New! View global litigation for patent families

WO2006127402A1 - Version-controlled cached data store - Google Patents

Version-controlled cached data store

Info

Publication number
WO2006127402A1
WO2006127402A1 PCT/US2006/019289 US2006019289W WO2006127402A1 WO 2006127402 A1 WO2006127402 A1 WO 2006127402A1 US 2006019289 W US2006019289 W US 2006019289W WO 2006127402 A1 WO2006127402 A1 WO 2006127402A1
Authority
WO
Grant status
Application
Patent type
Prior art keywords
node
file
version
module
control
Prior art date
Application number
PCT/US2006/019289
Other languages
French (fr)
Inventor
David E. Tewksbary
Original Assignee
Matrixone, Inc.
Priority date (The priority date 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 date listed.)
Filing date
Publication date

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRICAL DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/30Information retrieval; Database structures therefor ; File system structures therefor
    • G06F17/30067File systems; File servers
    • G06F17/30129Details of further file system functionalities
    • G06F17/30132Caching or prefetching or hoarding of files
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRICAL DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/30Information retrieval; Database structures therefor ; File system structures therefor
    • G06F17/30067File systems; File servers
    • G06F17/30129Details of further file system functionalities
    • G06F17/30165Support for shared access to files, file-sharing support
    • G06F17/30168Concurrency control, e.g. optimistic or pessimistic approaches
    • G06F17/30171Locking methods, e.g. locking methods for file systems allowing shared and concurrent access to files

Abstract

Described are methods, systems, and apparatus, including computer program products for a version-controlled cached data store. A first node is associated with a first identifier and a first version. A control file is received, where the control file includes information associating a second node with a second identifier and a second version. The first version is compared with the second version when the first identifier and the second identifier are identical.

Description

VERSION-CONTROLLED CACHED DATA STORE

FIELD OF THE INVENTION

[0001] The present invention relates generally to data storage and, more specifically, to a version-controlled cached data store.

BACKGROUND

[0002] Many search schemes and indexing approaches have been developed to manage file systems and/or databases. One example of an approach is a binary tree approach. In a binary tree approach, a node is established as the root node and the first node that has a greater value than the root node becomes the root node's right child node. Likewise, the first node with a value lower than the root node becomes the root node's left child node. A node that has a value greater than the root node, but less than the root node's right child node becomes the right child node's left child node. The tree structure is built this way as nodes are added to the tree. Periodically, the tree may be rebalanced to even the distribution of nodes from one side of the root node to the other. Ideally, the root node is the middle value of all nodes so that all searches take approximately the same amount of time to complete (i.e., so roughly half the searches traverse the left child node branch and half traverse the right child node branch). In an unbalanced tree, that is, where one node branch of the root node is significantly larger than the other node branch, search performance suffers because one branch of the search tree takes significantly longer to search through. After rebalancing the tree, neither side of the tree has generally more or less nodes than the other side.

[0003] Another approach to searching is the "B-Tree" approach. In a B-Tree, the nodes of the tree may have a larger number of data elements per node compared to the one data element of the binary tree. Correspondingly, a B-Tree may have a larger number of child nodes per parent node since the child nodes contain values between the parent node element values. Unlike the binary tree node which has one data element and two pointers, one for nodes greater than the current node and one for nodes less than the current node, the B-Tree node may have, for example, four data elements per node. The searched-for term is compared against the elements at the current node level, and if the searched-for term falls between nodes, a pointer for that gap is followed.

[0004] Using a B+-tree, or other variants, another approach caches portions of the index which are frequently accessed. For example, see US patent publication number 2003/0033328. The nodes that are cached and read are acquired from the database without a latch. There is a scheme described that uses a latch bit and a version number to ensure that the cached files are accurate and consistent. During the read operation, a version of the node is copied into a register. The contents of the node are read. A latch bit of the read node is examined to ensure that the node has not been latched by another process. The version number of the read node is compared to the version number in the register to ensure that the contents have not been updated during the read operation. If both the latch bit is not set and the version has not changed, then the node is in a consistent state and the node is used.

SUMMARY OF THE INVENTION

[0005] The invention described herein provides systems and methods for achieving a version- controlled cached data store through one or more implementations of various aspects. In one aspect, there is a method for a version-controlled cached data store. In the method, a first node is associated with a first identifier and a first version. The method also involves receiving a control file, which includes information associating a second node with a second identifier and a second version. The method further includes comparing the first version with the second version when the first identifier and the second identifier are identical.

[0006] In another aspect, there is another method for a version-controlled cached data store. The method includes transmitting a first node with a first identifier and a first version and transmitting a control file. The control file includes information associating a second node with a second identifier and a second version and the control file is used by a receiving component for comparing the first version with the second version when the first identifier and the second identifier are identical.

[0007] In some implementations of any of the aspects herein, the method also includes removing the first node if the first version and the second version are different and the first identifier and the second identifier are identical. The method can also include providing a set of cached nodes including the first node. Some of those implementations include receiving the second node if the first node has been removed and in some implementations, receiving involves receiving the second node from a server.

[0008] In another implementation, the method involves preventing the control file from being modified by a first process. The first node may then be generated by a second process, the second process optionally generating a list including new nodes generated by the second process. The second process can also modify the information of the control file that associates the second node with the second version and the second identifier. In some implementations, modifying the information of the control file includes replacing the second version with the first version when the first identifier is identical to the second identifier, with the first version representing a version subsequent to the second version. In those implementations, the first process can be prevented from modifying the control file until the second process has completed modifying the control file.

[0009] In some implementations, the second process generates the first node by copying the first node and modifying the copy. The second process then associates the copy of the first node with the first identifier and a new, third version. Once the new version of the file is associated with the third version, the second process modifies the control file, replacing the second version with the third version where the first identifier is identical to the second identifier. The first node is then removed, leaving the modified copy of the first node (i.e., the copy being associated with the third version).

[0010] In some implementations, the control file is received from a server. In those implementations, the method involves preventing a first process from modifying the control file on the server. A second process modifies the second node associated with the second identifier and a second version. The second process also modifies the received control file such that the first version is set to the second version if the first identifier equals the second identifier. The modified control file is then transmitted to the server. [0011] In another aspect, there is a system for a version-controlled cached data store. The system includes a first node associated with a first identifier and a first version, a control file comprising information associating a second node with a second identifier and a second version, and a comparison module configured to compare the first version with the second version. In a preferred implementation, the system also includes a local store, which in turn includes the first node and the comparison module. The system also includes a remote store which includes the control file and the second node, wherein the comparison module is further configured to retrieve the control file from the remote store. The system also has a locking module for preventing changes to the control file. In some implementations the system also includes a first writing module configured to direct the locking module to prevent modifications to the control file by a second writing module. One example of a locking module uses a lock file to prevent access. [0012] Some implementations include a reading module configured to read the first node if the comparison module determines that the first version and the second version are identical. In some of these implementations, the first node has a maximum number of data elements. The maximum number of data elements may be based on a size of an Input/Output read operation or is equivalent to a maximum number of elements for a single Input/Output read operation. In one implementation, the maximum number of data elements is 128 elements. [0013] In another aspect, there is a computer program product, tangibly embodied in an information carrier for use in a version-controlled cached data store. The computer program product includes instructions that are operable to cause a data processing apparatus to perform a variety of functions. The instructions are operable to associate a first node with a first identifier and a first version. The instructions are also operable to receive a control file, the control file comprising information associating a second node with a second identifier and a second version. The instructions are further operable to compare the first version with the second version when the first identifier and the second identifier are identical.

[0014] In still another aspect, there is a means for a version-controlled cached data store. The means includes means for associating a first node with a first identifier and a first version. The means for the version-controlled cached data store also includes means for receiving a control file, the control file including information that associates a second node with a second identifier and a second version. The means also includes means for comparing the first version with the second version when the first identifier and the second identifier are identical. [0015] Implementations can realize one or more of the following advantages. A latch bit is not needed for each node. Cached nodes are kept in cache and used until the version of that node is changed. The need to read nodes (e.g., to obtain version information and latch information) prior to each read operation is eliminated. A control file is used to transfer version information. Nodes with incorrect version numbers are deleted from the cache and only replaced on an as needed basis. Having a larger number of data elements per node, as well as more child nodes per node, flattens the B-tree, lowering the number of node traversals necessary to find the sought- after value. The larger number of data elements per node also allows a B-Tree to remain unbalanced longer as data elements are inserted and deleted (e.g., compared to a binary tree, which has one element per node). Allowing for an unbalanced tree reduces the resources used to maintain a balanced tree. By setting the number of data elements equal to an Input/Output operation, efficiency is maximized by utilizing the maximum retrieval of data elements during a single request for a node. [0016] Other aspects and advantages of the present invention will become apparent from the following detailed description, taken in conjunction with the accompanying drawings, illustrating the principles of the invention by way of example only. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The foregoing and other objects, features, and advantages of the present invention, as well as the invention itself, will be more fully understood from the following description of various embodiments, when read together with the accompanying drawings, in which: FIG. 1 depicts an architecture for a version-controlled cached data store;

FIG. 2 is a flowchart depicting a process for writing used in a version-controlled cached data store; and

FIG. 3 is a flowchart depicting another process for reading used in a version-controlled cached data store. DETAILED DESCRIPTION

[0018] FIG. 1 depicts an architecture 100 for a version-controlled cached data store. As illustrated in FIG. 1, there is a data store module 105 and one or more processing modules 110a, 110b (generally 110). The data store module 105 includes a control file 115 and one or more nodes (12Ox, 120y, 120z, generally 120) that represent the database. The control file 115 and nodes 120 reside on the file system 125 of the data store module 105. The control file 115 includes information about nodes 120 in the data store module 105. In one embodiment, the nodes 120 are organized as a B-Tree with multiple data elements per node and references to other data nodes. [0019] The information in the control file 115 associates each node 120 with an identifier and a version. The identifier allows the processing modules 110 to compare different nodes, distinguishing one node 12Oi from another node 12On. The version information allows the processing modules 110 to compare versions of nodes stored in their respective cache with the versions in the control file 115 associated with identical identifiers. For example, one version of node 120x (e.g., version 1) may reflect node 12Ox before a write operation and another version may reflect node 12Ox after a write operation (e.g., version 2). The processing module 110 can determine if their cached nodes are the latest versions using a simple comparison of their cached contents to the control file 115. Table 1 illustrates an example of the association of an identifier to a version that can be included in the control file 115. Table 1 uses nodes 120x, 12Oy, and 120z for illustration. Table 1

