CN103888516A - Cloud storage platform with QoS guarantee - Google Patents
Cloud storage platform with QoS guarantee Download PDFInfo
- Publication number
- CN103888516A CN103888516A CN201410073780.2A CN201410073780A CN103888516A CN 103888516 A CN103888516 A CN 103888516A CN 201410073780 A CN201410073780 A CN 201410073780A CN 103888516 A CN103888516 A CN 103888516A
- Authority
- CN
- China
- Prior art keywords
- data
- data center
- cloud storage
- storage platform
- data centers
- 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
Images
Abstract
The invention provides a cloud storage platform with QoS guarantee. The cloud storage platform comprises four levels of data centers, and virtual data centers containing a plurality of levels of the data centers are divided according to geographic positions; duplicates of the same data are cached on different levels of data center storage systems in the virtual data centers; the first-level data center is provided with a global resource controller which controls data access loads of services of different QoS levels by allocating network and storage system resources cloud-stored in all data centers; all other levels of the data centers are provided with partial resource controllers so that network flow of the data centers and storage system load conditions can be reported to the global controller in real time; a user downloads data from the fastest data center through an API of the cloud storage platform. By means of the cloud storage platform, the balance condition of the network flow and the storage system load of cloud storage platform of a wide area network is obviously improved, the access speed of the user is improved, and the overall throughput rate of the system is improved.
Description
Technical field
The present invention relates to the technical field of computer information management, the distributed cloud storage system of a kind of wide area network specifically, particularly its system syndeton and resource regulating method.
Background technology
Along with constantly popularizing of cloud computing technology, the data service quality (QoS) of user to cloud storage platform has higher demand.The cloud storage platform conventionally multiple data centers based on wide area network builds.But the disposal ability that network between different pieces of information center connects bandwidth and data center's internal storage system there are differences, can cause the cloud storage platform cannot be for user provides in real time, reliable data, services as traditional enterprise storage system.For example, system can run into the access request peak of burst, causes localized network or global network to occur congested; In addition, the reliability of wide area network link is not high, cannot ensure platform service quality yet.
The object of cloud storage system being carried out to QoS scheduling is exactly to adopt the controlled network equipment and the storage system of resource, realize the dynamic management of the interior network traffics of cloud storage platform and system load, make manager can effectively control data traffic and the route of physics and virtual network, prevent that the service quality that network or system congestion cause from declining, the impact of avoiding link failure to bring, and promote cloud storage system overall throughput.
Summary of the invention
The object of the present invention is to provide one to there is QoS and ensure cloud storage system, build highly reliable a, high availability and high performance cloud storage platform.Network traffics and the storage system load of each data center monitored and assessed to cloud storage platform resouce controller, distribute and dispatch the resource of different QoS rank data service in each data center, ensure the service quality of QoS data services at different levels, and make entire system resource utilization optimum.
To achieve these goals, the present invention is by the following technical solutions:
There is the cloud storage platform that QoS ensures, adopt the data center of the multiple different stages on hierarchy type architecture management wide area network.As shown in Figure 2, comprise the data center of four ranks, one-level data center only has one, and other data centers at different levels can have multiple;
The tree network of a stratification of data centers at different levels nodes composition, covers taking fourth stage data center as minimal geographical, using this data center and higher level data center thereof as a virtual data center;
Virtual data center, for the copy of same data is buffered in to the different pieces of information central store system in virtual data center, is divided into three kinds of different QoS ranks by data service;
Global network resouce controller is installed in one-level data center, localized network resouce controller is installed by other data centers at different levels, to real-time traffic situation and the access load of the data services at different levels of global resource controller reporting, global resource controller is by each data center of intelligent dispatching algorithm control cloud storage platform, the data traffic flow amount of different QoS rank.
In the present invention, data center is always divided into 4 grades, according to the scale of data center and rank, and descending one-level, secondary, three grades and the level Four data center of being respectively.
In the present invention, can be according to location distribution, divide the virtual data center that comprises multiple rank data center.
The copy of same data is buffered in the different pieces of information central store system in virtual data center.One-level data center has total data copy, and it covers all data trnascriptions to be assigned to each secondary data center by geography, and secondary data center is assigned to data trnascription subordinate's three DBMS centers again;
Data service is divided into three kinds of different QoS ranks, is respectively one-level, secondary and three grades.Three kinds of data services reduce successively to the susceptibility of time delay, also step-down successively of priority.First order service priority is the highest, and requirement can make an immediate response; Second level service priority allows to complete within a certain period of time response in being; Third level service priority is minimum, in the time that system is busy, can suspend response.
User, by cloud storage platform interface API, downloads required data from the fastest data center, or by Parallel download, from multiple data centers downloading data simultaneously.
