WO2015039569A1 - Dispositif de stockage de copies et procédé de stockage de copies - Google Patents

Dispositif de stockage de copies et procédé de stockage de copies Download PDF

Info

Publication number
WO2015039569A1
WO2015039569A1 PCT/CN2014/085993 CN2014085993W WO2015039569A1 WO 2015039569 A1 WO2015039569 A1 WO 2015039569A1 CN 2014085993 W CN2014085993 W CN 2014085993W WO 2015039569 A1 WO2015039569 A1 WO 2015039569A1
Authority
WO
WIPO (PCT)
Prior art keywords
copy
write request
memory table
slave
primary
Prior art date
Application number
PCT/CN2014/085993
Other languages
English (en)
Chinese (zh)
Inventor
梁晓豪
Original Assignee
华为技术有限公司
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
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2015039569A1 publication Critical patent/WO2015039569A1/fr

Links

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
    • G06F11/16—Error detection or correction of the data by redundancy in hardware
    • G06F11/20—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
    • G06F11/2053—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where persistent mass storage functionality or persistent mass storage control functionality is redundant
    • G06F11/2056—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where persistent mass storage functionality or persistent mass storage control functionality is redundant by mirroring
    • G06F11/2087—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where persistent mass storage functionality or persistent mass storage control functionality is redundant by mirroring with a common controller

