CN104767643A - Disaster recovery backup system based on virtual machine - Google Patents
Disaster recovery backup system based on virtual machine Download PDFInfo
- Publication number
- CN104767643A CN104767643A CN201510167677.9A CN201510167677A CN104767643A CN 104767643 A CN104767643 A CN 104767643A CN 201510167677 A CN201510167677 A CN 201510167677A CN 104767643 A CN104767643 A CN 104767643A
- Authority
- CN
- China
- Prior art keywords
- module
- virtual machine
- migration
- copy
- backup
- 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.)
- Pending
Links
Abstract
The invention relates to a disaster recovery backup system based on a virtual machine. The system comprises a deploying management module, a monitoring management module, a disaster recovery management module, a virtual machine deploying module, a virtual machine migration module and a virtual machine backup module. The deploying management module, the monitoring management module, the disaster recovery management module and the virtual machine migration module are connected in sequence. The virtual machine migration module and the virtual machine backup module are connected with a virtual resource pool respectively. The virtual resource pool is connected with the virtual machine deploying module. The virtual machine backup module comprises a copy server submodule, a copy engine submodule, a heartbeat detecting submodule and an external equipment buffering submodule. The copy server submodule and the copy engine submodule are used for transmitting a control system process. The external equipment buffering submodule comprises network buffering and peripheral disk storing buffering. Compared with the prior art, the fact that data are not lost, and applications are ceaseless is guaranteed, a user can have data which can be recovered at a crucial time, and customer production applications are ceaseless.
Description
[technical field]
The present invention relates to virtual disaster tolerance technology field, specifically a kind of disaster tolerant backup system based on virtual machine.
[background technology]
At present, according to the level implementing disaster tolerance, disaster tolerance system can be divided into data disaster tolerance, application disaster tolerance and business disaster tolerance.Data disaster tolerance and application disaster tolerance are owing to studying starting comparatively early, and existing utility system can be used.So-called business disaster tolerance is the thinner application of granularity by service decomposition, when a disaster occurs can by recombinating to business thus ensureing the continuity of business.By the reduction causing application to physical computing resources demand after service decomposition, and the performance of single computing system hardware is more and more higher simultaneously, wastes computational resource.Correlative study according to HP shows: the average utilization of server only has about 40%, and a considerable amount of server only runs the application program of small B/S or C/S structure.Even so, but traditional disaster tolerance system still needs for these servers configure corresponding backup server, thus causes the scale of disaster tolerance system excessive, high cost.
For different users, only need guarantee two point, first, whether can ensure that data are not lost, secondly, whether can ensure that application is uninterrupted.Backup is only used to ensure data a kind of technological means of not losing, and object to allow user can have at crucial moment the data that can recover and make the production application of client uninterrupted.But it is constant that the production environment of client is not one deck, increase by a root memory bar sometimes, although a hard disk has no impact to data Replica, when doing data and moving back, the failure that data may be caused to move back because of the difference of bottom hardware framework.In fact; now that what all to pursue than the data Replica means based on SAN of higher-end or fairly simple file copy is degree of protection to data security; if thus have ignored the possibility guaranteeing that data are moved back, even and if data that some cannot be moved back are lost the production environment of user also cannot be made to reenter serve level normally less again.
On the other hand, along with the development of intelligence science and technology of future generation, Intel Virtualization Technology more and more adopt by enterprise, but a lot of enterprise is not large-scale to be adopted.Or attempting the virtualization product of different brands, but, be difficult between the virtualization product of different brands carry out mutual migration, be also difficult to move back on physical platform, also bring very large difficulty to backup.So the mixing platform how between managing physical platform and different virtual product has also been trend of the times.
[summary of the invention]
Object of the present invention is exactly will solve above-mentioned deficiency and provide a kind of disaster tolerant backup system based on virtual machine, not only can ensure that data are not lost and apply uninterrupted, and achieve user and can have at crucial moment the data that can recover and make the production application of client uninterrupted.
Design a kind of disaster tolerant backup system based on virtual machine for achieving the above object, comprise deployment administration module, monitoring management module, disaster tolerance administration module, deploying virtual machine module, virtual machine (vm) migration module and virtual machine backup module, described deployment administration module, monitoring management module, disaster tolerance administration module, virtual machine (vm) migration module connects successively, described virtual machine (vm) migration module, virtual machine backup module connects virtual resource pond respectively, described virtual resource pond is connected with deploying virtual machine module, described virtual machine backup module comprises replication server, replication engine, heartbeat detection and external equipment cushion four submodules, described heartbeat detection submodule is used for backup node and carries out heartbeat detection to multiple live-vertex simultaneously safeguarding live-vertex list, described replication server, replication engine submodule is used for transmission control system process, described external equipment buffer sublayer module comprises meshwork buffering and external disk stores buffering.