The present invention can significantly improve wide area network cloud storage platform system load balancing situation, elevator system overall throughput.For user chooses optimum access node, accelerate user's access speed by automatically, promote user data services quality.In addition, platform can ensure to break down at the network link of Wide Area Network or certain data center's node, and business service is without interruption.
Brief description of the drawings
Fig. 1 is cloud storage inner structure and network connection diagram.
Fig. 2 is each data center's hierarchical relationship schematic diagram in cloud storage
Fig. 3 is resouce controller schematic diagram in cloud storage.
Embodiment
Adopt the data center of the multiple different stages on hierarchy type architecture management wide area network.As shown in Figure 1, one-level data center is located at regional center area machine room conventionally, is convenient to the network coverage.One-level data center disposes two core switch at network boundary place, converge two up links at each secondary data center, avoids Single Point of Faliure.
Secondary data center is the shunting to one-level data center, and machine room is built in the central area of level-2 area conventionally.Secondary data center adopts two optical fiber private-line mode access one-level data centers, to realize the high-speed data exchange between operation system.Article two, special line both can increase bandwidth by polymerization, also can realize link redundancy.
Three DBMS centers are the further shuntings to secondary data center, and machine room is built in the edge zone of level-2 area, and three DBMS centers are by optical fiber access via telephone line one-level data center and its higher level's secondary data center.The effect that two special lines are set is that meanwhile, secondary data center is as the convergent point at administer three DBMS centers in order to realize link redundancy and load balancing equally.
Level Four data center covers the minimum in region, adopts special line to realize and its higher level's three DBMS centers and the access of the dual link at secondary data center.
The bottom of cloud storage is in the distributed objects storage system of using based on wide area network.System builds global name space in wide area network memory node cluster, realizes the redundancy of data and fault-tolerant by copying object; The ability that simultaneity factor has automatic reparation and automatically manages, for example can be according to the node of file access temperature Autonomic Migration Framework file storage; In addition, system is also supported business zero interruption of upgrading and dilation process.
As shown in Figure 3, system control framework is divided into three layers: i.e. resource apparatus layer, Local resource key-course and global resource key-course.
Resource apparatus layer refers to physical equipment, comprises the network equipment and memory device, is deployed in each data center, and the system resource based on distributing provides data, services to user.
Local resource key-course refers to Local resource Control Server, is deployed in each data center's outlet, processes the scheduling of the network bandwidth and memory load, to global resource key-course reporting state information.Local resource controller can send to global controller the loading condition of the Link State of switch, bandwidth information, storage system, distributes to carry out bandwidth adjustment, path planning and storage resources.
Global resource controller is deployed in one-level data center, collects network link and memory load information that the local resouce controller of each data center reports, for different QoS level traffic, realizes cloud storage system is carried out to overall unified resource control.
The network bandwidth of global optimum and storage system resource allocation policy are to realize the core technology that cloud storage system QoS ensures, according to the flow of different QoS level traffic and loading condition, use global optimum's resource scheduling algorithm to calculate flow bandwidth and the memory load resource that each data center should distribute.In the time that new business data start to transmit, can consumable bandwidth and storage resources according to this business, be that it selects an optimum path.The resource information of distributing can be issued in the Local resource controller of each data center, again by the resource apparatus of controller control data center, thereby ensure the performance requirement of different QoS level traffic, promote cloud storage system overall throughput, realize cloud storage overall network and utilization ratio of storage resources optimization.
Claims (1)
1. have the cloud storage platform that QoS ensures, comprise the data center of four ranks, one-level data center only has one, and other data centers at different levels can have multiple;
The tree network of a stratification of data centers at different levels nodes composition, covers taking fourth stage data center as minimal geographical, using this data center and higher level data center thereof as a virtual data center;
Virtual data center, for the copy of same data is buffered in to the different pieces of information central store system in virtual data center, is divided into three kinds of different QoS ranks by data service;
Global resource controller is installed in one-level data center; Local resource controller is installed by other data centers at different levels; to real-time network flow and the system load situation of the data services at different levels of global resource controller reporting; global resource controller is by each data center of intelligent dispatching algorithm control cloud storage platform the whole network, the data access load of different QoS level traffic;