Definitions

  • data is fragmented or partitioned.
  • multiple copies of the data slice or data area need to be backed up.
  • a write request sent by the client (Master write request) is sent first.
  • Master master copy
  • slave slave data for peer or isochronous backup.
  • the master and slave replicas of different pieces of data can be stored on the same node.
  • Embodiments of the present invention provide a replica storage device and a replica storage method, which improve the overall write capability and read capability of a distributed system, and reduce resource consumption of the distributed system.
  • an embodiment of the present invention provides a replica storage device, where the device includes: a distribution module;
  • the primary copy space includes a first memory table and a first copy file;
  • the first memory table is configured to store a record carried by the primary copy write request, and is in the first memory table When all the memory occupied by the record reaches the first preset value, all the records in the first memory table are stored. Stored in the first copy file;
  • the slave copy space includes a second memory table and a second copy file;
  • the second memory table is configured to store the record carried by the slave copy write request, and in the second memory table When the memory size occupied by all the records reaches the second preset value, all the records in the second memory table are stored in the second copy file;
  • the distribution module receives a write request sent by the sending node, and determines a type of the write request: if it is determined that the write request is a primary copy write request, submitting the primary copy write request to the primary copy space a first memory table; if it is determined that the write request is a copy request from a copy, submitting the slave copy write request to the second memory table of the slave copy space.
  • the number of the first memory table of the primary copy space is at least one; and/or the second memory table of the secondary copy space The number is at least one.
  • the primary copy space further includes a quick selection module, and the fast selection module is configured to receive the submission by the distribution module. Determining, according to the primary key of the record carried by the primary copy write request, determining, from each of the first memory tables, a target first memory table for storing the record carried by the primary copy write request, And submitting the primary copy write request to the target first memory table;
  • the slave copy space further includes a queue module; the queue module is configured to receive the slave copy write request submitted by the distribution module, insert the slave copy write request into a queue, and when the queue reaches a preset queue At the threshold, all slave copy write requests in the queue are committed to the second memory table.
  • the queue module is further configured to send, after the insert request is inserted into the queue, the sending node The copy writes a request response.
  • the first memory table is further configured to store the primary copy write request After the carried record, a slave copy write request carrying the record is sent to a preset copy storage node.
  • an embodiment of the present invention provides a copy storage method, where the method includes: applying to a copy storage device, where the copy storage device includes a distribution module, a primary copy space, and a secondary copy space;
  • the primary copy space includes a first memory table and a first copy file;
  • the first memory table is configured to store a record carried by the primary copy write request, and is occupied by all records in the first memory table.
  • the memory size reaches the first preset value, all the records in the first memory table are stored into the first copy file;
  • the slave copy space includes a second memory table and a second copy file;
  • the second memory table is configured to store the record carried by the slave copy write request, and is occupied by all records in the second memory table When the memory size reaches the second preset value, all the records in the second memory table are stored in the second copy file;
  • the copy storage method includes:
  • the distribution module receives a write request sent by the sending node, and determines a type of the write request: if it is determined that the write request is a primary copy write request, submitting the primary copy write request to the primary copy space An in-memory table;
  • the slave copy write request is submitted to the second memory table of the slave copy space.
  • the number of the first memory table of the primary copy space is at least one; and/or the second memory table of the secondary copy space The number is at least one.
  • the primary copy space further includes a fast selection module;
  • the secondary copy space further includes a queue module;
  • the submitting the primary copy write request to the first memory table of the primary copy space includes: the distribution module submitting the primary copy write request to the quick selection module;
  • the quick selection module receives the primary copy write request submitted by the distribution module, and determines, according to the primary key of the record carried by the primary copy write request, from each of the first memory tables in the primary copy space. a target first memory table for storing the record carried by the primary copy write request, and submitting the primary copy write request to the target first memory table;
  • the submitting the copy request from the copy to the second memory table of the slave copy space includes: the distribution module submitting the master copy write request to the queue module;
  • the queue module receives the slave copy write request submitted by the distribution module, inserts the slave copy write request into a queue, and when the queue reaches a preset queue threshold, all the slave copies in the queue are A write request is submitted to the second memory table.
  • the queue module receives the slave copy write request submitted by the distribution module, and inserts the slave copy write request After the queue, the method further includes: The queue module sends a copy write request response to the sending node.
  • the first to the third possible implementation manners of the second aspect in a fourth possible implementation manner, after the first memory table stores the record carried by the primary copy write request, The method further includes:
  • the first memory table sends a slave copy request carrying the record to a preset copy storage node.
  • the copy storage device and the copy storage method provided by the embodiments of the present invention divide the primary copy space and the secondary copy space independent of the primary copy and the secondary copy, and read and write data by using the primary copy space and the copy space as boundaries. Operation: By processing the write request to the primary copy and the write request from the copy, the processing delay of the write request is reduced, and the write capability of the distributed system is improved; the main storage is separately stored by using different files on the non-volatile storage medium.
  • FIG. 1 is a schematic structural diagram of a copy storage device according to an embodiment of the present invention.