The progress of work of described meshwork buffering comprises the steps:
(1) snapshot of virtual machine in generating run, that copies all states changes in buffering;
(2) change of state is kept in buffering, and virtual machine continues to perform;
(3) state be buffered is sent in backup host;
(4) when state receives, notice enlivens main frame;
(5) last, the network of buffer memory exports and is released.
Described virtual machine backup module also comprises multiple virtual machine copy, and described multiple virtual machine copy is stored in multiple backup node.
Described virtual machine (vm) migration module is by monitoring transferring module, running transferring module, pre-copy module, freezing module and active module forms, described monitoring transferring module, run transferring module, active module and connect successively, described pre-copy module, freeze module and connect operation transferring module respectively.
Described monitoring transferring module for starting migration, selecting destination host and predetermined resource, and sends migration signal.
Described operation transferring module is used for after monitoring transferring module sends a migration signal, collect the system status information of Source client, this system status information comprises the communications status of the state of VCPU, the time state of system, the state of register and I/O equipment, and the state collected is encapsulated, then send freeze signal to freezing module, Source client is freezed, copy page, after copy terminates, send activation signal to Destination client, activate Destination client.
Described module of freezing is for ensureing to move the consistency of rear Source client and Destination client.
Described active module is used for freezing after module freezes Source client, run module and continue the remaining page of migration, until all pages all copy destination host to, run transferring module and can send an activation signal to active module, start active module, then active module activates the migration client on destination host.
Described pre-copy module is for recording the amendment situation of page in a period of time, and then iteration transmits the not high page of Modification Frequency.
Described pre-copy module comprises HIT internal memory migration module and dirty page slowdown module, the page that described HIT internal memory migration module is used for the needs of locating after each iteration transmit does write record, this record is utilized to judge the change of current dirty page rate, to call the dirty page decelerator in dirty page slowdown module.
Compared with the existing technology, tool has the following advantages in the present invention:
(1) heterogeneous platform of efficient managing physical and different virtual product composition: based on hardware platform and irrelevant (operating system, application and the data) transportable technology of virtual machine platform; the physics of the hybird environments such as all isomery virtual platforms and isomerization hardware physical services applicator platform such as seamless supported data center VMWare, Hyper-V, VMWARE and in the future cloud computing platform and virtual work load intelligent management, can realize virtualization integration, migration, calamity for Conservation and instauration and intelligent scheduling management etc.
(2) integrated disaster tolerance is achieved: use very easy; Unpacking plug and play, installs client software without the need to manual on source and target server.Use WEB administration interface, provide a ready-to-use instrument board, the state of protection plan can be checked at any time.Easily can manage, monitor and report the various aspects of production environment.When this comprises the situation of production development server outage or disaster, automatically send e-mail alert etc.Can manage whenever and wherever possible, investment and O&M cost low.For whole server workload (comprising data, application program and operating system) provides protection.Contrary with data-centered restoration methods, the data that general protection live load function can be recovery provide an environment.The interference because manual reconstructing system, decentralized system and recovery data cause can be avoided.By this system, in economical and practical mode, can protect or remote protecting for the physical working environment in wide area network (WAN) and virtual work environment provide local.And if production development server interrupts or the situation of disaster, only need a mouse click, can recover shielded production application without any confusion, TR equipment can adapter production application, makes it while the environment that resumes production, still can continue normal operation.After production environment recovers on line state, the production application for storing is provided option flexibly by TR.If original production server is through repairing, and hardware is unaffected, then by performing the virtual to physics (V2P) operating load transmitting function, thus production application can be retracted primary platform from virtual recovery environment.If original hardware cannot be repaired, then V2P transmitting function can be used production application to be returned in new hardware.Also can like a cork production application be migrated in virtual production environment.Also mean that you need not worry that whether new hardware be same brand, model or configuration with original hardware flexibly with the memory function of hardware independent.Also need not consider that whether new virtual environment is identical with former virtual environment brand.
(3) realize the standby rehearsal of calamity simply: a key restore funcitons of Help of System, easily can test rapidly and copy and the integrality of recovery plan.No matter be in physics and virtual environment, if perform Failure Transfer test, system will arrange production environment restoration point.Like this, can fast verification recovery plan and related service service in emulation test system in the unbroken situation of production environment.DRP is after checking, and system can abandon all changes of resuming work in load immediately, and recovers normal operating load copy operation.
