CN102521063B - Shared storage method suitable for migration and fault tolerance of virtual machine - Google Patents
Shared storage method suitable for migration and fault tolerance of virtual machine Download PDFInfo
- Publication number
- CN102521063B CN102521063B CN 201110390549 CN201110390549A CN102521063B CN 102521063 B CN102521063 B CN 102521063B CN 201110390549 CN201110390549 CN 201110390549 CN 201110390549 A CN201110390549 A CN 201110390549A CN 102521063 B CN102521063 B CN 102521063B
- Authority
- CN
- China
- Prior art keywords
- virtual machine
- server
- data
- image file
- data storage
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The invention relates to the field of cloud calculation, in particular to a shared storage method suitable for migration and fault tolerance of a virtual machine. According to the invention, a data storage server and a metadata server are respectively loaded with a network driving module. The metadata server actively searches each data storage server on the network through a fixed multicast group method. Each data storage server actively adds into a network storage cluster after identifying the metadata server through key words information. A plurality of backups can be assigned for data in the data storage server according to configuration files, meanwhile a plurality of metadata servers can be assigned to provide fault-tolerant when occasions happen such as the downtime of the metadata servers due to excessive network loads or data read-write error on the data storage server. By adopting the method, the failure and transmission problems of an integrated storage system can be effectively solved; and the method can be widely used for storage of migration and fault tolerance of the virtual machine.
Description
Technical field
The present invention relates to the cloud computing field, especially a kind of virtual machine (vm) migration and fault-tolerant shared storage means of being applicable to.
Background technology
Cloud computing (Cloud Computing) is the new term be born in 2007, still there is no generally acknowledged definition at present.Cloud computing is a kind of business computation model, on it forms calculation task branch resource pool at a large amount of computing machines, makes the user can obtain as required computing power, storage space and information service.And cloud computing operating system is the holistic management operation system on cloud computing backstage, it refers to that framework is in the underlying hardware resource such as server, storage, network and single-set operation system, middleware, the underlying hardware of database basic software management magnanimity, the cloud platform total management system on soft resources.
Distributed file system (Distributed File System) refers to that the physical store resource of file system management not necessarily directly is connected on local node, but is connected with node by computer network.The design of distributed file system is based on Client/Server pattern.A typical network may comprise a plurality of servers for multi-user access.In addition, reciprocity characteristic allows some systems to play the part of the dual role of client-server.For example, the user can " deliver " catalogue that allows other client access, once accessed, this catalogue concerning client computer as using local drive.
Cloud computing platform is when carrying out system migration and System Fault Tolerance, and there is Single Point of Faliure in traditional centralised storage system, and the not high phenomenon of transmission performance.
Summary of the invention
The technical matters that the present invention solves is to provide a kind of virtual machine (vm) migration and fault-tolerant shared storage means of being applicable to, and effectively solves the centralised storage system and has fault, transporting problem.
The technical scheme that the present invention solves the problems of the technologies described above is:
Data storage server and meta data server be the load networks driver module respectively, by meta data server with each data storage server on fixing multicast prescription formula active searching network, each data storage server passes through keyword message, after the identification meta data server, initiatively add network storage cluster;
Data can be specified a plurality of backups according to configuration file on data storage server, can specify a plurality of meta data servers simultaneously, the meta data server machine of delaying because offered load is too high is provided, or fault-tolerant when on data storage server, the reading and writing data error situation occurs;
Specifically comprise the steps:
The 1st step, physical server is shared storage directory as host's carry of virtual machine, for depositing the image file of virtual machine;
The 2nd step creates virtual machine on physical server, and the image file of virtual machine is kept under the shared storage directory of carry, and the virtual machine image file form of support comprises Raw and VHD;
The 3rd step, share the first accesses meta-data server of client of storage on physical server, meta data server is according to the distributed store record on each data storage server of HASH algorithm statistics, follow the principle of balanced load rate to virtual machine image file memory allocated space simultaneously, in addition, virtual machine image file can create the loss of data of a plurality of backups to prevent that Single Point of Faliure from causing according to configuration;
The 4th step, physical server, according to the space path distributed, carries out IO with data server and communicates by letter, and virtual machine image file is kept under the specified path under data server;
The 5th step, virtual machine (vm) migration and fault-tolerant in, the original physical machine first discharges the shared resource of virtual machine;
The 6th step, the machine-readable listed files of getting under share directory of target physical, find virtual machine image file;
The 7th step, the target physical machine is again applied for the resource of virtual machine and is taken over the image file of virtual machine.
