 |
| FR2643734B1 (en) | 1994-03-04 |
| SG68493G (en) | 1993-08-06 |
| CA2007691A1 (en) | 1990-08-24 |
| HK101193A (en) | 1993-10-08 |
| FR2643734A1 (en) | 1990-08-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5905990A (en) | File system viewpath mechanism | |
| US6055527A (en) | System, method and computer program product for superimposing attributes on hierarchically organized file systems | |
| US5313646A (en) | Method and apparatus for translucent file system | |
| US7234077B2 (en) | Rapid restoration of file system usage in very large file systems | |
| US7099900B1 (en) | Mapping driver for use in data backup systems | |
| US5781908A (en) | File data synchronizer in a distributed data computer network | |
| EP0629950B1 (en) | File backup file system | |
| US6374268B1 (en) | Methods and systems for an incremental file system | |
| US5463774A (en) | Object oriented computer architecture using directory objects | |
| US6944790B2 (en) | System and method for collecting and restoring user environment data using removable storage | |
| US8719767B2 (en) | Utilizing snapshots to provide builds to developer computing devices | |
| EP0526035B1 (en) | Method and apparatus for operating a computer-based file system | |
| Golub et al. | UNIX as an Application Program. | |
| US20020091710A1 (en) | Method for backup and restore of a multi-lingual network file server | |
| WO1996041283A9 (en) | System and method for superimposing attributes on hierarchically organized file systems | |
| JPS62165251A (en) | Apparatus for controlling input/output operation in virtual memory/virtual computer type date processing system | |
| EP0871134A2 (en) | Accessing database information | |
| CA2210147A1 (en) | A method for accessing a file in a multi-processor computer system using pipes and fifos | |
| CA2007691C (en) | Method and apparatus for pre-process mounting of file systems in a hierarchical file system environment | |
| US5655154A (en) | Method and system for sharing utilities between operating systems | |
| US6256751B1 (en) | Restoring checkpointed processes without restoring attributes of external data referenced by the processes | |
| US20060253858A1 (en) | Software service application and method of servicing a software application | |
| EP0579368A1 (en) | File system for a computer | |
| EP0992898A1 (en) | Using a namespace extension to selectively display files read from a computer readable drive | |
| US7107289B2 (en) | Process file systems having multiple personalities and methods therefor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| EEER | Examination request | ||
| MKLA | Lapsed |