User, by cloud storage platform interface API, downloads required data from the fastest data center, or by Parallel download, from multiple data centers downloading data simultaneously.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410073780.2A CN103888516A (en) | 2014-02-28 | 2014-02-28 | Cloud storage platform with QoS guarantee |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410073780.2A CN103888516A (en) | 2014-02-28 | 2014-02-28 | Cloud storage platform with QoS guarantee |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103888516A true CN103888516A (en) | 2014-06-25 |
Family
ID=50957232
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410073780.2A Pending CN103888516A (en) | 2014-02-28 | 2014-02-28 | Cloud storage platform with QoS guarantee |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103888516A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104199741A (en) * | 2014-08-29 | 2014-12-10 | 曙光信息产业(北京)有限公司 | Virtual data management method for cloud computing environment |
| CN107197055A (en) * | 2017-08-01 | 2017-09-22 | 成都鼎智汇科技有限公司 | One kind realizes storage resource assigned unit beyond the clouds |
| CN107241448A (en) * | 2017-08-01 | 2017-10-10 | 成都鼎智汇科技有限公司 | A kind of cloud storage node load balancing method |
| CN107645731A (en) * | 2017-09-21 | 2018-01-30 | 北京邮电大学 | Load-balancing method based on self-organizing resource allocation in a kind of non-orthogonal multiple access system |
| CN108512692A (en) * | 2018-02-24 | 2018-09-07 | 国家计算机网络与信息安全管理中心 | A kind of resource integrated method and device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102281329A (en) * | 2011-08-02 | 2011-12-14 | 北京邮电大学 | Resource scheduling method and system for platform as a service (Paas) cloud platform |
| US20140053226A1 (en) * | 2012-08-14 | 2014-02-20 | Ca, Inc. | Self-adaptive and proactive virtual machine images adjustment to environmental security risks in a cloud environment |
-
2014
- 2014-02-28 CN CN201410073780.2A patent/CN103888516A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102281329A (en) * | 2011-08-02 | 2011-12-14 | 北京邮电大学 | Resource scheduling method and system for platform as a service (Paas) cloud platform |
| US20140053226A1 (en) * | 2012-08-14 | 2014-02-20 | Ca, Inc. | Self-adaptive and proactive virtual machine images adjustment to environmental security risks in a cloud environment |
Non-Patent Citations (1)
| Title |
|---|
| 夏纯中，宋顺林: ""基于商空间的层次式数据网格资源调度算法"", 《通信学报》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104199741A (en) * | 2014-08-29 | 2014-12-10 | 曙光信息产业(北京)有限公司 | Virtual data management method for cloud computing environment |
| CN107197055A (en) * | 2017-08-01 | 2017-09-22 | 成都鼎智汇科技有限公司 | One kind realizes storage resource assigned unit beyond the clouds |
| CN107241448A (en) * | 2017-08-01 | 2017-10-10 | 成都鼎智汇科技有限公司 | A kind of cloud storage node load balancing method |
| CN107645731A (en) * | 2017-09-21 | 2018-01-30 | 北京邮电大学 | Load-balancing method based on self-organizing resource allocation in a kind of non-orthogonal multiple access system |
| CN107645731B (en) * | 2017-09-21 | 2019-10-08 | 北京邮电大学 | Load-balancing method based on self-organizing resource allocation in a kind of NOMA system |
| CN108512692A (en) * | 2018-02-24 | 2018-09-07 | 国家计算机网络与信息安全管理中心 | A kind of resource integrated method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111953526B (en) | Hierarchical computational power network arrangement method, device and storage medium | |
| CN102427475B (en) | Load balance scheduling system in cloud computing environment | |
| US8918566B2 (en) | System and methods for allocating shared storage resources | |
| CN102469023B (en) | Based on the dispatching method of cloud computing, unit and system | |
| JP5324637B2 (en) | Dynamic flowlet scheduling system, flow scheduling method, and flow scheduling program | |
| CN105610715B (en) | A kind of cloud data center multi-dummy machine migration scheduling method of planning based on SDN | |
| CN107548549A (en) | Equilibrium of stock in DCE | |
| CN103888516A (en) | Cloud storage platform with QoS guarantee | |
| CN105468435A (en) | NFV dynamic resource distribution method | |
| CN104679594B (en) | A kind of middleware distributed computing method | |
| CN101227416A (en) | Method for distributing link bandwidth in communication network | |
| CN107046567A (en) | A kind of distributed cloud computing system for the centralized management being made up of thin cloud system | |
| CN104468407A (en) | Method and device for performing service platform resource elastic allocation | |
| CN110221920B (en) | Deployment method, device, storage medium and system | |
| CN104539744B (en) | A kind of the media edge cloud dispatching method and device of two benches cooperation | |
| CN102567106A (en) | Task scheduling method, system and device | |
| CN105893113A (en) | Management system and management method of virtual machine | |
| CN101427533A (en) | Broadband access network capacity management | |
| CN106293933A (en) | A kind of cluster resource configuration supporting much data Computational frames and dispatching method | |
| CN102480502A (en) | I/O load equilibrium method and I/O server | |
| CN114416355A (en) | Resource scheduling method, device, system, electronic equipment and medium | |
| CN102448120B (en) | Multipath load sharing method and device | |
| CN107197039B (en) | A kind of PAAS platform service packet distribution method and system based on CDN | |
| CN105721545A (en) | Multi-level cluster management realization method | |
| Bhatia et al. | Perspective study on load balancing paradigms in cloud 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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140625 |
|
| WD01 | Invention patent application deemed withdrawn after publication |