Adopt method of the present invention, can effectively improve virtual machine (vm) migration and fault-tolerant speed, not traffic affecting operation, the availability of raising virtual machine image.
The accompanying drawing explanation
Below in conjunction with accompanying drawing, the present invention is further described:
Fig. 1 is the inventive method schematic diagram;
Fig. 2 is customer instance schematic diagram of the present invention.
Embodiment
As shown in Figure 1, data storage server of the present invention is referred to as DS, and meta data server is referred to as MS, DS, MS load respectively this network-driven module, and by MS, with each DS on fixing multicast prescription formula active searching network, each DS passes through keyword message, after identification MS, initiatively add network storage cluster.
The present invention be take the Iinux2.6.38.2 kernel as the shared memory systems deployment platform, by self-defining operation system kernel characteristic, and the applying virtual patch, makes the network-in-dialing of system virtual support machine, I/O storage, distributed storage etc.
The present invention adopts Iinux network distribution type transmission algorithm, based on the Iinux operating system characteristic of increasing income, by the communication module between each node of the form encapsulation network storage driven, to adapt to the driving of distributed storage, called after gcloud-dnet.ko, replace original system drive Inet, with the virtual support application.
The present invention, on the basis of revising Iinux2.6.38.2 kernel source code, makes distributed memory system supporting (ext2, ext3, ext4, ntfs, fat) beyond traditional file system format, also supports the VHD virtual disk form that Microsoft releases.
On the basis of revising Iinux2.6.38.2 kernel source code, by revising original distributed storage I/O, drive, carry out the distributed store record of statistics on each DS, by fast distributed HASH algorithm, improve client reading out data efficiency, balancing network I/O load factor, improve the overall performance of storage.
The guarantee of data storages multiple spot, data can be specified a plurality of backups according to configuration file on DS, can specify a plurality of MS, and the MS machine of delaying because offered load is too high is provided, or the fault-tolerant networks of the situation such as the upper reading and writing data mistake of DS while occurring simultaneously.
The present invention includes following implementation step:
The 1st step, physical server is shared storage directory as host's carry of virtual machine, for depositing the image file of virtual machine;
The 2nd step creates virtual machine on physical server, and the image file of virtual machine is kept under the shared storage directory of carry, and the virtual machine image file form of support comprises Raw and VHD etc.;
The 3rd step, share the first accesses meta-data server of client of storage on physical server, meta data server is according to the distributed store record on each DS of HASH algorithm statistics, follow the principle of balanced load rate to virtual machine image file memory allocated space simultaneously, in addition, virtual machine image file can create the loss of data of a plurality of backups to prevent that Single Point of Faliure from causing according to configuration;
The 4th step, physical server, according to the space path distributed, carries out IO with data server and communicates by letter, and virtual machine image file is kept under the specified path under database server;
The 5th step, virtual machine (vm) migration and fault-tolerant in, the original physical machine first discharges the shared resource of virtual machine;
The 6th step, the machine-readable listed files of getting under share directory of target physical, find virtual machine image file;
The 7th step, the target physical machine is again applied for the resource of virtual machine and is taken over the image file of virtual machine.
Particularly, the inventive method can have multiple realization, and a kind of implementation will be introduced in this part.Referring to Fig. 2, it is as follows that the present invention is applicable to the concrete applying step of virtual machine (vm) migration and fault-tolerant shared storage means:
The 1st step, we use the operating system of CentOS x86_64 as host, the IP of two hosts is 192.168.1.6 and 192.168.1.7, the virtual kernel of operation Xen4.0 on it, 192.168.1.6 the fully virtualized virtual machine name of upper operation is called vm, id is 1, and under this share directory of leave in/vm of virtual machine image file, shared-file system is built with ceph, configures as follows:
The 2nd step, suppose to have two DS, and virtual machine image file has two copies, leaves in so respectively in DS, and name is called vm.img Part1 and vm.img Part2.
The 3rd step, call the migration order of Libvirt and carry out virtual machine (vm) migration,
On the host that the virtual machine (vm) migration that # is id 1 is 192.168.1.7 to the IP address
xm?migrate1192.168.1.7
The 4th step, virtual machine vm has operated on this physical machine of 192.168.1.7, and approximately 2 seconds used time, the pilot process network is less than 500 milliseconds trip time, on the impact of business, can ignore, and proves that the method is applicable to virtual machine (vm) migration and fault-tolerant.
