CN103051681A - Collaborative type log system facing to distribution-type file system - Google Patents
Collaborative type log system facing to distribution-type file system Download PDFInfo
- Publication number
- CN103051681A CN103051681A CN2012105216766A CN201210521676A CN103051681A CN 103051681 A CN103051681 A CN 103051681A CN 2012105216766 A CN2012105216766 A CN 2012105216766A CN 201210521676 A CN201210521676 A CN 201210521676A CN 103051681 A CN103051681 A CN 103051681A
- Authority
- CN
- China
- Prior art keywords
- module
- request
- log
- log recording
- record
- 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.)
- Granted
Links
Images
Landscapes
- Debugging And Monitoring (AREA)
- Computer And Data Communications (AREA)
Abstract
The invention discloses a collaborative type log system facing to a distribution-type file system, and belongs to a distribution type file system. According to the collaborative type log system facing to the distribution-type file system, the problem that the system recovery operation can not be carried out caused by that a log record is lost due to equipment faults can be solved, and the write performance and the recovery performance of the distribution-type file system are improved. The collaborative type log system facing to the distribution-type file system comprises a log request module, a distribution module, N transmission modules, N consistency guaranteeing modules, N record adding modules, N record recovery modules and N restoring modules, wherein the log request module and the distribution module are operated on a metadata management node or a data storage node; and the transmission module, the consistency guaranteeing module, the record adding module, the record recovery module and the restoring module are all operated on each data storage node. According to the collaborative type log system facing to the distribution-type file system, which is disclosed by the invention, on the premise that the log reliability is guaranteed and the normal log action is performed, the write performance and the recovery performance of the distribution-type file system are improved by the characteristics that the network transmission efficiency in the distribution file system is higher than the low-speed transmission efficiency of storage equipment.
Description
Technical field
The invention belongs to distributed file system, be specifically related to a kind of collaboration type log system of Based on Distributed file system, to improve write performance, the restorability of distributed file system.
Background technology
The Internet and multimedia application are so that digital information is explosive growth trend, also increasing to the demand of storage aspect quality and quantity two, in the face of such mass data storage demand, the large-scale distributed file system that runs on hundreds and thousands of even thousands of memory devices arises at the historic moment, and is applied to multiple fields rapidly.Distributed file system comprises a plurality of metadata management nodes and a plurality of client node, and each client node connects respectively each metadata management node, and each metadata management node connects again a plurality of data memory nodes.The memory device that uses along with distributed file system increases, and the probability that distributed file system meets with unexpected event also increases greatly, and for safeguards system provides time of service, the correctness of assurance system operation must the usage log system.Many distributed file systems such as Lustre, Google FileSystem, Hadoop File System etc. has used log system.
Log system is made of each generic module and the daily record of admin log record, at first file system is opened up additional storage space at disk for log system, making amendment operation to file data before, file system can and be written in the daily record these operation generation log recordings, fixedly log recording.In this case, in case meet with the machine of delaying, log system just can utilize log recording in the daily record with the state of file system recovery before the machine of delaying.The greatest drawback of log system be All Files more new capital must be first fixing log recording, each file update operation all will be experienced twice write operation, the fixedly log recording operation of write-once daily record, once be only real writing in files system, extra fixedly log recording operation can cause file system performance to descend, particularly use this low speed storage device of disk to fix log recording, file system performance descends even more serious.
Because metadata management node or data memory node all are independent operatings under the distributed file system, all have the memory device of oneself and the log recording of access oneself alone.When equipment fault appears in memory device, may cause log recording to read, cause system's recovery operation to carry out, finally cause overall distribution formula file system to use.
Summary of the invention
The invention provides a kind of collaboration type log system of Based on Distributed file system, solve because equipment fault causes log recording to lose the problem that the system recovery operation of causing can't be carried out, improve write performance, the restorability of distributed file system.