[accompanying drawing explanation]
Fig. 1 is structural principle block diagram of the present invention;
Fig. 2 is the block schematic illustration of virtual machine (vm) migration module in the present invention;
[embodiment]
Below in conjunction with accompanying drawing, following further illustrating is done to the present invention:
As shown in drawings, the present invention includes: dispose administration module, monitoring management module, disaster tolerance administration module, deploying virtual machine module, virtual machine (vm) migration module and virtual machine backup module, dispose administration module, monitoring management module, disaster tolerance administration module, virtual machine (vm) migration module connects successively, virtual machine (vm) migration module, virtual machine backup module connects virtual resource pond respectively, virtual resource pond is connected with deploying virtual machine module, virtual machine backup module comprises replication server, replication engine, heartbeat detection and external equipment cushion four submodules, heartbeat detection submodule is used for backup node and carries out heartbeat detection to multiple live-vertex simultaneously safeguarding live-vertex list, replication server, replication engine submodule is used for transmission control system process, external equipment buffer sublayer module comprises meshwork buffering and external disk stores buffering.
In the present invention, virtual machine backup module also comprises multiple virtual machine copy, and multiple virtual machine copy is stored in multiple backup node.Virtual machine (vm) migration module is by monitoring transferring module, running transferring module, pre-copy module, freezing module and active module forms, monitor transferring module, run transferring module, active module and connect successively, pre-copy module, freeze module and connect operation transferring module respectively.Wherein, monitor transferring module for starting migration, selecting destination host and predetermined resource, and send migration signal.Running transferring module is used for after monitoring transferring module sends a migration signal, collect the system status information of Source client, this system status information comprises the communications status of the state of VCPU, the time state of system, the state of register and I/O equipment, and the state collected is encapsulated, then send freeze signal to freezing module, Source client is freezed, copy page, after copy terminates, send activation signal to Destination client, activate Destination client.Freeze module for ensureing to move the consistency of rear Source client and Destination client.Active module is used for freezing after module freezes Source client, run module and continue the remaining page of migration, until all pages all copy destination host to, run transferring module and can send an activation signal to active module, start active module, then active module activates the migration client on destination host.Pre-copy module is for recording the amendment situation of page in a period of time, and then iteration transmits the not high page of Modification Frequency.This pre-copy module comprises HIT internal memory migration module and dirty page slowdown module, the page that HIT internal memory migration module is used for the needs of locating after each iteration transmit does write record, this record is utilized to judge the change of current dirty page rate, to call the dirty page decelerator in dirty page slowdown module.
Traditionally in order to the isolation between guarantee business, if adopt the mode of traditional business of directly disposing on the computer systems, then a station server preferably only disposes a business, and by multiple service deployment on a station server, multiple business must interact.Virtual machine of the present invention to reasonable distribution again after Resource Abstract, can provide packaging effects, isolation, hardware independence simultaneously.But by service deployment on virtual platform, solve the wasting of resources and the interactive problem of multi-service, the performance of server is not fully exerted, when same physical device, effectively can improve the capability to resist destruction of whole disaster tolerance system.In such a case, business backup with recover problem transforms in order to virtual machine disaster-tolerant backup with recover problem.
The present invention utilizes Intel Virtualization Technology, multiple physical resource cluster can be abstracted into the virtual resource with hardware independent, virtual machine operates on the virtual resource that distributed by VMM, the operating system of VM detection of dynamic virtual machine internal and the running status of application program, if operating system or application crash, then restart virtual machine at corresponding physical node, thus realize the fast restoration capabilities [24] of software faults.Meanwhile, the availability of VM meeting detection of dynamic physical node, if certain physical node lost efficacy, then restarted the virtual machine of this physical nodes all impact, thus realized the fast restoration capabilities of hardware error on physical node available in addition.
This system by protected service encapsulates in a virtual machine.System realizes based on the VMM of Vmware, extends the dynamic migration mechanism of Vmware, to support fine-grained checkpoint.Virtual machine image in backup host resides in internal memory, can run immediately when enlivening hostdown or being ruined.Because backup virtual machine can only ensure and enliven the periodic consistency of virtual machine, must be buffered until state is synchronized to backup virtual machine so network exports.When one complete, after consistent mirror image is received, meshwork buffering is released.Checkpoint, buffering and release cycle are with very high frequency run.Distribution part system cushions four parts by replication server, replication engine, heartbeat detection and external equipment and forms, and wherein, the heart beat detection module of backup node carries out heartbeat detection to N number of live-vertex simultaneously and safeguards live-vertex list; Replication server and replication engine are transmission control processes, and it is used for the calculating of the transmission of various message in control system, the updating maintenance of each list and hashed value; External equipment buffering is divided into two parts meshwork buffering and disk storage buffering.
Because most of network all can not provide reliable data transport service, transparent to ensure to back up in realtime to user, web application must process packet loss, repetition and the problem such as out of order, or uses upper layer network agreement as TCP, to support to serve more reliably.Native system uses meshwork buffering to coordinate upper layer network agreement to simplify meshwork buffering problem: the bag of loss does not need to retransmit; because packet loss is counted as transparent network failure; do not affect the correctness of protected virtual machine state; before the state produced in the checkpoint cycle is simultaneously submitted to backup host, queue to be transmitted will get clogged.
In VMWARE, the network interface of Domain U is driven by the rear end in the front-end driven of Domain U and corresponding Domain 0 and forms.It is exactly network input that the network of Domain U exports for Domain 0, so add meshwork buffering middleware, to support the network I/O of Domain U in we Hypercall between Domain U and Domain 0.Network data is copy transmissions between Domain U and Domain 0 not, but shared by internal memory, but the same time only have a small amount of page to be shared, if transmit very important message between Domain U and Domain 0, just need to wait for the long time, this is intolerable sometimes.So meshwork buffering middleware directly copies these message to local internal memory, then release is mapped to local shared data.
The invention describes meshwork buffering workflow, network input is directly transferred to protected virtual machine, but the network produced after previous checkpoint exports and gets clogged until current check point is backed up main frame receives.Meshwork buffering is applied to the network equipment of client computer with Linux queue mechanism, returns two RT-netlink message.Because client computer can continue to perform after checkpoint, when meshwork buffering receives a checkpoint message, network output queue before receiving release message is caused to get clogged.Backup host receives checkpoint, and then meshwork buffering receives release message, and at this moment the packet of network output queue exports.Be below the concise and to the point course of work of meshwork buffering: the snapshot of virtual machine in (1) generating run, that copies all states changes in buffering;
(2) change of state is kept in buffering, and virtual machine continues to perform; (3) state be buffered is sent in backup host; (4) when state receives, notice enlivens main frame; (5) last, the network of buffer memory exports and is released.Wherein, some is different in the process of exterior storage state and network state, and the change of exterior storage state is divided into two kinds of situations: read and write.When needs are from exterior storage read data, require no special processing.And when needs externally store write data time, handling process with process network flow process be the same.
In the present invention, virtual machine system of backing up in realtime realizes based on VMM, extends the dynamic migration mechanism of Vmware, to support fine-grained checkpoint.Virtual machine copy in backup host resides in internal memory, can run immediately when enlivening hostdown or being ruined.Because backup virtual machine can only ensure and enliven the periodic consistency of virtual machine, must be buffered until state is synchronized to backup virtual machine so network exports.After complete, a consistent virtual machine copy is received, meshwork buffering is released.Checkpoint, buffering and release cycle, with very high frequency run, ensure that whole process is transparent concerning user.
Fast and effeciently can recover during in order to ensure that operation system is destroyed, the virtual machine copy of live-vertex must be preserved on multiple backup node.Multiple copy is kept in multiple node, effectively can improve and be ruined node and at least obtain the chance of an effective copy, improve the survivability of system.But the increase of copy also brings new problem: the consistency how keeping multiple copy when virtual machine upgrades; How to select rapidly an optimum backup node to become live-vertex when having an accident.These all can bring extra expense to system, certainly will have influence on the performance of live-vertex.For the problems referred to above, propose a kind of virtual machine copy herein and safeguard and selection strategy, and carried out performance test and analysis in prototype system.
Wherein, virtual machine copy maintenance strategy is: when (1) each backup node is to live-vertex heartbeat detection, transmitting time is packed, sends to the live-vertex with token transmission route; (2) network load condition of each backup node of live-vertex record; (3) live-vertex provides token to the minimum node of offered load; (4) backup node holding token receives the incremental checkpoint of live-vertex transmission; (5) all the other backup nodes receive the incremental checkpoint of the backup node transmission holding token.Virtual machine copy selection strategy is: when each backup node heart beat detection module finds that certain live-vertex is ruined or destroyed, (1) token is held on backup node and is initiated once " election ", respectively can reach the incremental checkpoint hashed value of replica node final updating; (2) when the hashed value of all nodes is all identical, then hold node by current token and recover; (3) when the different hashed value of appearance two and affiliated node number does not wait time, select credible, an optimum node with majority rule; (4) when two different hashed values and affiliated node number is equal, then hold node by current token and recover; (5) then think that group of alliance suffers large-scale destruction if there is more than two different hashed values, must request management person artificially get involved; (6) backup node becomes live-vertex and automatically obtains token granting power.
Before planization technology maturation, migration calculation task can not be disposed widely because of its higher complexity.The advantage of virtual machine technique be just its by operating system and on application program be abstracted into one or several file, thus real hardware resource and software service to be separated.Therefore, the Transport And Transformation of calculation task is in order to the migration of virtual machine.The dynamic migration technology of virtual machine can significantly improve the service quality of system, effectively can solve the business continuance of operation system.The wide area network used according to disaster tolerance system herein postpones larger feature, proposes and optimizes the dynamic migration of virtual machine algorithm towards disaster tolerance, and having carried out system performance testing and analysis.
Vmware dynamic migration framework is by monitoring transferring module, running transferring module, pre-copy module, freezing module and active module forms.Monitor transferring module: the major function of this module is the initial work of migration, comprise and start migration, select destination host and predetermined resource, send migration signal.Run transferring module: the major part work being responsible for migration, after monitoring module sends a migration signal, collect the state information of Source client, the system status information collected, comprising: the state of VCPU, the time state of system, the in addition state of each register, the communications status etc. of I/O equipment, encapsulate the OS state collected, then sending " freeze signal " to freezing module, Source client is freezed, copy page.After copy terminates, send " activation signal " to Destination client, activate Destination client.Freeze module: this module is to solve migration client computer and Destination client internal memory stationary problem when moving.Namely in order to ensure the consistency of moving rear source host and destination host.Active module: freezing after module freezes Source client, run module and continue the remaining page of migration, until all pages all copy destination host to, now, run transferring module and can send an activation signal to active module, start active module, then active module activates the migration client on destination host.Pre-copy module: the main task of this module is the amendment situation of page in record a period of time, the page that then iteration transmission Modification Frequency is not high, has reached the object reducing downtime in transition process.
On the basis that have studied Vmware dynamic migration framework, for the principal element affecting dynamic migration performance, DRO (Disaster Recovery Oriented) the dynamic migration framework of this problem is proposed.DRO dynamic migration framework realizes based on the VMM of Vmware, and the basis of Vmware dynamic migration framework with the addition of HIT (Hierarchical Iterative Transmit) internal memory migration module and dirty page decelerator.HIT internal memory migration module is by the original working set assay method of amendment Vmware, improves the efficiency of page copy.Dirty page decelerator mainly reduces by appropriate the cpu resource distributing to migration client computer, thus reduce the throughput rate of dirty page.To cooperatively interact work to realize dirty page decelerator and HIT internal memory migration module, the page needing the needs of locating after to each iteration to transmit in HIT internal memory migration module writes relevant record, and these records can be utilized like this to judge the change of current dirty page rate.When finding that the number of pages of iterative migration is greater than previous round, just should call dirty page decelerator, reaching by the activity reducing client computer the object reducing dirty page rate.Reduction due to dirty page rate can reduce the workload freezing stage page copy, thus reduces downtime.
When using shared storage technique, the center of gravity of dynamic migration of virtual machine is internal memory migration.The internal memory migration principle of Vmware is: the method first adopting pre-copy, page is iteratively copied on destination host, the first round copies all pages, second takes turns the page that only copy revised in first round iterative process, by that analogy, n-th what take turns copy is take turns (n-1)th the page revised in iterative process, the page revised again in former transmission becomes working set, when the number of dirty page arrives certain constant or iterations arrival some, the pre-copy stage terminates, enter and shut down the copy stage, at this moment shut down and the information such as remaining dirty page and running status are all copied over.This method balances the contradiction between downtime and gross migration time preferably, is a kind of more satisfactory internal memory migration method.
The present invention, by the restriction of above-mentioned execution mode, change, the modification done, substitutes, combines, simplifies under other any does not deviate from Spirit Essence of the present invention and principle, all should be equivalent substitute mode, is included within protection scope of the present invention.
Claims (10)
1. the disaster tolerant backup system based on virtual machine, it is characterized in that: comprise deployment administration module, monitoring management module, disaster tolerance administration module, deploying virtual machine module, virtual machine (vm) migration module and virtual machine backup module, described deployment administration module, monitoring management module, disaster tolerance administration module, virtual machine (vm) migration module connects successively, described virtual machine (vm) migration module, virtual machine backup module connects virtual resource pond respectively, described virtual resource pond is connected with deploying virtual machine module, described virtual machine backup module comprises replication server, replication engine, heartbeat detection and external equipment cushion four submodules, described heartbeat detection submodule is used for backup node and carries out heartbeat detection to multiple live-vertex simultaneously safeguarding live-vertex list, described replication server, replication engine submodule is used for transmission control system process, described external equipment buffer sublayer module comprises meshwork buffering and external disk stores buffering.
2., as claimed in claim 1 based on the disaster tolerant backup system of virtual machine, it is characterized in that, the progress of work of described meshwork buffering comprises the steps:
(1) snapshot of virtual machine in generating run, that copies all states changes in buffering;
(2) change of state is kept in buffering, and virtual machine continues to perform;
(3) state be buffered is sent in backup host;
(4) when state receives, notice enlivens main frame;
(5) last, the network of buffer memory exports and is released.
3., as claimed in claim 1 or 2 based on the disaster tolerant backup system of virtual machine, it is characterized in that: described virtual machine backup module also comprises multiple virtual machine copy, described multiple virtual machine copy is stored in multiple backup node.
4. as claimed in claim 3 based on the disaster tolerant backup system of virtual machine, it is characterized in that: described virtual machine (vm) migration module is by monitoring transferring module, running transferring module, pre-copy module, freezing module and active module forms, described monitoring transferring module, run transferring module, active module and connect successively, described pre-copy module, freeze module and connect operation transferring module respectively.
5. as claimed in claim 4 based on the disaster tolerant backup system of virtual machine, it is characterized in that: described monitoring transferring module for starting migration, selecting destination host and predetermined resource, and sends migration signal.
6. as claimed in claim 4 based on the disaster tolerant backup system of virtual machine, it is characterized in that: described operation transferring module is used for after monitoring transferring module sends a migration signal, collect the system status information of Source client, this system status information comprises the state of VCPU, the time state of system, the state of register and the communications status of I/O equipment, and the state collected is encapsulated, then freeze signal is sent to freezing module, Source client is freezed, copy page, after copy terminates, activation signal is sent to Destination client, activate Destination client.
7., as claimed in claim 4 based on the disaster tolerant backup system of virtual machine, it is characterized in that: described in freeze module for ensureing to move the consistency of rear Source client and Destination client.
8. as claimed in claim 4 based on the disaster tolerant backup system of virtual machine, it is characterized in that: described active module is used for freezing after module freezes Source client, run module and continue the remaining page of migration, until all pages all copy destination host to, run transferring module and can send an activation signal to active module, start active module, then active module activates the migration client on destination host.
9. as claimed in claim 4 based on the disaster tolerant backup system of virtual machine, it is characterized in that: described pre-copy module is for recording the amendment situation of page in a period of time, and then iteration transmits the not high page of Modification Frequency.
10. as claimed in claim 4 based on the disaster tolerant backup system of virtual machine, it is characterized in that: described pre-copy module comprises HIT internal memory migration module and dirty page slowdown module, the page that described HIT internal memory migration module is used for the needs of locating after each iteration transmit does write record, this record is utilized to judge the change of current dirty page rate, to call the dirty page decelerator in dirty page slowdown module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510167677.9A CN104767643A (en) | 2015-04-09 | 2015-04-09 | Disaster recovery backup system based on virtual machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510167677.9A CN104767643A (en) | 2015-04-09 | 2015-04-09 | Disaster recovery backup system based on virtual machine |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104767643A true CN104767643A (en) | 2015-07-08 |
Family
ID=53649271
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510167677.9A Pending CN104767643A (en) | 2015-04-09 | 2015-04-09 | Disaster recovery backup system based on virtual machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104767643A (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105553629A (en) * | 2016-03-15 | 2016-05-04 | 山东超越数控电子有限公司 | Safe and credible calculation master and slave system |
| CN106201702A (en) * | 2016-07-19 | 2016-12-07 | 浪潮(北京)电子信息产业有限公司 | The method of virtual machine (vm) migration, system and cloud computing management platform |
| CN106790713A (en) * | 2017-03-08 | 2017-05-31 | 中国人民解放军国防科学技术大学 | Across data center virtual machine migration method under cloud computing environment |
| CN107135097A (en) * | 2017-04-14 | 2017-09-05 | 北京金融资产交易所有限公司 | The disaster tolerance system and disaster recovery method filed based on bookkeeping |
| CN107329806A (en) * | 2017-07-24 | 2017-11-07 | 郑州云海信息技术有限公司 | A kind of development environment construction method and device |
| CN107368349A (en) * | 2017-07-17 | 2017-11-21 | 郑州云海信息技术有限公司 | A kind of method for managing resource and device |
| CN107832118A (en) * | 2017-11-18 | 2018-03-23 | 浙江网新恒天软件有限公司 | A kind of KVM live migration of virtual machine optimization methods of reduction VCPU temperatures |
| CN108153491A (en) * | 2017-12-22 | 2018-06-12 | 深圳市瑞驰信息技术有限公司 | A kind of storage method and framework for closing part server |
| CN109240855A (en) * | 2018-09-11 | 2019-01-18 | 郑州云海信息技术有限公司 | A kind of method and device for realizing dual-active disaster tolerance for server virtualization system |
| CN109753340A (en) * | 2017-11-08 | 2019-05-14 | 阿里巴巴集团控股有限公司 | Virtual machine snapshot processing method, apparatus and system |
| CN110427246A (en) * | 2019-06-29 | 2019-11-08 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Multi-core virtual subregion reconfiguration system |
| CN110609764A (en) * | 2018-06-15 | 2019-12-24 | 伊姆西Ip控股有限责任公司 | Method, apparatus and computer program product for data backup |
| CN111831490A (en) * | 2020-06-30 | 2020-10-27 | 中国船舶重工集团公司第七一六研究所 | Method and system for synchronizing memories between redundant main and standby nodes |
| CN111858150A (en) * | 2019-04-26 | 2020-10-30 | Emc知识产权控股有限公司 | System and method for management of large-scale data backup |
| US10887382B2 (en) | 2018-12-18 | 2021-01-05 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud-based disaster recovery |
| US10958720B2 (en) | 2018-12-18 | 2021-03-23 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud based disaster recovery |
| US10983886B2 (en) | 2018-12-18 | 2021-04-20 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud-based disaster recovery |
| CN112910676A (en) * | 2019-12-04 | 2021-06-04 | 财团法人工业技术研究院 | Fault tolerant system and control method thereof |
| CN112954264A (en) * | 2019-12-10 | 2021-06-11 | 浙江宇视科技有限公司 | Platform backup protection method and device |
| US11178221B2 (en) | 2018-12-18 | 2021-11-16 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud-based disaster recovery |
| US11176002B2 (en) | 2018-12-18 | 2021-11-16 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud-based disaster recovery |
| US11252019B2 (en) | 2018-12-18 | 2022-02-15 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud-based disaster recovery |
| US11489730B2 (en) | 2018-12-18 | 2022-11-01 | Storage Engine, Inc. | Methods, apparatuses and systems for configuring a network environment for a server |
-
2015
- 2015-04-09 CN CN201510167677.9A patent/CN104767643A/en active Pending
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105553629A (en) * | 2016-03-15 | 2016-05-04 | 山东超越数控电子有限公司 | Safe and credible calculation master and slave system |
| CN106201702A (en) * | 2016-07-19 | 2016-12-07 | 浪潮(北京)电子信息产业有限公司 | The method of virtual machine (vm) migration, system and cloud computing management platform |
| CN106790713A (en) * | 2017-03-08 | 2017-05-31 | 中国人民解放军国防科学技术大学 | Across data center virtual machine migration method under cloud computing environment |
| CN106790713B (en) * | 2017-03-08 | 2019-09-20 | 中国人民解放军国防科学技术大学 | Across data center virtual machine migration method under cloud computing environment |
| CN107135097A (en) * | 2017-04-14 | 2017-09-05 | 北京金融资产交易所有限公司 | The disaster tolerance system and disaster recovery method filed based on bookkeeping |
| CN107368349A (en) * | 2017-07-17 | 2017-11-21 | 郑州云海信息技术有限公司 | A kind of method for managing resource and device |
| CN107368349B (en) * | 2017-07-17 | 2020-09-08 | 郑州云海信息技术有限公司 | Resource management method and device |
| CN107329806A (en) * | 2017-07-24 | 2017-11-07 | 郑州云海信息技术有限公司 | A kind of development environment construction method and device |
| CN109753340A (en) * | 2017-11-08 | 2019-05-14 | 阿里巴巴集团控股有限公司 | Virtual machine snapshot processing method, apparatus and system |
| CN107832118A (en) * | 2017-11-18 | 2018-03-23 | 浙江网新恒天软件有限公司 | A kind of KVM live migration of virtual machine optimization methods of reduction VCPU temperatures |
| CN108153491A (en) * | 2017-12-22 | 2018-06-12 | 深圳市瑞驰信息技术有限公司 | A kind of storage method and framework for closing part server |
| CN110609764B (en) * | 2018-06-15 | 2023-07-25 | 伊姆西Ip控股有限责任公司 | Method, apparatus and computer program product for data backup |
| CN110609764A (en) * | 2018-06-15 | 2019-12-24 | 伊姆西Ip控股有限责任公司 | Method, apparatus and computer program product for data backup |
| CN109240855A (en) * | 2018-09-11 | 2019-01-18 | 郑州云海信息技术有限公司 | A kind of method and device for realizing dual-active disaster tolerance for server virtualization system |
| US11178221B2 (en) | 2018-12-18 | 2021-11-16 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud-based disaster recovery |
| US11489730B2 (en) | 2018-12-18 | 2022-11-01 | Storage Engine, Inc. | Methods, apparatuses and systems for configuring a network environment for a server |
| US10887382B2 (en) | 2018-12-18 | 2021-01-05 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud-based disaster recovery |
| US10958720B2 (en) | 2018-12-18 | 2021-03-23 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud based disaster recovery |
| US10983886B2 (en) | 2018-12-18 | 2021-04-20 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud-based disaster recovery |
| US11252019B2 (en) | 2018-12-18 | 2022-02-15 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud-based disaster recovery |
| US11176002B2 (en) | 2018-12-18 | 2021-11-16 | Storage Engine, Inc. | Methods, apparatuses and systems for cloud-based disaster recovery |
| CN111858150A (en) * | 2019-04-26 | 2020-10-30 | Emc知识产权控股有限公司 | System and method for management of large-scale data backup |
| CN110427246B (en) * | 2019-06-29 | 2023-06-20 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Multi-core virtual partition reconstruction system |
| CN110427246A (en) * | 2019-06-29 | 2019-11-08 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Multi-core virtual subregion reconfiguration system |
| CN112910676A (en) * | 2019-12-04 | 2021-06-04 | 财团法人工业技术研究院 | Fault tolerant system and control method thereof |
| CN112954264A (en) * | 2019-12-10 | 2021-06-11 | 浙江宇视科技有限公司 | Platform backup protection method and device |
| CN112954264B (en) * | 2019-12-10 | 2023-04-18 | 浙江宇视科技有限公司 | Platform backup protection method and device |
| CN111831490A (en) * | 2020-06-30 | 2020-10-27 | 中国船舶重工集团公司第七一六研究所 | Method and system for synchronizing memories between redundant main and standby nodes |
| CN111831490B (en) * | 2020-06-30 | 2023-11-10 | 中国船舶集团有限公司第七一六研究所 | Method and system for synchronizing memories between redundant main and standby nodes |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104767643A (en) | Disaster recovery backup system based on virtual machine | |
| US11016857B2 (en) | Microcheckpointing with service processor | |
| US11669409B2 (en) | Application migration between environments | |
| US9323550B2 (en) | Mechanism for providing virtual machines for use by multiple users | |
| US9411628B2 (en) | Virtual machine cluster backup in a multi-node environment | |
| US6795966B1 (en) | Mechanism for restoring, porting, replicating and checkpointing computer systems using state extraction | |
| CA2839014C (en) | Managing replicated virtual storage at recovery sites | |
| US8689047B2 (en) | Virtual disk replication using log files | |
| US7523344B2 (en) | Method and apparatus for facilitating process migration | |
| US8214685B2 (en) | Recovering from a backup copy of data in a multi-site storage system | |
| CN105612503A (en) | Virtual retry queue | |
| AU2012273366A1 (en) | Managing replicated virtual storage at recovery sites | |
| Zhao et al. | Exploring reliability of exascale systems through simulations. | |
| WO2012109876A1 (en) | Virtualized data backup method, and virtualized data reorganization method, device and system | |
| US10824522B2 (en) | Method, apparatus, and computer program product for generating consistent snapshots without quiescing applications | |
| US11550491B2 (en) | Container-based application processing | |
| KR20140068046A (en) | System and method for persisting transaction records in a transactional middleware machine environment | |
| Ouyang et al. | RDMA-based job migration framework for MPI over infiniband | |
| US20220269414A1 (en) | Snapshotting a containerized application | |
| US20220350497A1 (en) | Managing replication journal in a distributed replication system | |
| Jung et al. | Design and Implementation of Multiple Fault-Tolerant MPI over Myrinet (M^ 3) | |
| Ni | Mitigation of failures in high performance computing via runtime techniques | |
| CN111459607A (en) | Virtual server cluster building method, system and medium based on cloud desktop virtualization | |
| Cerveira et al. | Fast local VM migration against hypervisor corruption | |
| US11972266B2 (en) | Hibernating and resuming nodes of a computing cluster |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150708 |
|
| WD01 | Invention patent application deemed withdrawn after publication |