  • FIG. 2 is another schematic structural diagram of a copy storage device according to an embodiment of the present invention.
  • FIG. 3 is still another schematic structural diagram of a copy storage device according to an embodiment of the present disclosure.
  • FIG. 4 is a flowchart of a method for storing a copy according to an embodiment of the present invention.
  • FIG. 5 is another flowchart of a copy storage method according to an embodiment of the present invention.
  • FIG. 6 is a signaling flowchart of a copy storage method according to an embodiment of the present invention.
  • the technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention.
  • the embodiments are a part of the embodiments of the invention, and not all of the embodiments. Based on the present invention All other embodiments obtained by those skilled in the art without creative efforts are within the scope of the present invention.
  • FIG. 1 is a schematic structural diagram of a replica storage device according to an embodiment of the present invention.
  • the copy storage device 100 provided in this embodiment includes: a distribution module 101, a primary copy space 102, and a secondary copy space 103, wherein:
  • the main copy space 102 includes a first memory table 1021 and a first copy file 1022.
  • the first memory table 1021 is configured to store a record carried by the primary copy write request, and is in the first memory table 1021. All the records in the first memory table 1021 are stored in the first copy file 1022 when the memory size occupied by all the records reaches the first preset value;
  • the second memory table 1031 is configured to store the record carried by the slave copy write request, and in the second memory table 1031 When all the memory occupied by the record reaches the second preset value, all the records in the second memory table 1031 are stored in the second copy file 1032;
  • the distribution module 101 receives a write request sent by the sending node, and determines the type of the write request: if it is determined that the write request is a primary copy write request, submitting the primary copy write request to the primary copy The first memory table 1021 of the space 102; if it is determined that the write request is a copy request from the copy, the copy request is submitted to the second memory table 1031 of the slave copy space 103.
  • the replica storage device 100 can be used in a scenario that needs to process massive data and services, such as a distributed database, a distributed file system, or a distributed storage system; the replica storage device 100 can be a node or a node in a distributed system.
  • the functional entity can store the primary copy and the secondary copy of different pieces of data, can perform the difference between the primary copy write request and the copy write request, and store the primary copy write request separately by using different files on the non-volatile storage medium.
  • the record and the record carried by the copy request, the non-volatile storage medium for example, may be a hard disk, a magnetic array, or the like. As shown in FIG.
  • the storage space of the copy storage device 100 is divided into a primary copy space 102 and a secondary copy space 103, and the primary copy space 102 is responsible for processing the primary copy write request and storing the primary copy write request.
  • Recording, from the copy space 103, is responsible for processing the slave copy write request and storing the record carried by the slave copy write request; wherein the master copy space 102 and the slave copy space 103 are distinguished from each other, for example, The two are distinguished by means of directories, file names or file attributes.
  • the distribution module 101 of the replica storage device 100 receives the write request sent by the sending node, first determines the type of the write request, and the type of the write request may be a primary copy write request or a copy write request: if the distribution Module 101 determines that the write request is a primary copy write request, and then writes the primary copy Requesting a first memory table 1021 submitted to the primary copy space 102; the first memory table 1021 storing a record carried by the primary copy write request; when all records stored in the first memory table 1021 are occupied When the memory size reaches the first preset value, all the records in the first memory table 1021 are merged and dumped into the first copy file 1022 on the non-volatile storage medium.
  • the distribution module 101 determines that the write request is a copy request from a copy, submit the copy request to the second memory table 1031 of the slave copy space 103; the second memory table 1031 stores the Recording the record carried by the request from the copy; when the memory size occupied by all the records stored in the second memory table 1031 reaches a second preset value, all the records in the second memory table 1031 are merged and dumped. Go to the second copy file 1032 on the non-volatile storage medium.
  • the copy storage device provided by the embodiment of the present invention separately divides the primary storage space and the secondary copy into separate storage spaces, and uses the divided primary copy space and the secondary copy space to separately perform a write request to the primary copy and a separate write request from the copy space, thereby reducing
  • the processing delay of the write request improves the write capability of the distributed system; the primary and secondary copies are stored in the storage by storing the records carried by the primary and secondary copy write requests by using different files on the non-volatile storage medium. It can be separated, which reduces the number and size of files that need to be queried when querying the primary copy record, improves the reading ability of the distributed system, and reduces the amount of data to be filtered when the primary and secondary copy ranges are changed. , reducing the resource consumption of distributed systems.
  • FIG. 2 is another schematic structural diagram of a replica storage device according to an embodiment of the present invention. This embodiment is implemented based on the embodiment shown in Fig. 1. As shown in FIG. 2, the copy storage device 200 provided in this embodiment also includes: a distribution module 101, a primary copy space 102, and a secondary copy space 103.
  • the number of the first memory table 1021 of the primary copy space 102 is at least one; and/or the number of the second memory tables 1031 of the secondary copy space 103 is at least one.
  • the primary copy space 102 further includes a quick selection module 1023.
  • the fast selection module 1023 is configured to receive the primary copy write request submitted by the distribution module 101, and carry the request according to the primary copy write request. a primary key of the record, determining, from each of the first memory tables 1021, a target first memory table for storing a record carried by the primary copy write request, and submitting the primary copy write request to the target An in-memory table;
  • the slave copy space 103 further includes a queue module 1033; the queue module 1033 is configured to receive the slave copy write request submitted by the distribution module 101, insert the slave copy write request into a queue, and when the queue reaches When a preset queue threshold is reached, all slave copy write requests in the queue are committed to the second memory table 1031. Further, the queue module 1033 is further configured to send a copy write request response to the sending node after inserting the slave copy write request into the queue.