For convenience of description, collaboration type log system provided by the present invention operates on N the data memory node of a metadata management node and connection thereof, stores the daily record of self on each data memory node, N 〉=2;
The collaboration type log system of a kind of Based on Distributed file system provided by the present invention, comprise that log request module, distribution module, a N transport module, a N consistency guarantee that module, a N record add module, a N record recycling module and N recovery module, N 〉=2; It is characterized in that:
(1) log request module runs on metadata management node or data memory node,
When receiving the retouching operation order of file system, judging that record adds module and whether sends the time-out request, is the recovery end signal that then waits recycling module to be recorded to send, and receives this recovery end signal, form again log recording and add request, add request otherwise directly form log recording; Then log recording is added request and send to distribution module; Log recording adds request and is comprised of request type and many log recordings, and request type is for adding record, and each bar log recording all has sequence number;
When receiving the recovery order of file system, form recovery request, recovery request sent to distribution module and wait for, receive recover each bar log recording that module transmits after, transfer them to file system; Recovery request is comprised of request type and a plurality of log recording sequence number, and request type is for recovering;
(2) distribution module runs on metadata management node or data memory node, sends connection request to each transport module respectively, sets up the Internet Transmission passage; After receiving log recording interpolation request or recovery request, it is arrived each transport module by the multicast mode parallel transmission; Connection request is comprised of request type, public network address, destination network addresses and port numbers, request type is for connecting, public network address is that IP address, the destination network addresses of distribution module place node is the IP address of transport module place node, and port numbers is the portal number that the Internet Transmission passage takies;
(3) N transport module is identical, runs on respectively each data memory node, receive connection request after, set up the Internet Transmission passage; After receiving log recording interpolation request, the consistency that sends it to the place data memory node guarantees module; After receiving recovery request, send it to the recovery module of place data memory node; After receiving record recovery request, send it to and respectively record recycling module;
(4) N consistency guarantees that module is identical, run on respectively each data memory node, guarantee that to the consistency of other each data memory node module sends institute's log recording that receives and adds log recording maximum sequence number in asking, the consistency that receives simultaneously other each data memory node guarantees the maximum sequence number of log recording that module sends, and the number of the identical log recording that the receives maximum sequence number of the maximum sequence number of log recording that sends with self of record, judge whether this number surpasses half of data memory node sum, be then the log recording that receives to be added the record interpolation module that request sends to the place data memory node, otherwise continue to wait for that the next log recording of transport module adds request;
(5) N record interpolation module is identical, run on respectively each data memory node, judge whether the daily record remaining space satisfies the total length that described log recording adds each bar log recording in the request on the data memory node of place, is then each bar log recording to be added in the daily record; Otherwise the transport module to the place data memory node sends record recovery request, and sends the time-out request to the log request module; Record to reclaim and ask by request type and reclaim record count to form, request type is to reclaim, and the recovery record count is M times that log recording adds log recording number in the request, M 〉=2;
(6) N record recycling module is identical, run on respectively each data memory node, after receiving that record reclaims request, the daily record original position begins to reclaim the log recording that reclaims record count the record recovery request from the notebook data memory node, and recovery is finished backward log request module and sent the recovery end signal;
(7) N to recover module identical, run on respectively each data memory node, receive recovery request after, according to the log recording sequence number in the recovery request log recording of corresponding sequence number is transferred to the log request module.
Client node sends the metadata operation request to the metadata management node, and the metadata management node produces corresponding operation log recording needs storage, forms log recording by the log request module and adds request; Need a part of log recording to recover after the metadata management node machine of delaying is restarted, form recovery request by the log request module; Distribution module operates in equally on the metadata management node and has set up the Internet Transmission passage with a plurality of data memory nodes; Transport module operates in and receives request on the data memory node; Consistency guarantees that module operates on the data memory node, guarantee module communication by the consistency with each data memory node, send log recording and add max log LSN in the request, only receive the identical ability with it of numbering that the data memory node above half sends and log recording is added request send to record interpolation module; Record interpolation module operates in the interpolation of carrying out log recording on the data memory node and fixes; The record recycling module operates on the data memory node and carries out the log recording reclaimer operation and notify the daily record request module after the receiving record recovery request; Recovering module operates on the data memory node according to the log recording sequence number in the recovery request log recording of corresponding sequence number is transferred to the log request module.