Claims (1)
1. one kind is applicable to virtual machine (vm) migration and fault-tolerant shared storage means, it is characterized in that: data storage server and meta data server be the load networks driver module respectively, by meta data server with each data storage server on fixing multicast prescription formula active searching network, each data storage server passes through keyword message, after the identification meta data server, initiatively add network storage cluster;
Data can be specified a plurality of backups according to configuration file on data storage server, can specify a plurality of meta data servers simultaneously, the meta data server machine of delaying because offered load is too high is provided, or fault-tolerant when on data storage server, the reading and writing data error situation occurs;
Specifically comprise the steps:
The 1st step, physical server is shared storage directory as host's carry of virtual machine, for depositing the image file of virtual machine;
The 2nd step creates virtual machine on physical server, and the image file of virtual machine is kept under the shared storage directory of carry, and the virtual machine image file form of support comprises Raw and VHD;
The 3rd step, share the first accesses meta-data server of client of storage on physical server, meta data server is according to the distributed store record on each data storage server of HASH algorithm statistics, follow the principle of balanced load rate to virtual machine image file memory allocated space simultaneously, in addition, virtual machine image file can create the loss of data of a plurality of backups to prevent that Single Point of Faliure from causing according to configuration;
The 4th step, physical server, according to the space path distributed, carries out IO with data server and communicates by letter, and virtual machine image file is kept under the specified path under data server;
The 5th step, virtual machine (vm) migration and fault-tolerant in, the original physical machine first discharges the shared resource of virtual machine;
The 6th step, the machine-readable listed files of getting under share directory of target physical, find virtual machine image file;
The 7th step, the target physical machine is again applied for the resource of virtual machine and is taken over the image file of virtual machine.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110390549 CN102521063B (en) | 2011-11-30 | 2011-11-30 | Shared storage method suitable for migration and fault tolerance of virtual machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110390549 CN102521063B (en) | 2011-11-30 | 2011-11-30 | Shared storage method suitable for migration and fault tolerance of virtual machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102521063A CN102521063A (en) | 2012-06-27 |
| CN102521063B true CN102521063B (en) | 2013-12-25 |
Family
ID=46291998
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201110390549 Active CN102521063B (en) | 2011-11-30 | 2011-11-30 | Shared storage method suitable for migration and fault tolerance of virtual machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102521063B (en) |
Families Citing this family (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102882864B (en) * | 2012-09-21 | 2015-10-28 | 南京斯坦德云科技股份有限公司 | A kind of virtualization system based on InfiniBand system for cloud computing |
| CN102905059A (en) * | 2012-09-27 | 2013-01-30 | 曙光信息产业(北京)有限公司 | Continuous radio and television broadcast system |
| CN103731292A (en) * | 2012-10-16 | 2014-04-16 | 无锡云捷科技有限公司 | Fault-tolerant method enabling server to realize fault seamless recovery |
| WO2014101218A1 (en) * | 2012-12-31 | 2014-07-03 | 华为技术有限公司 | Computing and storage integrated cluster system |
| CN103458055A (en) * | 2013-09-22 | 2013-12-18 | 广州中国科学院软件应用技术研究所 | Clout competing platform |
| CN103902401B (en) * | 2014-03-14 | 2018-01-12 | 汉柏科技有限公司 | Virtual machine fault-tolerance approach and device based on monitoring |
| US9678839B2 (en) | 2014-09-12 | 2017-06-13 | Microsoft Technology Licensing, Llc | Scalable data storage pools |
| TWI621942B (en) * | 2015-01-16 | 2018-04-21 | 緯創資通股份有限公司 | Methods for session failover in os (operating system) level and systems using the same |
| CN106293994A (en) * | 2015-05-15 | 2017-01-04 | 株式会社日立制作所 | Virtual machine cloning process in NFS and NFS |
| CN106446159B (en) * | 2016-09-23 | 2019-11-12 | 华为技术有限公司 | A kind of method of storage file, the first virtual machine and name node |
| CN107967165B (en) * | 2016-10-20 | 2021-09-24 | 上海盛霄云计算技术有限公司 | Virtual machine offline migration method based on LVM |
| WO2018090290A1 (en) | 2016-11-17 | 2018-05-24 | 华为技术有限公司 | Mirror image file conversion method and apparatus |
| CN108234551B (en) * | 2016-12-15 | 2021-06-25 | 腾讯科技（深圳）有限公司 | Data processing method and device |
| CN108206750A (en) * | 2016-12-16 | 2018-06-26 | 北京国双科技有限公司 | The configuration method and device of virtual machine network interface card |
| 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 |
| CN107018193A (en) * | 2017-04-01 | 2017-08-04 | 济南浪潮高新科技投资发展有限公司 | A kind of VTPM migratory systems and method based on shared storage |
| CN109213566B (en) * | 2017-06-29 | 2022-05-13 | 华为技术有限公司 | Virtual machine migration method, device and equipment |
| KR102474582B1 (en) * | 2017-08-28 | 2022-12-05 | 삼성전자주식회사 | Method and system for preventing execution of a dirty virtual machine on an undesirable host server in a virtualization cluster environment |
| TWI669605B (en) * | 2018-06-29 | 2019-08-21 | 財團法人工業技術研究院 | Fault tolerance method and system for virtual machine group |
| US10909002B2 (en) | 2018-06-29 | 2021-02-02 | Industrial Technology Research Institute | Fault tolerance method and system for virtual machine group |
| CN109525432A (en) * | 2018-12-10 | 2019-03-26 | 浙江大华技术股份有限公司 | A kind of indoor power method for inspecting, apparatus and system |
| CN112269759B (en) * | 2020-10-23 | 2023-12-22 | 北京浪潮数据技术有限公司 | Migration method and related device for shared file storage |
| CN114780272B (en) * | 2022-04-18 | 2023-03-17 | 北京亚康万玮信息技术股份有限公司 | Intelligent fault self-healing scheduling method and device based on shared storage and virtualization |
| CN117112135A (en) * | 2023-08-25 | 2023-11-24 | 合芯科技有限公司 | Chip design platform based on container and platform architecture method |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101667187B (en) * | 2009-09-23 | 2011-09-07 | 南京邮电大学 | Compound intellectual cloud method based on P2P computing technology |
| CN101937693B (en) * | 2010-08-17 | 2012-04-04 | 深圳市子栋科技有限公司 | Video and audio playing method and system based on voice command |
| CN101951553B (en) * | 2010-08-17 | 2012-10-10 | 深圳市车音网科技有限公司 | Navigation method and system based on speech command |
| CN101969475A (en) * | 2010-11-15 | 2011-02-09 | 张军 | Business data controllable distribution and fusion application system based on cloud computing |
-
2011
- 2011-11-30 CN CN 201110390549 patent/CN102521063B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN102521063A (en) | 2012-06-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102521063B (en) | Shared storage method suitable for migration and fault tolerance of virtual machine | |
| US11120152B2 (en) | Dynamic quorum membership changes | |
| US11153380B2 (en) | Continuous backup of data in a distributed data store | |
| US10929428B1 (en) | Adaptive database replication for database copies | |
| EP3117326B1 (en) | Reducing data volume durability state for block-based storage | |
| CA2913036C (en) | Index update pipeline | |
| KR101771246B1 (en) | System-wide checkpoint avoidance for distributed database systems | |
| US20170329528A1 (en) | Efficient data volume replication for block-based storage | |
| US9053167B1 (en) | Storage device selection for database partition replicas | |
| US9880933B1 (en) | Distributed in-memory buffer cache system using buffer cache nodes | |
| CN105144121B (en) | Cache content addressable data block is for Storage Virtualization | |
| US9251003B1 (en) | Database cache survivability across database failures | |
| US10157214B1 (en) | Process for data migration between document stores | |
| US20130111471A1 (en) | Compute and storage provisioning in a cloud environment | |
| US20130047156A1 (en) | Method and apparatus of assigning virtual machine resources | |
| US11080253B1 (en) | Dynamic splitting of contentious index data pages | |
| AU2014235433A1 (en) | Fast crash recovery for distributed database systems | |
| CN102904949A (en) | Replica-based dynamic metadata cluster system | |
| CN103516549B (en) | A kind of file system metadata log mechanism based on shared object storage | |
| US10708379B1 (en) | Dynamic proxy for databases | |
| CN103581332A (en) | HDFS framework and pressure decomposition method for NameNodes in HDFS framework | |
| CN103150225B (en) | Disk full abnormity fault tolerance method of object parallel storage system based on application level agent | |
| CN105468296A (en) | No-sharing storage management method based on virtualization platform | |
| CN106383671A (en) | Block device storage cluster capability expansion system and method | |
| CN107203639A (en) | Parallel file system based on High Performance Computing |
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 |