  • the first memory table 1021 is further configured to send, after storing the record carried by the primary copy write request, a secondary copy write request carrying the record to a preset copy storage node.
  • one or more of the first memory tables 1021 may exist in one node at the same time; the quick selection module 1023 of the primary replica space 102 receives the submission by the distribution module 101.
  • the corresponding value is calculated according to the primary key of the record carried by the primary copy write request, and is determined from each of the first memory tables 1021 for storing the primary according to the corresponding interval of the corresponding value.
  • the copy writes the target first memory table of the record carried by the request, and submits the main copy write request to the target first memory table, wherein the manner of calculating the corresponding value may be, for example, a hash, a prefix, or the like.
  • the quick selection module 1023 before submitting the primary copy write request to the target first memory table, first The primary copy write request is written to the log file.
  • the first memory table 1021 receives and processes the primary copy write request submitted by the quick selection module 1023, and stores the record carried by the primary copy write request; the first memory table 1021 stores the primary copy After the record carried by the request is written, the slave copy write request carrying the record is sent to the preset copy storage node according to the routing table: when the mode of the write slave copy is the peer mode, waiting for the copy storage node to return Writing a request response from the copy, and sending a primary copy write request response to the sending node after receiving the secondary copy write request response; not waiting for the copy storage node when the mode of the write secondary copy is in the meta mode The returned slave write request response, but the master copy write request response is sent directly to the sending node after sending the slave copy write request to the copy storage node.
  • one or more of the second memory tables 1031 may be present in one node at the same time.
  • the queue module 1033 of the copy space 103 determines whether the queue reaches a preset queue threshold (such as time or queue size): if the queue does not reach the stated Queue threshold, the queue module 1033 inserts the slave copy write request into the queue, and then sends a copy write request response to the sending node; if the queue reaches the queue threshold, the queue module 1033 switches the queue to An empty queue, then inserting the slave copy write request into the empty queue, and sending a copy write request response to the sending node, and simultaneously, all the slave copy write requests in the queue reaching the queue threshold Submit (merge dump) to the second memory table 1031.
  • the queue module 1033 submits the slave copy write request to the target second memory. Prior to Table 1031, the slave copy write request
  • the copy storage device allocates a plurality of first memory tables by using a primary copy space, and determines a target first memory table from each of the first memory tables according to a key value of the record carried by the primary copy write request. , submitting the primary copy write request to the first target memory table, reducing the write delay of the primary copy write request; returning the write request from the copy node after inserting the received copy request from the copy into the queue In response, all of the slave copy write requests in the queue are submitted to the second memory table in the copy space when the queue reaches the queue threshold, which reduces the write latency from the copy write request; thereby improving the write capability of the distributed system.
  • FIG. 3 is still another schematic structural diagram of a copy storage device according to an embodiment of the present invention. This embodiment is implemented based on the embodiment shown in Fig. 2. As shown in FIG. 3, the copy storage device 300 provided in this embodiment also includes: a distribution module 101, a primary copy space 102, and a secondary copy space 103.
  • the copy storage device 300 further includes a file management module 301.
  • the file management module 301 is configured to receive a storage request submitted by the memory table, and determine the indication information carried in the storage request: if the storage request is determined to be the The storage request submitted by the first memory table 1021 merges all the records in the first memory table 1021 into the first copy file 1022 on the non-volatile storage medium; if it is determined that the storage request is the first When the two memory tables 1031 are submitted, all the records in the second memory table 1031 are merged and dumped into the second copy file 1032 on the non-volatile storage medium.
  • the storage request is submitted to the file management module 301; the file management module 301 determines to determine the When storing the storage request submitted by the first memory table 1021, all the records in the first memory table 1021 are merged and dumped into the first copy file 1022 on the non-volatile storage medium.
  • the file management module 301 determines to determine that the storage request is When the second memory table 1031 submits the storage request, all the records in the first memory table 1031 are merged and dumped into the first copy file 1032 on the non-volatile storage medium.
  • the file management module 301 is further configured to process a primary copy space and a file merge request from the copy space.
  • the file management module 301 may include a primary copy space file request queue, a copy space file request queue, and a request scheduling unit: the primary copy file merge request is submitted to the primary copy space file request queue, and the copy file merge request is submitted to the copy file.
  • a request queue is requested from the copy space file, and the request scheduling unit triggers execution of the file merge request according to conditions such as priority.
  • Request scheduling algorithm For example, it can be: Each request scheduling unit first queries the primary copy space file request queue (with high priority), and schedules a file merge request from the copy space file request queue without a request, wherein the primary copy The file operation priority of the space is greater than the copy space.
  • FIG. 4 is a flowchart of a copy storage method according to an embodiment of the present invention.
  • the method may be applied to a copy storage device, where the copy storage device includes a distribution module, a primary copy space, and a secondary copy space; the primary copy space includes a first memory table and a first copy file; a table, configured to store the record carried in the primary copy write request, and when all the records in the first memory table occupy a first preset value, all the records in the first memory table are Stored in the first copy file; the secondary copy space includes a second memory table and a second copy file; the second memory table is configured to store the record carried by the copy request, and When the memory size occupied by all the records in the second memory table reaches the second preset value, all the records in the second memory table are stored in the second copy file.