[0020] The processing modules 110 perform comparisons and read/write operations on the nodes 120 and control file 115 as well as copies of the same stored in the processing module's memory. The comparisons done by the processing module 110 can be performed by a comparison module (not shown) that is included as part of the processing module 110. The processing module 110 includes a copy (e.g., 130a, 130b (generally 130)) of the control file 115 as well as copies (e.g., cache 135a, 135b, generally 135) of some of the nodes 120 of the data store module 105. [0021] In a process more fully described with respect to FIG. 2, the processing module 110a, which performs a write operation, creates a lock file 140 on the file system 125 of the data store module 105. The presence of the lock file 140 prevents other processing modules HOb from changing, modifying, and/or writing to the control file 115 or the nodes 120 that represent the database. During the processing module's HOa write operations, the processing module HOa adds, deletes, and/or modifies nodes located in the processing modules cache 135a as well as modifies the copy 130a of the control file. As these additions, subtractions, and modifications are made to the cache 135a and control file copy 130, a list of changed nodes 145 is created on the processing module 110a. The list of changed nodes 145 represents the subset of the cached nodes 135a that the processing module 110a has generated and/or modified during a write operation. [0022] If a processing module 110b is performing read operations, that processing module 110b has a copy 130b of the control file 115 and a cache 135b of nodes representing a subset of the nodes 120 on the data store module 105. Typically no nodes are changed during a read operation so no list of changed nodes is present or generated on the processing module 110b. [0023] FIG. 2 is a flowchart depicting a process 200 for writing used in a version-controlled cached data store. For illustrative purposes, the system 100 is used as an example to describe the process 200. The processing module 110a attempts to create (205) a lock on the control file 115. If the processing module 110a determines (210) locking the control file 115 is not successful, the processing module 110a determines (215) if a timeout occurred. If a timeout occurred, the processing module 110a enters (220) an error state. If a timeout did not occur, the processing module 11 Oa attempts to create (205) a lock on the control file 115 again. If processing module determines (210) that locking the control file 115 is successful, the processing module 110a retrieves the control file 115 and reads (225) a copy 130a of the control file 115 into memory (e.g., the control file 115 exists on disk 125 in the data store module 105 and the copy 130a exists in the processing module's file system or memory).

[0024] The control file 115 (and copy 130a thereof) includes information associating nodes 120 on the file system 125 (120, 125 from FIG. 1) with node identifiers and versions. For example, node 12Ox is associated on the file system 125 with an identifier, "X" and a version "2", node 12Oy is associated with identifier "Y" and version "4", and node 12Oz is associated with identifier "Z", version "1". When the processing module 110a initially reads (225) the copy 130a of the control file 115, the information associating nodes 120 with identifiers and versions is the same in the copy 130a and the original control file 115. Once the copy 130a is read (225) into memory, the processing module 110a then synchronizes (230) the cache 135a of nodes. [0025] Synchronization (230) is accomplished by examining the control file 130a, which represents nodes 120 on the file system 125, and comparing the information associating nodes with identifiers and versions in the control file 130a with the cache 135a of nodes in the processing module 110a. The processing module 110a determines, for nodes with identical identifiers, which nodes in the cache 135a are associated with the same version as nodes 120 on the file system 125. For example, if a node associated with identifier "X" 12Ox in the control file 130a is associated with version 2 and node X2 (associated with identifier "X") in the cache 135a is version 2, then the processing module 110a deems node X2 in the cache 135a to be synchronized. If a node in the cache 135a is associated with a different version than the node information in the control file 130a (where the nodes are associated with the same identifier), the node in the cache 135a is deemed out-of-synch and is removed. For example, if a node 12Oy (associated with identifier "Y") is associated with version "4" in the control file 130, but the node Y3 in the cache 135a associated with identifier "Y" is associated with version "3", the node Y3 in the cache 135a is out-of-synch. Because that node Y3 in the cache 135a is out of synch, that node Y3 in the cache 135a is removed. While synchronizing (230) the cache 135a, the processing module 110a compares the version of every node in the copy 130 of the control file to the versions of nodes in the cache 135a associated with the same identifier, keeping or removing nodes as described above.

[0026] After the cache in memory is synchronized (230), the processing module 110a traverses the tree to find (235) the node that will be modified and/or inserted. During and/or after traversing the tree, the processing module 110a determines (240) if the node was found. If processing module 110a determines (240) that the node was not found after the tree is traversed, the processing module 110a determines that the value should be inserted into the appropriate node and gets (245) the node from the and reads (245) the node into cache. Inserting a node (and re-balancing of the tree if necessary) can be performed using known techniques.

[0027] Once the processing module 110a determines to insert (245) the value into the tree, the value is set (250) in the modified/new node and the node is associated with a new version (and additionally associated with an identifier in the case of a newly created node). In one implementation, associating an existing node with a new version involves examining the version the node is associated with (e.g., "2" for node 120x) and incrementing the version (e.g., "3" for node 12Ox). Newly created nodes are associated with an initial version (e.g., version "1"). The processing module 110a then updates the information in the copy 130a of the control file associating the node with that identifier with the new/initial version. Thus the node in the cache 135a is associated with an identifier and a new/initial version and the information in the copy 130a of the control file associates the node with the same identifier and the same version. The new and/or modified nodes are also added to the list of changed nodes 145 in the processing module HOa.

[0028] Once an insertion/modification is complete (235 through 250), the processing module 110a determines (255) if any more nodes need to the processed (i.e., the processing module determines whether this set of write operations have been completed). If the processing module 110a determines (255) that additional nodes need to be processed, the processing module finds (235) the next value in the tree, and the processing module 110a repeats 235 through 255 until the processing module 110a determines (255) all write procedures have been completed. [0029] Once the processing module HOa determines (255) all write procedures have been completed, the processing module 110a flushes (260) the list of changed nodes (145 from FIG. 1). Flushing (260) the list of changed nodes 145 involves writing the nodes in memory that have changed versions to the file system with the new version. For example, the list of changed nodes 145 can represent the nodes in memory that are marked as "dirty" (e.g., the node has changed). Additionally, the nodes in memory that have been changed (and the ones that have not been changed) are removed from memory. At this point, nodes associated with the previous versions are on the file system 125, as are nodes associated with new versions. The control file 115 on file system, however, still contains information associating the nodes on the file system with the previous versions. [0030] The processing module 110a then flushes (265) the control file 130. Flushing the control file 130 involves writing the copy 130 of the control file (made in 225) to the file system, replacing control file 115 (i.e., control file 115 becomes control file 130). Once the copy 130 of the control file is written to the file system 125, the only file containing information associating nodes with versions is the control file 130 on the file system. In some implementations, the copy 130 of the control file does not replace the control file 115 on the file system 125, rather the control file 115 is updated to reflect the changes made to the copy 130 of the control file. For example if node 12Ox is associated with the identifier "X" in the control file 115 and the version "2" was updated during the set value (250) operation, node 12Ox in the copy 130 of the control file is associated with identifier "X" and version "3". When the control file is flushed (265), the control file 115 is updated such that the node associated with identifier "X" and version "2" is instead associated with identifier "X" and version "3" (i.e., the same version as the updated version in the cache 135a). It can be seen that prior to the flushing (e.g., 260 and/or 265), any changes to nodes generated on the processing module 110a can be rolled back by not incorporating the changes into the control file 115 of the data store module 105 and simply deleting any new or modified nodes in the processing module 110a without transmitting them to the data store module 105 or deleting nodes on the data store module 105 that have no corresponding entries in the control file 115. [0031] The processing module 110a then deletes (270) the lock on the control file 115. The processing module 110a then deletes (275) the obsolete nodes. An obsolete node is a node on the file system 125 associated with a version that is not included in the information contained in the control file 130 on the file system (i.e., nodes associated with "old" versions). After the processing module 110a finishes deleting (275) the obsolete nodes, the processing module 110a is finished with the write operation and enters an end state (280) for this operation. [0032] FIG. 3 is a flowchart depicting a process 300 for reading used in a version-controlled cached data store. For illustrative purposes, the system 100 is used as an example to describe the process 300. A processing module (e.g., module 110b) reads (305) a copy 130 of the control file 115 into memory (e.g., the control file 115 exists on disk and a copy 130b exists in the processing module's memory). The processing module 110b then synchronizes (310) the in- memory cache of nodes (e.g., 135b) in a manner similar to the manner described herein with respect to synchronizing the cache in FIG. 2 (i.e., 230).

[0033] After the cache in memory is synchronized (310), the processing module 110b traverses the tree to find (315) the node that will be read. During and/or after the tree is traversed, the processing module HOb determines (320) if the node was found. If processing module HOb determines (320) that the node was not found (e.g., because that node was deleted during the synchronization process), the processing module 110b attempts to get and read (325) the node from the data store module 105. This advantageously ensures that the removed node that is received is the latest version included in the data store module 125 at the time of the read operation. If the read operation attempts to retrieve a node that is being modified by a processing module 110a (and is therefore locked), the read operation gets the last committed version of the node from the data store module 105. The last committed version of the node is the version that existed before the processing module 110a locked the node for writing (because the new, modified node is not retrievable until the processing module 110a commits the changed node and the changed copy 130 of the control file to the file system 125). Retrieving the last committed version of the node ensures that the retrieved node is read-consistent with the reading processing module 110b. [0034] To retrieve the node, first, the processing module 110b determines (330) if the path that the processing module 110b expects to find the node in on the data store module 105 is a valid path. Path determination is accomplished by examining the file system 125 location that corresponds to the node identifier and node version. If the node with that identifier and that version does not exist, the path is not valid. Using path validation is advantageous in that no locking or latching is needed to read nodes or determine if the node being retrieved is the latest committed node. If the processing module 110b determines (330) that the path is not a valid path (e.g., that the node does not exist on the file system 125 of the data store module 105 at the expected location, or the folder where the node is expected to be does not exist), the processing module 110b reads (305) into memory a copy 130 of the control file 115 from data store module 105. Re-reading (305) the control file 115 into memory is done to determine if any writes (process 200 of FIG. 2) occurred between when the copy 130 of the control file 115 was read into memory the first time and when the node was determined (330) not to have a valid path. [0035] If the processing module 110b determines (330) that the path to the node on data store module 105 is correct, the processing module 110b reads (335) a copy 135b of the node into memory and the value in the node is retrieved (340). When the nodes 120 are organized as a B- Tree with multiple data elements per node, the nodes 120 can have a maximum number of data elements per node 120 depending on the tree implementation. For example, in some implementations, the maximum number of data elements is based on the size of an Input/Output read operation (e.g., reading 512 kilobytes per read operation). In other implementations, the maximum number of elements is equivalent to a maximum number of elements for a single Input/Output read operation such as reading a maximum of 128 data elements. In these other implementations, reading 128 data elements to retrieve a single node beneficially allows the processing of the maximum number of allowable data elements in a single operation. [0036] Once the value is retrieved (345), the processing module 110b determines (345) if any more nodes need to be read (e.g., is the processing module "done" reading nodes?). If the processing module 110b determines (345) that more nodes need to be read, the processing module 11 Ob returns to 315 and repeats 315 through 345 until the processing module 11 Ob determines (345) that the processing module 110b is done reading nodes. Once the processing module 110b determines (345) that the processing module 110b is done reading nodes, the processing module 110b enters an end state (350) for this read operation. [0037] The above described processes are beneficial for a number of applications. In particular, architectures where the files system 125 is shared, remote, or distributed benefit from caching and version-controlling nodes. In some of these architectures, nodes 120 and the control file 115 are on the file system 125 and the copy 130 of the control file and the cache 135 of nodes are located in memory. Operating on the copy 130 of the control file and the cache 135 in memory increases the performance of the operations because the aspects of read and write operations that are typically latent, e.g., network or disk access, are minimized. Instead, data is transmitted over a network or read from or written to data store module 105 only when updates occur (i.e., a version associated with a node has changed). Also, the techniques advantageously allow the access layer to be located anywhere.