The present invention utilizes network transmission efficiency in the distributed file system to be higher than the characteristics of memory device low speed transmissions efficient, the daily record data of single metadata management node or data memory node generation is arrived each data memory node by network with the multicast mode parallel transmission, the log recording fixing operation is assigned on a plurality of nodes carries out simultaneously, improved the write performance of individual node.Simultaneously since with important log recording distributed storage on a plurality of data memory nodes, the expendable harm of having avoided the storage device failure of individual data memory node to cause has improved the reliability of distributed file system integral body.
When recovering in system, the present invention obtains the partial log record by express network is parallel from each data memory node, finally be spliced to form complete log information, compare the individual data memory node and read complete log recording from memory device and carry out recovery operation and have higher performance, improved the performance that distributed file system is recovered.
The present invention meets the basic function requirement of log system, comprise availability and versatility, and do not reduce file system performance, solved because equipment fault causes log recording to lose the problem that the system recovery operation of causing can't be carried out, improve write performance and the restorability of distributed file system.
Description of drawings
Fig. 1 is the distributed file system structural representation;
Fig. 2 is the embodiments of the invention schematic diagram;
Fig. 3 is the log record schematic diagram;
Fig. 4 is that log recording is distributed to each data memory node process schematic diagram;
Fig. 5 is that data memory node carries out the schematic flow sheet that log recording adds and reclaims;
Fig. 6 is journal recovery process schematic diagram.
Embodiment
The present invention is further described below in conjunction with drawings and Examples.
As shown in Figure 1, distributed file system is made of a plurality of metadata management nodes, a plurality of client node, a plurality of data memory node and the express network that is used for interconnecting between node.Each client node connects respectively each metadata management node, and each metadata management node connects again a plurality of data memory nodes, and client node is the client terminal that uses whole distributed file system; The metadata management node is in charge of the metadata information of file system; Data memory node is responsible for providing memory space and stores service to client node.
As shown in Figure 2, embodiments of the invention comprise that log request module, distribution module, 3 transport modules, 3 consistency guarantee that module, 3 records add module, 3 record recycling modules and 3 recovery modules; Present embodiment operates on 3 data memory nodes of a metadata management node and connection thereof, stores the daily record of self on each data memory node;
Log request module and distribution module run on the metadata management node, all move transport module, consistency assurance module, record interpolation module, record recycling module on 3 data memory nodes and recover module;
As shown in Figure 3, a log recording comprises seven parts: magic number (Magic Number), daily record numbering (Journal Number), LSN (Sequence Number), record type (Type), record length (Length), log recording data (Log data) and verification with (Checksum).Wherein, magic number is used for verifying the legitimacy of log recording; The daily record numbering is exactly the numbering of the daily record under this log recording, may manage the daily record of a plurality of systems on the data memory node, so need usage log to number to determine certain bar log recording is to belong to which daily record; LSN is this sequence number in daily record of log recording, the log request module can be calculated log recording according to this LSN and be distributed on which data memory node, and record interpolation module also can be inserted into log recording on the corresponding position according to this LSN and manage; Record type is used for representing the type of log recording, comprises that affairs begin type, write operation type, deletion action type, affairs end type etc.; Record length is used for representing the space size of these log recording needs, because log record types is different, the log recording data length also may change; The log recording data are exactly the record data that file system produces, and file system is delayed and can be recovered by these data behind the machine; The integrality of log recording storage is verified in verification and being used for.