  • the method can be used in scenarios where large amounts of data and services need to be processed, such as distributed databases, distributed file systems, or distributed storage systems. As shown in FIG. 4, the method includes:
  • the distribution module receives a write request sent by the sending node.
  • the distribution module determines a type of the write request.
  • the write request is a primary copy write request
  • the write request is a copy request from a copy
  • the number of the first memory table in the primary copy space is at least one; and/or the number of the second memory table in the secondary copy space is at least one.
  • the primary copy space further includes a quick selection module; the submitting the primary copy write request to the first memory table of the primary copy space, including: the distribution module submitting the primary copy write request Providing the quick selection module, the quick selection module, receiving the primary copy write request submitted by the distribution module, according to the primary key of the record carried by the primary copy write request, from the primary copy space Determining, in the first memory table, a target first memory table for storing a record carried by the primary copy write request, and submitting the primary copy write request to the target first memory table.
  • the slave copy space further includes a queue module; the submitting the slave copy write request to the second memory table of the slave copy space, comprising: the distribution module submitting the master copy write request to the queue a module; receiving, by the queue module, the slave copy request submitted by the distribution module, The slave copy write request insertion queue, when the queue reaches a preset queue threshold, submits all slave copy write requests in the queue to the second memory table.
  • the method further includes: the queue module sending the slave node to the sending node The copy writes a request response.
  • the method further includes: sending, by the first memory table, a slave carrying the record to a preset copy storage node. A copy write request.
  • the copy storage method provided by the embodiment of the present invention separately divides the independent storage space by using the primary copy space and the secondary copy space, and separately uses the divided primary copy space and the secondary copy space to separately write the primary copy write request and the copy copy request, thereby reducing
  • the processing delay of the write request improves the write capability of the distributed system; the primary and secondary copies are stored in the storage by storing the records carried by the primary and secondary copy write requests by using different files on the non-volatile storage medium. It can be separated, which reduces the number and size of files that need to be queried when querying the primary copy record, improves the reading ability of the distributed system, and reduces the amount of data to be filtered when the primary and secondary copy ranges are changed. Resource consumption of distributed systems.
  • FIG. 5 is another flowchart of a copy storage method according to an embodiment of the present invention. This embodiment is implemented based on the embodiment shown in FIG. 4, and the execution subject is a copy storage device. As shown in FIG. 5, the method includes:
  • the distribution module receives a write request sent by the sending node.
  • step 502. Determine the type of the write request: if it is determined that the type of the write request is a primary copy write request, then go to step 503; if it is determined that the write request is a copy write request, then go to step 506.
  • step 505 Send, according to the routing table, a slave copy write request that carries the record to a preset copy storage node; when the mode of the write slave copy is a peer mode, wait for the slave copy write request response returned by the copy storage node ( Heterogeneous mode does not need to wait); Go to step 509.
  • step 506. Determine whether the queue reaches the queue threshold: if yes, go to step 507; if no, go to step 508.
  • FIG. 6 is a signaling flowchart of a copy storage method according to an embodiment of the present invention. The method is performed by a transmitting node, a first copy storage node, and a second copy storage node. The first copy storage node and the second copy storage node may use any of the copy storage devices provided by the embodiments of the present invention. As shown in FIG. 6, the method includes:
  • the sending node sends a primary copy write request to the first copy storage node.
  • the first copy storage node receives the write request, and determines that the type of the write request is a primary copy write request; according to the primary key of the record carried by the primary copy write request, from the primary copy of the first copy storage node A target first memory table for storing a record carried by the master copy write request is determined in each of the first memory tables of the space, and the master copy write request is submitted to the target first memory table.
  • the first copy storage node sends a secondary copy write request carrying the record to a preset second copy storage node.
  • the second copy storage node receives the write request, determines that the type of the write request is a copy request from a copy, and determines whether the queue reaches a preset queue threshold: if the queue threshold is not reached, the secondary copy is A write request is inserted into the queue; if the queue threshold is reached, the queue is switched to an empty queue, and then the slave copy write request is inserted into the empty queue.
  • the second copy storage node sends a copy write request response to the first copy storage node.
  • all the slave copy write requests in the queue that reach the queue threshold are batch submitted (merged dumped) to the second memory table.
  • the first replica storage node sends a primary replica write request response to the sending node.
  • the disclosed apparatus and method may be implemented in other ways.
  • the device embodiments described above are merely illustrative.
  • the division of the unit is only a logical function division.
  • there may be another division manner for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored, or not executed.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be electrical, mechanical or otherwise.
  • each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
  • the above integrated unit can be implemented in the form of hardware or in the form of hardware plus software functional units.
  • the above-described integrated unit implemented in the form of a software functional unit can be stored in a computer readable storage medium.
  • the software functional unit is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor to perform the method of various embodiments of the present invention.
  • a computer device which may be a personal computer, a server, or a network device, etc.
  • the foregoing storage medium includes: a U disk, a removable hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like, which can store program code. .

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

Un mode de réalisation de l'invention concerne un dispositif de stockage de copies et un procédé de stockage de copies permettant de : partitionner un espace maître et un espace esclave mutuellement dépendants pour un maître et un esclave, et effectuer respectivement des opérations de lecture et d'écriture de données, etc., dans l'espace maître et l'espace esclave ; traiter de manière différenciée une demande d'écriture sur le maître et une demande d'écriture sur l'esclave, ce qui raccourcit le délai de traitement de la demande d'écriture et améliore la capacité d'écriture d'un système distribué ; utiliser différents fichiers présents sur un support de stockage non volatile pour stocker respectivement les enregistrements transportés dans la demande d'écriture sur le maître et la demande d'écriture sur l'esclave de façon à stocker séparément le maître et l'esclave, ce qui diminue le nombre et la taille des fichiers à trouver lors de la recherche des enregistrements du maître, et améliore la capacité de lecture du système distribué, tout en réduisant le trafic de données à filtrer lorsque la plage du maître et de l'esclave varie, et en réduisant la consommation de ressources du système distribué.
PCT/CN2014/085993 2013-09-22 2014-09-05 Dispositif de stockage de copies et procédé de stockage de copies WO2015039569A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310432857.6 2013-09-22
CN201310432857.6A CN104461768B (zh) 2013-09-22 2013-09-22 副本存储装置及副本存储方法

Publications (1)

Publication Number Publication Date
WO2015039569A1 true WO2015039569A1 (fr) 2015-03-26

Family

ID=52688228

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/085993 WO2015039569A1 (fr) 2013-09-22 2014-09-05 Dispositif de stockage de copies et procédé de stockage de copies

Country Status (2)

Country Link
CN (1) CN104461768B (fr)
WO (1) WO2015039569A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018223262A1 (fr) * 2017-06-05 2018-12-13 华为技术有限公司 Procédé, dispositif et équipement de traitement de transaction
CN111290850A (zh) * 2018-12-07 2020-06-16 中国移动通信集团陕西有限公司 一种数据存储方法、装置及设备
CN111611068A (zh) * 2019-02-25 2020-09-01 清华大学 分布式系统中的数据写方法、服务器及客户端

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107515723B (zh) * 2016-06-16 2020-04-24 伊姆西Ip控股有限责任公司 用于管理存储系统中的存储器的方法和系统
CN106406750A (zh) * 2016-08-23 2017-02-15 浪潮(北京)电子信息产业有限公司 一种数据操作方法及系统
CN106445737B (zh) * 2016-09-12 2023-03-10 恒为科技（上海）股份有限公司 一种多备份启动方法
CN106445409A (zh) * 2016-09-13 2017-02-22 郑州云海信息技术有限公司 一种分布式块存储的数据写入方法及装置
CN106331166B (zh) * 2016-10-11 2019-09-06 杭州宏杉科技股份有限公司 一种存储资源的访问方法及装置
CN108647228B (zh) * 2018-03-28 2021-08-24 中国电力科学研究院有限公司 可见光通信大数据实时处理方法和系统
CN111090629B (zh) * 2019-12-24 2024-02-06 上海达梦数据库有限公司 一种数据文件存储方法、装置、设备及存储介质
CN113076220B (zh) * 2020-01-06 2024-05-31 阿里巴巴集团控股有限公司 数据处理方法、装置、电子设备及计算机可读介质

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101706802A (zh) * 2009-11-24 2010-05-12 成都市华为赛门铁克科技有限公司 一种数据写入、修改及恢复的方法、装置及服务器
US20110161607A1 (en) * 2009-04-14 2011-06-30 Hitachi ,Ltd. Storage system and control method therefor
CN103294787A (zh) * 2013-05-21 2013-09-11 成都市欧冠信息技术有限责任公司 分布式数据库系统的多副本存储方法和系统

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110161607A1 (en) * 2009-04-14 2011-06-30 Hitachi ,Ltd. Storage system and control method therefor
CN101706802A (zh) * 2009-11-24 2010-05-12 成都市华为赛门铁克科技有限公司 一种数据写入、修改及恢复的方法、装置及服务器
CN103294787A (zh) * 2013-05-21 2013-09-11 成都市欧冠信息技术有限责任公司 分布式数据库系统的多副本存储方法和系统

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018223262A1 (fr) * 2017-06-05 2018-12-13 华为技术有限公司 Procédé, dispositif et équipement de traitement de transaction
CN111290850A (zh) * 2018-12-07 2020-06-16 中国移动通信集团陕西有限公司 一种数据存储方法、装置及设备
CN111290850B (zh) * 2018-12-07 2024-04-09 中国移动通信集团陕西有限公司 一种数据存储方法、装置及设备
CN111611068A (zh) * 2019-02-25 2020-09-01 清华大学 分布式系统中的数据写方法、服务器及客户端
CN111611068B (zh) * 2019-02-25 2022-07-15 清华大学 分布式系统中的数据写方法、服务器及客户端

Also Published As

Publication number Publication date
CN104461768B (zh) 2018-08-14
CN104461768A (zh) 2015-03-25

Similar Documents

Publication Publication Date Title
WO2015039569A1 (fr) Dispositif de stockage de copies et procédé de stockage de copies
US11153380B2 (en) Continuous backup of data in a distributed data store
JP6522812B2 (ja) 分散型データベースシステム用高速クラッシュ回復
JP6538780B2 (ja) 分散型データベースシステムのシステム全体のチェックポイント回避
US10579610B2 (en) Replicated database startup for common database storage
US10496669B2 (en) System and method for augmenting consensus election in a distributed database
CN110502507B (zh) 一种分布式数据库的管理系统、方法、设备和存储介质
US9946735B2 (en) Index structure navigation using page versions for read-only nodes
US9703803B2 (en) Replica identification and collision avoidance in file system replication
KR101862718B1 (ko) 블록-기반 저장을 위한 데이터 볼륨 지속성 상태 감소
US20220100600A1 (en) Method and apparatus to neutralize replication error and retain primary and secondary synchronization during synchronous replication
EP3803618B1 (fr) Transactions distribuées dans un environment de stockage en cloud avec directoire hierarchique
EP3811596B1 (fr) Espace de nommage hiérarchique à forte cohérence et évolutivité horizontale
US10725691B1 (en) Dynamic recycling algorithm to handle overlapping writes during synchronous replication of application workloads with large number of files
US11593016B2 (en) Serializing execution of replication operations
US20170024315A1 (en) Efficient garbage collection for a log-structured data store
US9424140B1 (en) Providing data volume recovery access in a distributed data store to multiple recovery agents
US10885023B1 (en) Asynchronous processing for synchronous requests in a database
WO2014124576A1 (fr) Procédé de gestion de mémoire, dispositif de gestion de mémoire et système numa
CN113901024A (zh) 数据存储系统、数据存储方法、可读介质和电子设备
WO2021057108A1 (fr) Procédé de lecture de données, procédé d'écriture de données et serveur
US20130006920A1 (en) Record operation mode setting
US10387384B1 (en) Method and system for semantic metadata compression in a two-tier storage system using copy-on-write
WO2014180395A1 (fr) Procédé et système de stockage de fusion de données massives
CN114661690A (zh) 多版本并发控制和日志清除方法、节点、设备和介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14845497

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14845497

Country of ref document: EP

Kind code of ref document: A1