[0038] One example of a benefit is realized when the nodes represent business objects, such as product components or parts, and many business objects are examined and/or updated in one transaction or operation. In this example, business objects stored on the data store module 105 are copied into the memory of the processing module 110. Without the latency imposed by file- opening and reading operations, read and update operations over a large number of business objects in memory are performed more efficiently.

[0039] In some implementations, the processing modules 110 are located on different computer systems than the data store module 105, and the processing modules 110 (reading and/or writing) access the data store module 105 over a network. In these implementations, the processing modules 110 are located on "local stores" and the data store modules 105 is considered a "remote store." In some of these implementations, the data store module 105 may be considered a server and the processing modules 110 considered clients. In those implementations, the processing modules 110 make a copy 130 of the control file 115 by accessing the server and making a file copy on the processing module's file system or memory. Retrieving the nodes 120 on the file system 125 to create the cache 135 is achieved in a similar manner (i.e., accessing the server and making local file/in-memory copies). Updates to the server by writing processing modules 110a are accomplished by transmitting the updated files (the copy 130a of the control file and updated nodes) to the server.

[0040] The nodes mentioned herein may be database rows or records or columns. The nodes may be represented as files located on a file system or as data within a database. The format of the node identifiers and the node versions described herein are for illustration only and any format can be used. For example, node identifiers may be the node's filename or, in the alternative, a property of the file. Regardless of implementation, the nodes are associated with identifiers that distinguish one node from another.

[0041] During cache synchronization (e.g., 310), nodes not found in the copy 130 of the control file may be deleted. The techniques above describe retrieving the latest versions of deleted nodes when they are arrived at during the search process, thus efficiently retrieving nodes (e.g., using network resources and creating latencies) only as needed to complete the search. In some implementations, the processing module 11 Oa can alternatively retrieve the latest version of a node during the synchronization process (e.g., 230). [0042] The above described techniques and systems can be implemented using a variety of data structures such as, but not limited to, databases, file systems, B-trees, pi-trees, hash tables and/or hash maps.

[0043] The above-described lock file 140 can be implemented as an advisory lock, an exclusive lock, or a shared lock. When a processing module 110a has a lock on the control file 115, other processing modules 110 are prevented from changing the control file 115 and the nodes 120. To achieve modification prevention, the other processing modules 110 determine if a lock file 145 exists. If the lock file 145 exists, the other processing modules 110 then do not attempt to change the control file 115. The other processing modules 110 then periodically determine if the control file is unlocked (i.e., the lock file 145 is deleted or removed). If the control file 115 is unlocked, the first processing module 110 able to lock the control file 115 gains write access (i.e., becomes processing module 110a) to the control file 115 and nodes 120.

[0044] The above-described techniques can be implemented in digital electronic circuitry, or in computer hardware, firmware, software, or in combinations of them. The implementation can be as a computer program product, i.e., a computer program tangibly embodied in an information carrier, e.g., in a machine-readable storage device or in a propagated signal, for execution by, or to control the operation of, data processing apparatus, e.g., a programmable processor, a computer, or multiple computers. A computer program can be written in any form of programming language, including compiled or interpreted languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program can be deployed to be executed on one computer or on multiple computers at one site or distributed across multiple sites and interconnected by a communication network. [0045] Method steps can be performed by one or more programmable processors executing a computer program to perform functions of the invention by operating on input data and generating output. Method steps can also be performed by, and apparatus can be implemented as, special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application-specific integrated circuit). Modules can refer to portions of the computer program and/or the processor/special circuitry that implements that functionality. [0046] Processors suitable for the execution of a computer program include, by way of example, both general and special purpose microprocessors, and any one or more processors of any kind of digital computer. Generally, a processor receives instructions and data from a read-only memory or a random access memory or both. The essential elements of a computer are a processor for executing instructions and one or more memory devices for storing instructions and data. Generally, a computer also includes, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto-optical disks, or optical disks. Data transmission and instructions can also occur over a communications network. Information carriers suitable for embodying computer program instructions and data include all forms of non- volatile memory, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks. The processor and the memory can be supplemented by, or incorporated in special purpose logic circuitry. [0047] To provide for interaction with a user, the above described techniques can be implemented on a computer having a display device, e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor, for displaying information to the user and a keyboard and a pointing device, e.g., a mouse or a trackball, by which the user can provide input to the computer (e.g., interact with a user interface element). Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback, e.g., visual feedback, auditory feedback, or tactile feedback; and input from the user can be received in any form, including acoustic, speech, or tactile input. [0048] The above described techniques can be implemented in a distributed computing system that includes a back-end component, e.g., as a data server, and/or a middleware component, e.g., an application server, and/or a front-end component, e.g., a client computer having a graphical user interface and/or a Web browser through which a user can interact with an example implementation, or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication, e.g., a communication network. Examples of communication networks include a local area network ("LAN") and a wide area network ("WAN"), e.g., the Internet, and include both wired and wireless networks.

[0049] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

[0050] The invention has been described in terms of particular embodiments. The alternatives described herein are examples for illustration only and not to limit the alternatives in any way. The steps of the invention can be performed in a different order and still achieve desirable results. Other embodiments are within the scope of the following claims. [0051] What is claimed is:

Claims

1. A method for a version-controlled cached data store comprising: associating a first node with a first identifier and a first version; receiving a control file, the control file comprising information associating a second node with a second identifier and a second version; and comparing the first version with the second version when the first identifier and the second identifier are identical.
2. The method of claim 1 further comprising removing the first node if the first version and the second version are different and the first identifier and the second identifier are identical.
3. The method of claim 2 further comprising providing a set of cached nodes including the first node.
4. The method of claim 2 further comprising receiving the second node if the first node has been removed.
5. The method of claim 4 wherein receiving the second node comprises receiving the second node from a server.
6. The method of claim 1 further comprising preventing a first process from modifying the control file.
7. The method of claim 6 further comprising generating the first node by a second process.
8. The method of claim 7 further comprising generating a list including new nodes generated by the second process.
9. The method of claim 7 further comprising modifying, by the second process, the information of the control file associating the second node with the second version and the second identifier.
10. The method of claim 9 wherein modifying the information of the control file comprises replacing the second version with the first version when the first identifier is identical to the second identifier, the first version representing a version subsequent to the second version.
11. The method of claim 7 wherein preventing comprises preventing the first process from modifying the control file until after the second process has completed modifying the control file.
12. The method of claim 7 wherein generating, by the second process, the first node comprises copying the first node and modifying the copy.
13. The method of claim 12 further comprising the second process associating the copy of the first node with the first identifier and a third version.
14. The method of claim 13 further comprising modifying, by the second process, the control file, replacing the second version with the third version where the first identifier is identical to the second identifier.
15. The method of claim 14 further comprising removing the first node and leaving the modified copy of the first node.
16. The method of claim 1 wherein the control file is received from a server.
17. The method of claim 16 further comprising preventing a first process from modifying the control file on the server; modifying, by a second process, the second node associated with the second identifier and a second version; and modifying, by the second process, the received control file such that the first version is set to the second version if the first identifier equals the second identifier.
18. The method of claim 17 further comprising transmitting the modified control file to the server.
19. A method for a version-controlled cached data store comprising: transmitting a first node with a first identifier and a first version; transmitting a control file, the control file comprising information associating a second node with a second identifier and a second version and being used by a receiving component for comparing the first version with the second version when the first identifier and the second identifier are identical.
20. A system for a version-controlled cached data store comprising: a first node associated with a first identifier and a first version; a control file comprising information associating a second node with a second identifier and a second version; and a comparison module configured to compare the first version with the second version.
21. The system of claim 20 further comprising a local store comprising the first node and the comparison module; and a remote store comprising the control file and the second node, wherein the comparison module is further configured to retrieve the control file from the remote store.
22. The system of claim 21, further comprising a locking module for preventing changes to the control file.
23. The system of claim 22, further comprising a first writing module configured to direct the locking module to prevent modifications to the control file by a second writing module.
24. The system of claim 22 wherein the locking module comprises a lock file.
25. The system of claim 20 further comprising a reading module configured to read the first node if the comparison module determines that the first version and the second version are identical.
26. The system of claim 20 wherein the first node further comprises a maximum number of data elements.
27. The system of claim 26 wherein the maximum number of data elements is based on a size of an Input/Output read operation.
28. The system of claim 26 wherein the maximum number of data elements is equivalent to a maximum number of elements for a single Input/Output read operation.
The system of claim 26 wherein the maximum number of data elements is 128 elements.
29. A computer program product, tangibly embodied in an information carrier for use in a version- controlled cached data store, the computer program product including instructions being operable to cause a data processing apparatus to: associate a first node with a first identifier and a first version; receive a control file, the control file comprising information associating a second node with a second identifier and a second version; and compare the first version with the second version when the first identifier and the second identifier are identical.
30. A means for a version-controlled cached data store comprising: means for associating a first node with a first identifier and a first version; means for receiving a control file, the control file comprising information associating a second node with a second identifier and a second version; and means for comparing the first version with the second version when the first identifier and the second identifier are identical.
PCT/US2006/019289 2005-05-25 2006-05-18 Version-controlled cached data store WO2006127402A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US11/137,460 2005-05-25
US11137460 US7716182B2 (en) 2005-05-25 2005-05-25 Version-controlled cached data store

Publications (1)

Publication Number Publication Date
WO2006127402A1 true true WO2006127402A1 (en) 2006-11-30

Family

ID=36947599

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2006/019289 WO2006127402A1 (en) 2005-05-25 2006-05-18 Version-controlled cached data store

Country Status (2)

Country Link
US (1) US7716182B2 (en)
WO (1) WO2006127402A1 (en)

Families Citing this family (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7996442B2 (en) * 2005-10-17 2011-08-09 Oracle International Corporation Method and system for comparing and re-comparing data item definitions
US20070088766A1 (en) * 2005-10-17 2007-04-19 Oracle International Corporation Method and system for capturing and storing multiple versions of data item definitions
US8015165B2 (en) * 2005-12-14 2011-09-06 Oracle International Corporation Efficient path-based operations while searching across versions in a repository
US7472140B2 (en) * 2005-12-20 2008-12-30 Oracle International Corporation Label-aware index for efficient queries in a versioning system
US7543004B2 (en) * 2005-12-22 2009-06-02 Oracle International Corporation Efficient support for workspace-local queries in a repository that supports file versioning
US7533136B2 (en) * 2005-12-22 2009-05-12 Oracle International Corporation Efficient implementation of multiple work areas in a file system like repository that supports file versioning
US7730032B2 (en) * 2006-01-12 2010-06-01 Oracle International Corporation Efficient queriability of version histories in a repository
JP4693669B2 (en) * 2006-03-23 2011-06-01 キヤノン株式会社 Image forming apparatus, an image processing apparatus, image forming method, an image processing method, program, storage medium
US8433693B2 (en) * 2007-04-02 2013-04-30 Microsoft Corporation Locking semantics for a storage system based on file types
US20080243847A1 (en) * 2007-04-02 2008-10-02 Microsoft Corporation Separating central locking services from distributed data fulfillment services in a storage system
US9449047B2 (en) 2007-06-19 2016-09-20 Sybase, Inc. Dynamic modification of schemas in streaming databases
US8745012B2 (en) * 2007-08-10 2014-06-03 Sybase, Inc. Log-structured store for streaming data
US8856206B2 (en) * 2007-08-28 2014-10-07 International Business Machines Corporation Maintaining message versions at nodes in a network
WO2009027138A1 (en) * 2007-08-30 2009-03-05 International Business Machines Corporation Accessing data entities
US8489654B2 (en) * 2009-08-28 2013-07-16 Beijing Innovation Works Technology Company Limited Method and system for forming a virtual file system at a computing device
US8630980B2 (en) 2010-04-06 2014-01-14 Microsoft Corporation Synchronization framework that restores a node from backup
CN101872163B (en) 2010-05-21 2012-02-01 杭州电子科技大学 Process monitoring method based on recursive non-linear partial least squares
US20130232495A1 (en) * 2012-03-01 2013-09-05 Microsoft Corporation Scheduling accelerator tasks on accelerators using graphs
CN102819585B (en) * 2012-07-31 2015-04-22 北大方正集团有限公司 Method for controlling document of extensive makeup language (XML) database
US8868531B2 (en) * 2012-09-10 2014-10-21 Apple Inc. Concurrent access methods for tree data structures
US9424079B2 (en) 2013-06-27 2016-08-23 Microsoft Technology Licensing, Llc Iteration support in a heterogeneous dataflow engine
US9542335B1 (en) 2014-07-25 2017-01-10 Google Inc. Methods and systems for rule-based flexible cache invalidation
US9792309B2 (en) 2014-08-04 2017-10-17 Cohesity, Inc. Write operations in a tree-based distributed file system

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0689140A2 (en) * 1994-06-24 1995-12-27 International Business Machines Corporation Client-server computer system and method for updating the client, server, and objects
EP0864129B1 (en) * 1995-12-01 2000-08-16 BRITISH TELECOMMUNICATIONS public limited company Database access

Family Cites Families (43)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4714992A (en) * 1985-11-26 1987-12-22 International Business Machines Corporation Communication for version management in a distributed information service
US5047918A (en) * 1985-12-31 1991-09-10 Tektronix, Inc. File management system
US5010478A (en) * 1986-04-11 1991-04-23 Deran Roger L Entity-attribute value database system with inverse attribute for selectively relating two different entities
US4875159A (en) * 1987-12-22 1989-10-17 Amdahl Corporation Version management system using plural control fields for synchronizing two versions of files in a multiprocessor system
US4912637A (en) * 1988-04-26 1990-03-27 Tandem Computers Incorporated Version management tool
US5278979A (en) * 1990-12-20 1994-01-11 International Business Machines Corp. Version management system using pointers shared by a plurality of versions for indicating active lines of a version
US5430869A (en) * 1991-05-29 1995-07-04 Hewlett-Packard Company System and method for restructuring a B-Tree
US5386559A (en) * 1992-07-16 1995-01-31 International Business Machines Corporation Variant domains and variant maps in a versioned database management system
US5649200A (en) * 1993-01-08 1997-07-15 Atria Software, Inc. Dynamic rule-based version control system
JP2863805B2 (en) * 1993-11-26 1999-03-03 富士通株式会社 Version management system
US5806078A (en) * 1994-06-09 1998-09-08 Softool Corporation Version management system
CA2137492C (en) * 1994-12-07 1998-07-28 Lenny Kwok-Ming Hon System for and method of providing delta-versioning of the contents of pcte file objects
US5745753A (en) * 1995-01-24 1998-04-28 Tandem Computers, Inc. Remote duplicate database facility with database replication support for online DDL operations
US5805889A (en) * 1995-10-20 1998-09-08 Sun Microsystems, Inc. System and method for integrating editing and versioning in data repositories
US6366933B1 (en) * 1995-10-27 2002-04-02 At&T Corp. Method and apparatus for tracking and viewing changes on the web
US5862325A (en) * 1996-02-29 1999-01-19 Intermind Corporation Computer-based communication system and method using metadata defining a control structure
US5806065A (en) * 1996-05-06 1998-09-08 Microsoft Corporation Data system with distributed tree indexes and method for maintaining the indexes
US6026413A (en) * 1997-08-01 2000-02-15 International Business Machines Corporation Determining how changes to underlying data affect cached objects
US6122645A (en) * 1997-08-25 2000-09-19 Lucent Technologies, Inc. System and method for physically versioning data in a main memory database
US6018747A (en) * 1997-11-26 2000-01-25 International Business Machines Corporation Method for generating and reconstructing in-place delta files
JP3912895B2 (en) * 1998-04-15 2007-05-09 富士通株式会社 Structured data management system, a computer-readable recording medium storing structured data management program, and structured data management method
US6868425B1 (en) * 1999-03-05 2005-03-15 Microsoft Corporation Versions and workspaces in an object repository
US6327584B1 (en) * 1999-07-30 2001-12-04 Hewlett-Packard Company Apparatus and method for using version control to dynamically update files while the files are available for access
US6449624B1 (en) * 1999-10-18 2002-09-10 Fisher-Rosemount Systems, Inc. Version control and audit trail in a process control system
US6757893B1 (en) * 1999-12-17 2004-06-29 Canon Kabushiki Kaisha Version control system for software code
US7243136B2 (en) * 2000-01-18 2007-07-10 Inktomi Corporation Approach for managing and providing content to users
US6681389B1 (en) * 2000-02-28 2004-01-20 Lucent Technologies Inc. Method for providing scaleable restart and backout of software upgrades for clustered computing
US6598059B1 (en) * 2000-04-22 2003-07-22 Oracle Corp. System and method of identifying and resolving conflicts among versions of a database table
US6631386B1 (en) * 2000-04-22 2003-10-07 Oracle Corp. Database version control subsystem and method for use with database management system
US7421541B2 (en) * 2000-05-12 2008-09-02 Oracle International Corporation Version management of cached permissions metadata
GB2369208B (en) * 2000-11-21 2004-10-20 Gawne Cain Res Ltd Database management systems
US20020073110A1 (en) * 2000-12-12 2002-06-13 Edouard Duvillier Version collection technique implemented on an intrinsic versioning information storage and retrieval system
US20020073082A1 (en) * 2000-12-12 2002-06-13 Edouard Duvillier System modification processing technique implemented on an information storage and retrieval system
US6868414B2 (en) * 2001-01-03 2005-03-15 International Business Machines Corporation Technique for serializing data structure updates and retrievals without requiring searchers to use locks
US7293028B2 (en) * 2001-06-08 2007-11-06 Sap Ag Cache-conscious concurrency control scheme for database systems
DE10152765B4 (en) * 2001-07-13 2015-11-12 Siemens Aktiengesellschaft Method for the electronic provision of services to equipment via a data communication link
CA2481687A1 (en) * 2002-04-11 2003-10-23 Linuxcare, Inc. Managing multiple virtual machines
US7117052B2 (en) * 2003-02-18 2006-10-03 Fisher-Rosemount Systems, Inc. Version control for objects in a process plant configuration system
US20040168115A1 (en) * 2003-02-21 2004-08-26 Bauernschmidt Bill G. Method and system for visualizing data from multiple, cached data sources with user defined treemap reports
WO2004090672A3 (en) * 2003-04-02 2004-11-18 Shaw Chuang Methods and systems for controlling network infrastructure devices
JP2005234860A (en) * 2004-02-19 2005-09-02 Hitachi Ltd Version management system, version management server device and storage device controller
JP2006134214A (en) * 2004-11-09 2006-05-25 Hitachi Ltd File version management method and computing system
US7921076B2 (en) * 2004-12-15 2011-04-05 Oracle International Corporation Performing an action in response to a file system event

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0689140A2 (en) * 1994-06-24 1995-12-27 International Business Machines Corporation Client-server computer system and method for updating the client, server, and objects
EP0864129B1 (en) * 1995-12-01 2000-08-16 BRITISH TELECOMMUNICATIONS public limited company Database access

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
"EFFICIENT, APPROXIMATE CACHE INVALIDATION FAR AN OBJECT SERVER" IBM TECHNICAL DISCLOSURE BULLETIN, IBM CORP. NEW YORK, US, vol. 37, no. 1, January 1994 (1994-01), pages 325-326, XP000428794 ISSN: 0018-8689 *
SIMONSON J ET AL: "Version augmented URIs for reference permanencevia an Apache module design" COMPUTER NETWORKS AND ISDN SYSTEMS, NORTH HOLLAND PUBLISHING. AMSTERDAM, NL, vol. 30, no. 1-7, April 1998 (1998-04), pages 337-345, XP004121413 ISSN: 0169-7552 *

Also Published As

Publication number Publication date Type
US20060271606A1 (en) 2006-11-30 application
US7716182B2 (en) 2010-05-11 grant

Similar Documents

Publication Publication Date Title
Kornacker et al. Concurrency and recovery in generalized search trees
US7305386B2 (en) Controlling visibility in multi-version database systems
US7257689B1 (en) System and method for loosely coupled temporal storage management
US6243718B1 (en) Building indexes on columns containing large objects
US5430869A (en) System and method for restructuring a B-Tree
US6728719B1 (en) Method and mechanism for dependency tracking for unique constraints
US5842196A (en) Database system with improved methods for updating records
US6772155B1 (en) Looking data in a database system
US7720892B1 (en) Bulk updates and tape synchronization
US20030196052A1 (en) Method, system, and program for grouping objects
US6567928B1 (en) Method and apparatus for efficiently recovering from a failure in a database that includes unlogged objects
US20100094806A1 (en) File storage system, cache appliance, and method
US5832516A (en) Caching data in recoverable objects
US5864849A (en) System and method for restoring a multiple checkpointed database in view of loss of volatile memory
US7016914B2 (en) Performant and scalable merge strategy for text indexing
US5247647A (en) Detection of deletion of stored data by concurrently executing processes in a multiprocessing data processing system
US20060059204A1 (en) System and method for selectively indexing file system content
US6631366B1 (en) Database system providing methodology for optimizing latching/copying costs in index scans on data-only locked tables
US20030204513A1 (en) System and methodology for providing compact B-Tree
US6456995B1 (en) System, method and computer program products for ordering objects corresponding to database operations that are performed on a relational database upon completion of a transaction by an object-oriented transaction system
US7293028B2 (en) Cache-conscious concurrency control scheme for database systems
US20060167960A1 (en) Lazy timestamping in transaction time database
US8868506B1 (en) Method and apparatus for digital asset management
US5134696A (en) Virtual lookaside facility
US6665675B1 (en) Shared file system having a token-ring style protocol for managing meta-data

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application
NENP Non-entry into the national phase in:

Ref country code: DE

NENP Non-entry into the national phase in:

Ref country code: RU

122 Ep: pct app. not ent. europ. phase

Ref document number: 06770591

Country of ref document: EP

Kind code of ref document: A1