As shown in Figure 4, present embodiment need to be that 0 to 2 log recording adds three data memory nodes to LSN; Specifically describe as follows:
The log request module forms log recording interpolation request and sends to distribution module; This log recording adds request and is comprised of request type and three log recordings, and request type is for adding record, and the LSN of three log recordings is respectively 0,1,2;
Distribution module has been set up the Internet Transmission passage with three data memory nodes; Receiving log recording adds after the request it by the transport module on multicast mode parallel transmission to the three data memory node;
Transport module receives the consistency that sends on the data memory node of place after the log recording interpolation request and guarantees module; Guarantee that to the consistency of other two data memory nodes module sends institute's log recording that receives and adds log recording maximum sequence number in asking, should equal 2; Receive simultaneously the maximum sequence number of log recording that other two data memory nodes send, and the number of the identical log recording that the receives maximum sequence number of the maximum sequence number of log recording that sends with self of record, when number equals 2, then the log recording that receives is added the record interpolation module that request sends to the place data memory node;
Record adds module to be responsible for that log recording is added three log recordings in the request and to add in the daily record on the data memory node of place.
As shown in Figure 5, data memory node carries out log recording and adds and recovery, and process is as follows:
1. the log request module produces corresponding daily record interpolation request after receiving file-system command;
2. add module and receive and attempt carrying out daily record after the request and add operation, need judge whether the daily record remaining space satisfies the total length that described log recording adds each bar log recording in the request, if satisfy then jumped to for the 3rd step, go on foot otherwise jump to the 4th;
3. the daily record remaining space satisfies, and coordinative coherence guarantees that module and record add the interpolation operation that module is carried out log recording, adds to jump to for the 7th step after operation is finished;
4. the daily record remaining space does not satisfy, and sends to the log request module and suspends request, and the generation log recording of requesting timeout adds request, and the transport module to the place data memory node sends record recovery request simultaneously;
5. transport module receives log recording and reclaims the rear Notification Record recycling module of request, and beginning log recording reclaimer operation begins to reclaim log recording from the daily record original position;
6. if log recording reclaims successfully then reclaims end signal to the transmission of log request module, the interpolation of recovery generation log recording was asked and is jumped to the 1st step and restarts, and went on foot otherwise continue the 7th;
7. end of operation.
As shown in Figure 6, it is that 0 to 2 log recording recovers that present embodiment needs LSN, specifically describes as follows:
The log request module forms recovery request and sends to distribution module; This recovery request is comprised of request type and three log recording sequence numbers, and request type is recovery, and three log recording sequence numbers are respectively 0,1,2;
Distribution module has been set up the Internet Transmission passage with three data memory nodes; Receive after the recovery request it by the transport module on multicast mode parallel transmission to the three data memory node;
Transport module receives the recovery module that sends to after the recovery request on the data memory node of place, recovers module and according to the log recording sequence number in the recovery request log recording of corresponding sequence number is transferred to the log request module; Wherein first data memory node transmission log sequence number is 0 and 1 log recording, and second data memory node transmission log sequence number is 1 and 2 log recording, and the 3rd data memory node transmission log sequence number is 0 and 2 log recording;
Now the log request module the complete LSN that receives be 0 to 2 log recording, transfer them to file system, file system can utilize these log recordings to recover.
Claims (1)
1. the collaboration type log system of a Based on Distributed file system comprises that log request module, distribution module, a N transport module, a N consistency guarantee that module, a N record add module, a N record recycling module and N recovery module, N 〉=2; It is characterized in that:
(1) log request module runs on metadata management node or data memory node,
When receiving the retouching operation order of file system, judging that record adds module and whether sends the time-out request, is the recovery end signal that then waits recycling module to be recorded to send, and receives this recovery end signal, form again log recording and add request, add request otherwise directly form log recording; Then log recording is added request and send to distribution module; Log recording adds request and is comprised of request type and many log recordings, and request type is for adding record, and each bar log recording all has sequence number;
When receiving the recovery order of file system, form recovery request, recovery request sent to distribution module and wait for, receive recover each bar log recording that module transmits after, transfer them to file system; Recovery request is comprised of request type and a plurality of log recording sequence number, and request type is for recovering;
(2) distribution module runs on metadata management node or data memory node, sends connection request to each transport module respectively, sets up the Internet Transmission passage; After receiving log recording interpolation request or recovery request, it is arrived each transport module by the multicast mode parallel transmission; Connection request is comprised of request type, public network address, destination network addresses and port numbers, request type is for connecting, public network address is that IP address, the destination network addresses of distribution module place node is the IP address of transport module place node, and port numbers is the portal number that the Internet Transmission passage takies;
(3) N transport module is identical, runs on respectively each data memory node, receive connection request after, set up the Internet Transmission passage; After receiving log recording interpolation request, the consistency that sends it to the place data memory node guarantees module; After receiving recovery request, send it to the recovery module of place data memory node; After receiving record recovery request, send it to and respectively record recycling module;
(4) N consistency guarantees that module is identical, run on respectively each data memory node, guarantee that to the consistency of other each data memory node module sends institute's log recording that receives and adds log recording maximum sequence number in asking, the consistency that receives simultaneously other each data memory node guarantees the maximum sequence number of log recording that module sends, and the number of the identical log recording that the receives maximum sequence number of the maximum sequence number of log recording that sends with self of record, judge whether this number surpasses half of data memory node sum, be then the log recording that receives to be added the record interpolation module that request sends to the place data memory node, otherwise continue to wait for that the next log recording of transport module adds request;
(5) N record interpolation module is identical, run on respectively each data memory node, judge whether the daily record remaining space satisfies the total length that described log recording adds each bar log recording in the request on the data memory node of place, is then each bar log recording to be added in the daily record; Otherwise the transport module to the place data memory node sends record recovery request, and sends the time-out request to the log request module; Record to reclaim and ask by request type and reclaim record count to form, request type is to reclaim, and the recovery record count is M times that log recording adds log recording number in the request, M 〉=2;
(6) N record recycling module is identical, run on respectively each data memory node, after receiving that record reclaims request, the daily record original position begins to reclaim the log recording that reclaims record count the record recovery request from the notebook data memory node, and recovery is finished backward log request module and sent the recovery end signal;
(7) N to recover module identical, run on respectively each data memory node, receive recovery request after, according to the log recording sequence number in the recovery request log recording of corresponding sequence number is transferred to the log request module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210521676.6A CN103051681B (en) | 2012-12-06 | 2012-12-06 | Collaborative type log system facing to distribution-type file system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210521676.6A CN103051681B (en) | 2012-12-06 | 2012-12-06 | Collaborative type log system facing to distribution-type file system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103051681A true CN103051681A (en) | 2013-04-17 |
| CN103051681B CN103051681B (en) | 2015-06-17 |
Family
ID=48064175
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210521676.6A Active CN103051681B (en) | 2012-12-06 | 2012-12-06 | Collaborative type log system facing to distribution-type file system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103051681B (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103235754A (en) * | 2013-04-24 | 2013-08-07 | 曙光信息产业(北京)有限公司 | Method and device for processing requests in distributed file system |
| CN103336785A (en) * | 2013-06-04 | 2013-10-02 | 华中科技大学 | Distributed storage method and distributed storage device based on network coding |
| CN104394023A (en) * | 2014-12-09 | 2015-03-04 | 上海斐讯数据通信技术有限公司 | Failure collecting method and system for network access terminal |
| CN104932986A (en) * | 2014-03-19 | 2015-09-23 | 中国移动通信集团公司 | Data redistribution method and apparatus |
| CN105487963A (en) * | 2015-11-26 | 2016-04-13 | 小米科技有限责任公司 | Message confirmation method and apparatus |
| CN105589887A (en) * | 2014-10-24 | 2016-05-18 | 中兴通讯股份有限公司 | Data processing method for distributed file system and distributed file system |
| CN106302577A (en) * | 2015-05-19 | 2017-01-04 | 广州四三九九信息科技有限公司 | Log processing method |
| CN106776130A (en) * | 2016-11-30 | 2017-05-31 | 华为技术有限公司 | A kind of journal recovery method, storage device and memory node |
| CN108108476A (en) * | 2018-01-03 | 2018-06-01 | 中科边缘智慧信息科技（苏州）有限公司 | The method of work of highly reliable distributed information log system |
| CN110019105A (en) * | 2019-01-24 | 2019-07-16 | 江苏中云科技有限公司 | A kind of reliable efficient distributed file system |
| CN111046009A (en) * | 2019-11-11 | 2020-04-21 | 杭州迪普科技股份有限公司 | Log storage method and device |
| CN115344539A (en) * | 2022-10-12 | 2022-11-15 | 北京奥星贝斯科技有限公司 | Log space recovery method and device for distributed database |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101183377A (en) * | 2007-12-10 | 2008-05-21 | 华中科技大学 | High availability data-base cluster based on message middleware |
| CN102024016A (en) * | 2010-11-04 | 2011-04-20 | 天津曙光计算机产业有限公司 | Rapid data restoration method for distributed file system (DFS) |
| US20120011398A1 (en) * | 2010-04-12 | 2012-01-12 | Eckhardt Andrew D | Failure recovery using consensus replication in a distributed flash memory system |
| CN102662795A (en) * | 2012-03-20 | 2012-09-12 | 浪潮电子信息产业股份有限公司 | Metadata fault-tolerant recovery method in distributed storage system |
| CN102664923A (en) * | 2012-03-30 | 2012-09-12 | 浪潮电子信息产业股份有限公司 | Method for realizing shared storage pool by utilizing Linux global file system |
-
2012
- 2012-12-06 CN CN201210521676.6A patent/CN103051681B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101183377A (en) * | 2007-12-10 | 2008-05-21 | 华中科技大学 | High availability data-base cluster based on message middleware |
| US20120011398A1 (en) * | 2010-04-12 | 2012-01-12 | Eckhardt Andrew D | Failure recovery using consensus replication in a distributed flash memory system |
| CN102024016A (en) * | 2010-11-04 | 2011-04-20 | 天津曙光计算机产业有限公司 | Rapid data restoration method for distributed file system (DFS) |
| CN102662795A (en) * | 2012-03-20 | 2012-09-12 | 浪潮电子信息产业股份有限公司 | Metadata fault-tolerant recovery method in distributed storage system |
| CN102664923A (en) * | 2012-03-30 | 2012-09-12 | 浪潮电子信息产业股份有限公司 | Method for realizing shared storage pool by utilizing Linux global file system |
Non-Patent Citations (1)
| Title |
|---|
| 熊劲等: "DCFS2的元数据一致性策略", 《计算机研究与发展》 * |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103235754A (en) * | 2013-04-24 | 2013-08-07 | 曙光信息产业(北京)有限公司 | Method and device for processing requests in distributed file system |
| CN103235754B (en) * | 2013-04-24 | 2016-10-05 | 曙光信息产业(北京)有限公司 | The treating method and apparatus of request in distributed file system |
| CN103336785A (en) * | 2013-06-04 | 2013-10-02 | 华中科技大学 | Distributed storage method and distributed storage device based on network coding |
| CN103336785B (en) * | 2013-06-04 | 2016-12-28 | 华中科技大学 | A kind of distributed storage method based on network code and device thereof |
| CN104932986A (en) * | 2014-03-19 | 2015-09-23 | 中国移动通信集团公司 | Data redistribution method and apparatus |
| CN105589887A (en) * | 2014-10-24 | 2016-05-18 | 中兴通讯股份有限公司 | Data processing method for distributed file system and distributed file system |
| CN104394023A (en) * | 2014-12-09 | 2015-03-04 | 上海斐讯数据通信技术有限公司 | Failure collecting method and system for network access terminal |
| CN104394023B (en) * | 2014-12-09 | 2018-05-01 | 上海斐讯数据通信技术有限公司 | Failure collection method and system for network insertion terminal |
| CN106302577B (en) * | 2015-05-19 | 2019-07-09 | 广州四三九九信息科技有限公司 | Log processing method |
| CN106302577A (en) * | 2015-05-19 | 2017-01-04 | 广州四三九九信息科技有限公司 | Log processing method |
| CN105487963B (en) * | 2015-11-26 | 2018-04-06 | 小米科技有限责任公司 | Message authentication method and device |
| CN105487963A (en) * | 2015-11-26 | 2016-04-13 | 小米科技有限责任公司 | Message confirmation method and apparatus |
| CN106776130A (en) * | 2016-11-30 | 2017-05-31 | 华为技术有限公司 | A kind of journal recovery method, storage device and memory node |
| WO2018098972A1 (en) * | 2016-11-30 | 2018-06-07 | 华为技术有限公司 | Log recovery method, storage device and storage node |
| CN108108476A (en) * | 2018-01-03 | 2018-06-01 | 中科边缘智慧信息科技（苏州）有限公司 | The method of work of highly reliable distributed information log system |
| CN110019105A (en) * | 2019-01-24 | 2019-07-16 | 江苏中云科技有限公司 | A kind of reliable efficient distributed file system |
| CN111046009A (en) * | 2019-11-11 | 2020-04-21 | 杭州迪普科技股份有限公司 | Log storage method and device |
| CN111046009B (en) * | 2019-11-11 | 2023-11-21 | 杭州迪普科技股份有限公司 | Log storage method and device |
| CN115344539A (en) * | 2022-10-12 | 2022-11-15 | 北京奥星贝斯科技有限公司 | Log space recovery method and device for distributed database |
| CN115344539B (en) * | 2022-10-12 | 2023-02-17 | 北京奥星贝斯科技有限公司 | Log space recovery method and device for distributed database |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103051681B (en) | 2015-06-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103051681B (en) | Collaborative type log system facing to distribution-type file system | |
| CN105335251B (en) | A kind of fault recovery method and system | |
| CN103226502B (en) | A kind of data calamity is for control system and data reconstruction method | |
| CN101997823B (en) | Distributed file system and data access method thereof | |
| CN102521072B (en) | Virtual tape library equipment and data recovery method | |
| CN102088490B (en) | Data storage method, device and system | |
| CN113641511B (en) | Message communication method and device | |
| CN109376197B (en) | Data synchronization method, server and computer storage medium | |
| KR102016095B1 (en) | System and method for persisting transaction records in a transactional middleware machine environment | |
| CN103929500A (en) | Method for data fragmentation of distributed storage system | |
| CN102142032B (en) | Method and system for reading and writing data of distributed file system | |
| CN102394914A (en) | Cluster brain-split processing method and device | |
| CN103902405B (en) | Quasi-continuity data replication method and device | |
| CN102609479A (en) | Memory database node copying method | |
| CN105518641A (en) | Method, device, and system for peer-to-peer data replication and method, device, and system for master node switching | |
| CN104077199A (en) | Shared disk based high availability cluster isolation method and system | |
| CN105721582A (en) | Multi-node file backup system | |
| CN107135097A (en) | The disaster tolerance system and disaster recovery method filed based on bookkeeping | |
| CN103516734A (en) | Data processing method, device and system | |
| CN102984009A (en) | Disaster recovery backup method for VoIP (Voice overInternet Protocol) system based on P2P | |
| CN110278222A (en) | The method, system and relevant device of data management in distributed file storage system | |
| CN101459544A (en) | Method and system for collecting performance data | |
| CN116010111A (en) | Cross-cluster resource scheduling method, system and terminal equipment | |
| CN114564458B (en) | Method, device, equipment and storage medium for synchronizing data among clusters | |
| CN103780433B (en) | Self-healing type virtual resource configuration management data architecture |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |