CN111338853B - Linux-based data real-time storage system and method - Google Patents
Linux-based data real-time storage system and method Download PDFInfo
- Publication number
- CN111338853B CN111338853B CN202010182063.9A CN202010182063A CN111338853B CN 111338853 B CN111338853 B CN 111338853B CN 202010182063 A CN202010182063 A CN 202010182063A CN 111338853 B CN111338853 B CN 111338853B
- Authority
- CN
- China
- Prior art keywords
- data
- index table
- memory
- written
- processing module
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/14—Error detection or correction of the data by redundancy in operation
- G06F11/1402—Saving, restoring, recovering or retrying
- G06F11/1446—Point-in-time backing up or restoration of persistent data
- G06F11/1448—Management of the data involved in backup or backup restore
- G06F11/1451—Management of the data involved in backup or backup restore by selection of backup contents
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/60—Software deployment
- G06F8/61—Installation
- G06F8/62—Uninstallation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The invention discloses a Linux-based data real-time storage system, which comprises a host, a processing module, a memory and at least one storage device, wherein the processing module, the memory and the at least one storage device are arranged in the host; the memory is used for storing data to be written into the storage device, an index table I and an index table II; the storage equipment is provided with an actual volume and a backup file for storing real-time data; the processing module selects the writing and reading positions of the data according to the written or read data in the memory and combines the first index table and the second index table, and updates the first index table and the second index table. The invention discloses a storage method based on the Linux-based data real-time storage system, wherein snapshot states and increment states in the storage system work alternately. The invention greatly reduces I/O operation and improves system performance; meanwhile, the normal operation of the system is not affected and the mounting point of the backup file is not limited in the process of copying the data by the user.
Description
Technical Field
The invention belongs to the field of computers, and particularly relates to a Linux-based data real-time storage system and method.
Background
When the internet rapidly develops, data is a serious issue, and in the actual process, reasons such as artificial improper operation, hardware faults, malicious attacks and the like all cause data loss or damage, especially important data, will cause catastrophic loss, and the backup of the data becomes extremely important.
The conventional backup is various, and the defects of the current backup technology in writing are obvious. The online backup of the production system cannot be operated in real time, so that the service use is influenced, the backup files occupy additional physical disk space, the original once I/O disk writing operation of the data copying mode needs to be realized by three times, the equipment volume is required to be read only and cannot write data in the traditional backup, the incremental backup process is slower, the full backup is required after the incremental backup fails, and the like.
Disclosure of Invention
The invention aims to: aiming at the problems existing in the prior art, the invention provides a Linux-based real-time data storage system capable of effectively improving the data backup efficiency.
The technical scheme is as follows: in order to achieve the above object, the present invention provides a Linux-based real-time data storage system, which includes a host, and a processing module, a memory and at least one storage device disposed in the host; the memory is used for storing data to be written into the storage device, an index table I and an index table II; the storage equipment is provided with an actual volume and a backup file for storing real-time data; index table I and index tableTwo are both key mapping value structures, wherein keys in the index table one 1 The value is the block number of the data block in the actual volume 1 A modification status value for recording whether the data block is modified; key in index table two 2 Value calculated for hash from stored data content 2 Recording the data content of which the block number, the reference number and the reference number of the data block referenced by each data content exceed the reference threshold value; the processing module selects the writing and reading positions of the data according to the written or read data in the memory and combines the first index table and the second index table, and updates the first index table and the second index table.
Wherein, the storage devices are two, one is used for storing the actual volume; and the other for saving the backup file. In the snapshot mode, the storage system can reduce I/O operation of the same equipment; thereby increasing the speed of data storage.
In order to ensure normal data reading and copying of a user while the storage system is capable of modifying the data stored therein, the storage system creates a dev device node in the processing module, the dev device node being for performing a read operation.
The invention also provides a storage method of the Linux-based data real-time storage system, which specifically comprises the following steps: in an initial state, creating an index table I and an index table II in a memory according to original data in an actual volume; the system works alternately in a snapshot state and an increment state; in the snapshot state, the processing module processes the data written into the memory, queries in the first index table and the second index table, judges the position where the data in the memory needs to be stored, writes the data in the corresponding data block in the actual volume needing to be modified into the backup file, and updates the first index table and the second index table; then carrying out incremental synthesis on the data in the backup file and the data in the actual volume through a PageRank algorithm; in the incremental state, the processing module cuts off the data backed up in the backup file, processes the data written into the memory, queries in the first index table and the second index table, and updates the first index table and the second index table according to the result.
And the processing module firstly judges whether the received operation instruction is read operation or write operation, if the operation instruction is read operation, the read operation is sent to the dev equipment node, and the required data is read through the dev equipment node. This allows the read and write operations to be performed in parallel.
Further, the method for processing the data written into the memory by the processing module comprises the following steps: splitting the data written into the memory according to the memory page size, and respectively calculating the written data by using a hash algorithm in a multithreading parallel mode to generate a key unique to each data in the index table II 2 Values. Thus, the data storage and reading speed can be effectively improved.
Working principle: the invention mainly creates an index table I and an index table II in a memory, continuously updates the relation between the index table I and the index table II according to written and read data content, sets two working states according to the two tables, maintains block cache data on the storage device by backup files on the storage device in a snapshot state, and uses a cache initialization default accounting for 10% of the total size of an original volume for realizing copy-on-write in the backup process. When the mode is changed into the increment mode, the backup file is deleted, so that the space can be released for the rest part of the file system; the processing module only needs to track and write in the memory and synchronize the memory to the index table periodically according to the requirement, so that the performance is greatly improved. In this mode, the processing module simply tracks the block number of the disk modification and records it into the index table. For incremental backup synthesis, the data state is restored by more rapid synthesis through parallel synthesis tools that assist in development. In the invention, two storage devices can be used for respectively storing the actual volume and the backup file, so that the I/O writing is transferred to the other storage device, the I/O writing load of the same storage device at one time is lightened, and the performance of the whole storage system can be greatly improved.
The beneficial effects are that: compared with the prior art, the invention has the following obvious advantages: the invention can backup or copy the later data without affecting the normal operation of the system, limiting the mounting point of the backup file, greatly reducing the I/O operation and improving the performance, and accelerating the incremental synthesis, and has wide application prospect in the fields of backup disaster recovery systems and the like.
Drawings
FIG. 1 is a workflow diagram of a write operation performed by a storage system provided by the present invention in a snapshot state;
FIG. 2 is a flow chart of the operation of the memory system of the present invention for performing a write operation in an incremental state;
FIG. 3 is a flowchart illustrating a memory system according to the present invention performing a write operation and a read operation simultaneously in a snapshot state;
FIG. 4 is a flowchart illustrating the operation of the memory system of the present invention for performing both write and read operations in an incremental state.
Detailed Description
The following description of the embodiments of the present invention will be made more clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The embodiment discloses a real-time data storage method based on a linux system, which is applied to a storage system, wherein the storage system comprises a host, and a processing module, a memory and at least one storage device which are arranged in the host; the method is mainly executed on a host; in this embodiment, the processing module selects the location of writing and reading data according to the specific data to be written and read. The memory is used for storing data to be written into the storage device, an index table I and an index table II; the storage device is provided with an actual volume and a backup file for storing real-time data. The host is mainly a terminal device such as a server, a notebook computer, a desktop computer and the like; the storage device may be a hard disk, a magnetic disk, or other hardware storage device.
Wherein, the first index table is the structure of key mapping value, and the keys in the first index table 1 For data in real volumesBlock number, value of block 1 A modification status value for recording whether the data block is modified. In the initial state, the modification state value corresponding to each data block is marked as 0, and when the data block is modified, the modification state value is marked as 1. The index table two is also the structure of the key mapping value. Key in index table two 2 Value calculated for hash from stored data content 2 The block number, the number of references, and the number of references of the data blocks referenced by each data content exceed the reference threshold are recorded. The reference threshold is set to 10 in this embodiment.
The real-time data storage method based on the linux system disclosed in the embodiment mainly comprises data storage in two states, wherein one is in a snapshot state and the other is in an increment state; these two states alternate. In the initial state, an index table I and an index table II are created in the memory according to the original data in the actual volume. The processing module periodically sends the index table I and the index table II in the memory to the backup file.
Wherein, as shown in fig. 1, the writing operation of the snapshot state includes the following steps:
step 11: splitting the data written into the memory according to the memory page size, and respectively calculating the written data by using a hash algorithm in a multithreading parallel mode to generate a key unique to each data in the index table II 2 A value; according to the written data, namely the block number of the data block to be written, the modification state value of the corresponding data block is found in the index table I, and if the modification state value of the corresponding data block is 0 at the moment; step 12 is performed; otherwise, executing the step 13;
step 12: the modification state value of the corresponding data block is 0, cloning the data content of the corresponding data block in the actual volume at the moment, writing the data content into the backup file, and modifying the modification state value of the corresponding data block in the index table I; step 13 is executed;
step 13: according to the key obtained in step 11 for writing data 2 Value inquiry index table II, if the written data is recorded in index table II, judging that the data block to be written of said written data has the data, if so, updating reference numberIf the data is not available, the block number of the data block to be written is written into the index table II and the reference number is updated; if the written data is not recorded in the index table II, creating a new index relation in the table II;
step 14: judging the reference times of the written data, and if the reference times corresponding to the written data exceeds a reference threshold value, storing the data in a memory; if the reference number corresponding to the written data does not exceed the reference threshold value, writing the data into an actual volume of the storage equipment;
step 15: and performing incremental synthesis on the data in the backup file and the data in the actual volume through a PageRank algorithm.
As shown in fig. 2, when performing a write operation in an incremental state mainly includes the steps of:
step 21: the processing module cuts back the data backed up in the backup file, and reserves an index table I and an index table II;
step 22: splitting the data written into the memory according to the memory page size, and respectively calculating the written data by using a hash algorithm in a multithreading parallel mode to generate a key unique to each data in the index table II 2 A value;
step 23: searching a modification state value of the data block in the index table I according to the block number of the data block to be written in by the data and modifying the modification state value;
step 24: according to the metadata query index table II of the written data obtained in the step 22, if the written data is recorded in the index table II, judging whether the data block to be written of the written data has the data, if so, updating the reference number, and if not, writing the block number of the data block to be written into the index table II and updating the reference number; if the written data is not recorded in the index table II, creating a new index relation in the table II;
the processing module also creates a dev equipment node for the user to read, so that the user can carry out data backup and simultaneously modify the data in the storage system. When the read operation and the write operation are performed simultaneously, as shown in fig. 3, when the host is in a snapshot state, the user backup data mainly includes the following steps:
step 31: the processing module judges whether the received instruction is a read operation or a write operation, if the received instruction is the read operation, the steps 32-33 are executed, and if the received instruction is the write operation, the steps 34-38 are executed;
step 32: splitting the data to be read according to the memory page size, and calculating the written data by using a hash algorithm in a multithreading parallel mode to generate a key unique to each data in the index table II 2 A value;
step 33: transmitting the read operation to the dev device node; judging a modification state value corresponding to a data block corresponding to data to be read in the index table I; if the corresponding modification state value is 0, directly reading data from the data block number corresponding to the storage slave actual volume; if the corresponding modification state value is 1, according to the key of the read data 2 The value is searched for an index table II, if the index number of the data to be read is larger than a threshold value, the data is directly read from the memory, and if the index number of the data to be read is not larger than the threshold value, the corresponding data block number is found in the index table II, and the data is read from the corresponding data block number in the actual volume;
step 34: splitting the data written into the memory according to the memory page size, and respectively calculating the written data by using a hash algorithm in a multithreading parallel mode to generate a key unique to each data in the index table II 2 A value; according to the written data, namely the block number of the data block to be written, the modification state value of the corresponding data block is found in the index table I, and if the modification state value of the corresponding data block is 0 at the moment; step 35 is performed; otherwise, go to step 36;
step 35: the modification state value of the corresponding data block is 0, cloning the data content of the corresponding data block in the actual volume at the moment, writing the data content into the backup file, and modifying the modification state value of the corresponding data block in the index table I; step 36 is performed;
step 36: according to the key obtained in step 34 for the write data 2 Value lookup index table two if the write data is already in the index tableSecondly, recording, judging whether a data block to be written of the written data has the data, if so, updating the reference times, and if not, writing the block number of the data block to be written into the index table II and updating the reference times; if the written data is not recorded in the index table II, creating a new index relation in the table II;
step 37: judging the reference times of the written data, and if the reference times corresponding to the written data exceeds a reference threshold value, storing the data in a memory; if the reference number corresponding to the written data does not exceed the reference threshold value, writing the data into an actual volume of the storage equipment;
step 38: and performing incremental synthesis on the data in the backup file and the data in the actual volume.
When the read operation and the write operation are performed simultaneously, as shown in fig. 4, when the host is in an incremental state, the user backup data mainly includes the following steps:
step 41: the processing module judges whether the received instruction is a read operation or a write operation, if the received instruction is the read operation, the steps 42-43 are executed, and if the received instruction is the write operation, the steps 44-47 are executed;
step 42: splitting the data to be read according to the memory page size, and calculating the written data by using a hash algorithm in a multithreading parallel mode to generate a key unique to each data in the index table II 2 A value;
step 43: transmitting the read operation to the dev device node; judging a modification state value corresponding to a data block corresponding to data to be read in the index table I; if the corresponding modification state value is 0, directly reading data from the data block number corresponding to the storage slave actual volume; if the corresponding modification state value is 1, inquiring an index table II according to the metadata of the read data, if the index number of the data to be read is greater than a threshold value, directly reading the data from the memory, and if the index number of the data to be read is not greater than the threshold value, finding a corresponding data block number in the index table II, and reading the data from the corresponding data block number in the actual volume;
step 44: the processing module cuts back the backup data in the backup file, and reserves an index table I and an index table II;
step 45: splitting the data written into the memory according to the memory page size, and respectively calculating the written data by using a hash algorithm in a multithreading parallel mode to generate a key unique to each data in the index table II 2 A value;
step 46: searching a modification state value of the data block in the index table I according to the block number of the data block to be written in by the data and modifying the modification state value;
step 47: according to the key obtained in step 45 for writing data 2 Value inquiry index table II, if the written data is recorded in index table II, judging whether the data block to be written of said written data has the data already, if so, updating reference number, if not, writing the block number of the data block to be written into index table II and updating reference number; if the written data is not recorded in the index table II, creating a new index relation in the table II;
in this embodiment, two storage devices may be used to store the actual volume and the backup file, respectively, and in the snapshot state, if data needs to be written in the system, the data in the actual volume needs to be read out first, written in the backup file, and then new data is written in the actual volume. Thus, the data I/O operation of writing the actual volume data I/O and the data I/O operation of writing the backup volume on the other storage device can be performed simultaneously, the write load of the actual volume I/O at one time is lightened, and the performance is greatly improved.
The file content is copied in the snapshot mode, the copy operation is not needed in the increment mode, and the I/O operation is reduced. In this way, the processor can maintain consistency of the file system and be a complete snapshot even if the original volume is written. The snapshot data is managed and accessed by the processing module using the index table. By this implementation, our copy-on-write system is enabled to reliably maintain the complete time point data state of the original volume at the block level, while not affecting the normal use of the original volume.
The processing module may also dynamically transition between different states of the tracked storage device, benefiting from monitoring of the mount operation. The storage device will be in an "active" state most of the time during normal use. In this state, the storage device is fully initialized and all in-memory data structures are active. When the storage device driver is correctly installed, the storage device will start to keep an active state, and when the system triggers the actions of unloading the volume and the like, the active state is triggered to enter a dormant state.
Meanwhile, after the block driver is unloaded, the storage device is also converted into a dormant state no matter the user actively operates or the computer is normally closed for unloading. In short, the "hibernate" state is a state in which the storage device can be safely uninstalled by the operating system. When capturing a trigger offload system call of a storage device that is being tracked by a processor, the processor first writes all memory data structures of the storage device to a backup volume for subsequent use, then closes file descriptors of the backup volume, and closes all working processing threads of the storage device. When operating with one storage device in the storage system, if the original volume is unloaded by the operating system, the storage system will remain in a hibernation mode until the storage device is reinstalled again, at which point the storage system will transition to an active state. When two storage devices are adopted in the storage system to work, the performance is improved greatly, but the processing logic is more complicated, the states of the two storage devices need to be monitored, any unloading of the two storage devices cannot work normally, the system must be in a dormant state, the system needs to be in an atomic reference counting mode, and if the unloading fails, the system needs to be restored in situ, and only the two storage devices can be converted into an active state normally. Initializing the atomic count to 0, and when the two storage devices of the verification tracking are normal, and the storage device pointed by the file structure is consistent with the monitoring, performing two atomic operations respectively, converting the atomic value to 2, wherein any verification in the mode needs the atomic value to be 2. Meanwhile, the unloading process needs to carry out restoration judgment of the secondary unloading logic, and the processing thread of the storage device cannot be repeatedly closed, so that an exception can be caused. In the process of restoring the storage device, only if all the storage devices are restored, the block device can be transited to an active state, and the processing thread is restored.
Based on the fact that the device can be opened for many times at the same time, the node reading operation of dev device has no competition relationship, the data consistency is not affected, the writing of the same block has no conflict, and the operation between tasks has no sequence, so that the storage system provided by the invention does not affect the normal operation of the storage system when the data backup and the copy are carried out.
Claims (6)
1. The online real-time data storage system based on Linux is characterized in that: the system comprises a host, a processing module arranged in the host, a memory and at least one storage device; the memory is used for storing data to be written into the storage device, an index table I and an index table II; the storage equipment is provided with an actual volume and a backup file for storing real-time data; the first index table and the second index table are both in key mapping value structures, wherein keys in the first index table 1 The value is the block number of the data block in the actual volume 1 A modification status value for recording whether the data block is modified; key in index table two 2 Value calculated for hash from stored data content 2 Recording the data content of which the block number, the reference number and the reference number of the data block referenced by each data content exceed the reference threshold value; the processing module selects the data writing and reading positions according to the written or read data in the memory and combining the first index table and the second index table, and updates the first index table and the second index table;
in an initial state, creating an index table I and an index table II in a memory according to original data in an actual volume; the system works alternately in a snapshot state and an increment state; in the snapshot state, the processing module processes the data written into the memory, queries in the first index table and the second index table, judges the position where the data in the memory needs to be stored, writes the data in the corresponding data block in the actual volume needing to be modified into the backup file, and updates the first index table and the second index table; then carrying out incremental synthesis on the data in the backup file and the data in the actual volume through a PageRank algorithm; in the incremental state, the processing module cuts off the data backed up in the backup file, processes the data written into the memory, queries in the first index table and the second index table, and updates the first index table and the second index table according to the result.
2. The Linux-based online real-time data storage system of claim 1, wherein: the storage devices are two, one is used for storing the actual volumes; and the other for saving the backup file.
3. The Linux-based online real-time data storage system of claim 1, wherein: and creating a dev equipment node in the processing module, wherein the dev equipment node is used for executing reading operation.
4. The online real-time data storage method of Linux according to claim 1, wherein: and creating a dev equipment node in the processing module, judging whether the received operation instruction is read operation or write operation by the processing module, if the operation instruction is read operation, sending the read operation to the dev equipment node, and reading the required data through the dev equipment node.
5. The online real-time data storage method of Linux according to claim 1, wherein: the storage devices are two, one is used for storing the actual volumes; and the other for saving the backup file.
6. The online real-time data storage method of Linux according to claim 1, wherein: the processing module processes the data written into the memory by the following steps: splitting the data written into the memory according to the memory page size, and respectively calculating the written data by using a hash algorithm in a multithreading parallel mode to generate a key unique to each data in the index table II 2 Values.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010182063.9A CN111338853B (en) | 2020-03-16 | 2020-03-16 | Linux-based data real-time storage system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010182063.9A CN111338853B (en) | 2020-03-16 | 2020-03-16 | Linux-based data real-time storage system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111338853A CN111338853A (en) | 2020-06-26 |
| CN111338853B true CN111338853B (en) | 2023-06-16 |
Family
ID=71184447
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010182063.9A Active CN111338853B (en) | 2020-03-16 | 2020-03-16 | Linux-based data real-time storage system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111338853B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113360495B (en) * | 2021-04-20 | 2023-02-24 | 山东英信计算机技术有限公司 | Database query interruption recovery method, device, equipment and readable medium |
| CN113535258B (en) * | 2021-06-25 | 2022-09-27 | 杭州加速科技有限公司 | Pattern file loading method and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101650692A (en) * | 2009-07-02 | 2010-02-17 | 北京飞天诚信科技有限公司 | Method and device for modifying data in memory |
| CN108763582A (en) * | 2018-06-11 | 2018-11-06 | 腾讯科技（深圳）有限公司 | File search method, device, terminal and storage medium |
| US10282099B1 (en) * | 2017-10-27 | 2019-05-07 | Netapp, Inc. | Intelligent snapshot tiering |
| CN110062925A (en) * | 2016-10-28 | 2019-07-26 | Netapp股份有限公司 | Snapshot metadata arrangement for converging |
| CN110109778A (en) * | 2019-07-02 | 2019-08-09 | 南京云信达科技有限公司 | A kind of a large amount of small data file backup methods and restoration methods |
-
2020
- 2020-03-16 CN CN202010182063.9A patent/CN111338853B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101650692A (en) * | 2009-07-02 | 2010-02-17 | 北京飞天诚信科技有限公司 | Method and device for modifying data in memory |
| CN110062925A (en) * | 2016-10-28 | 2019-07-26 | Netapp股份有限公司 | Snapshot metadata arrangement for converging |
| US10282099B1 (en) * | 2017-10-27 | 2019-05-07 | Netapp, Inc. | Intelligent snapshot tiering |
| CN108763582A (en) * | 2018-06-11 | 2018-11-06 | 腾讯科技（深圳）有限公司 | File search method, device, terminal and storage medium |
| CN110109778A (en) * | 2019-07-02 | 2019-08-09 | 南京云信达科技有限公司 | A kind of a large amount of small data file backup methods and restoration methods |
Non-Patent Citations (1)
| Title |
|---|
| 基于数据通道的非结构化数据多存储系统;蒋静远;《中国优秀硕士学位论文全文数据库信息科技辑》;20160215;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111338853A (en) | 2020-06-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3724764B1 (en) | Write-ahead style logging in a persistent memory device | |
| US10437662B2 (en) | Crash recovery using non-volatile memory | |
| US7076622B2 (en) | System and method for detecting and sharing common blocks in an object storage system | |
| US8874515B2 (en) | Low level object version tracking using non-volatile memory write generations | |
| US7814057B2 (en) | Page recovery using volume snapshots and logs | |
| CN108664359B (en) | Database recovery method, device, equipment and storage medium | |
| Nam et al. | Assuring demanded read performance of data deduplication storage with backup datasets | |
| US6678809B1 (en) | Write-ahead log in directory management for concurrent I/O access for block storage | |
| CN109542682B (en) | Data backup method, device, equipment and storage medium | |
| CN108460045B (en) | Snapshot processing method and distributed block storage system | |
| US20070055710A1 (en) | BLOCK SNAPSHOTS OVER iSCSI | |
| US20050065962A1 (en) | Virtual data store creation and use | |
| US7681001B2 (en) | Storage system | |
| US9542279B2 (en) | Shadow paging based log segment directory | |
| CN102402471A (en) | Method and system for backing up data in real time based on snapshot function of memory array | |
| KR102262409B1 (en) | NVM-based file system and method for data update using the same | |
| US11537582B2 (en) | Data access method, a data access control device, and a data access system | |
| CN111338853B (en) | Linux-based data real-time storage system and method | |
| US10459641B2 (en) | Efficient serialization of journal data | |
| KR20170054767A (en) | Database management system and method for modifying and recovering data the same | |
| US10671567B2 (en) | System and method for optimized lock detection | |
| US7082445B2 (en) | Fast data copy using a data copy track table | |
| US11860840B2 (en) | Update of deduplication fingerprint index in a cache memory | |
| KR101686340B1 (en) | Method for efficient non-volatile cache load management for high capacity storage apparatus | |
| KR20140049327A (en) | Apparatus for logging and recovering transactions in database installed in a mobile environment and method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |