TW201433124A - Cloud group type data center dynamic storage allocation system and method - Google Patents

Abstract

The present invention provides a cloud group type data center dynamic storage allocation system and method, in which the storages of the cloud data center are configured and managed in a manner of grouping. The method makes use of (1) a 2U/4U standard storage server subsystem, (2) a 42U/35U standard storage rack subsystem, (3) a single site distributed storage rack subsystem, (4) a multipoint group type storage site subsystem, and (5) a cloud group type data center dynamic storage allocation management subsystem or the like architecture to proceed with integrated arrangement and scheduling management, so as to achieve the purposes of automated intelligent storage configuration and data backup.

Description

Cloud group data room dynamic storage configuration system and method

The invention belongs to a dynamic storage configuration system and method in a cloud group data room, which configures and manages a storage body in a cloud data center in a group manner to achieve a smart configuration of a storage body. The purpose of automation and data backup.

Dynamic resource scheduling and management is the goal of cloud data room resource control. At present, there is a reference to failover. The data center system includes multiple data centers. Each data center includes a computer and a storage system. The computer contains a memory area for storing information about the priority order of operations between computers, and another memory area for storing the cluster service program. When a computer fails, the cluster service program is based on the operational priority order. Information, from the rest of the computer, choose a computer to take over the work of the faulty computer, but it does not mention the storage resource scheduling practices, which is obviously insufficient for the telecom operators who want to provide cloud services. The cloud group-type data room dynamic storage configuration system and method can dynamically group the computer room storage groups to serve specific application requirements, in addition to fully utilizing the existing storage resources of the equipment room, and can also achieve mutual preparation. Aid.

In the methods and systems mentioned in the existing technology, it can be supplied for virtual resources in a single computer room. The disadvantage is that the prior art can only be limited to resource management and control in a single computer room, and is intended to provide dispersion. The cloud service of the type is obviously insufficient in the management of resource scheduling. The invention provides a cloud group-type data room dynamic storage configuration system and method, which can uniformly allocate and centrally manage the storage room resources of the distributed computer rooms, and dynamically serve specific storage needs. The grouping of the equipment room is required, and the flexibility of the storage system resource scheduling can be greatly reduced and the storage equipment expenditure can be greatly reduced.

In view of the inadequacies of the foregoing systems and methods, the inventor of the present invention has invented and invented the invention, and after a long period of painstaking research, finally successfully developed and completed the dynamic storage configuration system and method of the cloud group-type data room.

The invention relates to a dynamic storage configuration system and method in a cloud group data room, which is configured and managed in a group manner by a storage body in a cloud data center (Cloud Data Center) to achieve a smart configuration of a storage body. Automation and data backup purposes.

The so-called cloud group-type machine room dynamic storage system means that the storage resources of the distributed computer room can be uniformly scheduled and centralizedly managed, and the computer room storage group can be dynamically grouped to serve specific application requirements, in addition to fully utilizing the existing storage of the machine room. Physical resources can also achieve mutual assistance. The storage systems of traditional data rooms are independent and have poor expansion. The storage systems of the computer rooms are scattered and cannot be grouped, and they cannot provide decentralized and automated resource scheduling and utilization, thus failing to effectively save capital expenditure on storage devices. . Compared with the traditional data room, the invention has the flexibility of storage system resource scheduling and can greatly reduce the effect of storage device expenditure.

The invention utilizes the cloud group type data room dynamic storage configuration management subsystem, can effectively integrate the storage bodies in the whole area, and introduces the functions of the automatic configuration into a cloud group type data room dynamic storage service system, Automatic storage system, optimal scheduling configuration and features across multiple sites. The cloud group data room dynamic storage configuration system and method proposed by the invention can simplify the storage system construction and management maintenance cost of the data room.

A cloud group-type data room dynamic storage configuration system and method, which configures and manages a storage body in a cloud data center in a group mode, and utilizes (1) 2U/4U standard storage server Subsystem, (2) 42U/35U standard storage rack subsystem, (3) single Site distributed storage rack subsystem, (4) multi-point group storage site subsystem, (5) cloud group data room dynamics Store configuration management subsystem and other architectures for integrated configuration and scheduling management to achieve storage intelligent configuration automation and data backup.

The virtual storage system of the present invention has high reliability of storage services and data, scalability of online capacity, support for multiple data copies (one or more), and Cache Data local access. The cloud group-type data room storage system service has a dynamic storage configuration mechanism, which enables users to dynamically configure according to the SLA mode, and can be based on the instant use situation of the local site storage system, including bandwidth, remaining storage space, power, etc. The fully automatic cross-region intelligent storage access strategy of the present invention dynamically configures the actual storage location of the user data to balance the resources of the storage system. The data frequently accessed by the AP (Application Program) is placed in the local location (in dual active mode), and the backed up data is configured on the remote site. The virtual storage system can support the multi-path transmission file mode, and users can obtain data from multiple group storage site subsystems simultaneously to improve the overall efficiency of the storage system. The cloud group data room dynamic storage configuration management subsystem can set data protection parameters and dynamically adjust the security, reliability and storage economy of the specified storage data. The Dynamic Storage Configuration Management Subsystem also provides a QoS-leveled control interface for customers in different regions to dynamically configure virtual storage space. It can also provide users with a control interface (GUI) to observe the current usage of the customer's stored data. .

The dynamic group configuration management subsystem of the cloud group data room of the present invention, the dynamic storage configuration management method thereof, comprises at least the following steps: mainly performing user demand storage system service points The steps of classifying and classifying the user storage system service categories according to the user SLA; providing the optimal storage room configuration and service assignment steps according to the user SLA classification, to provide the optimal configuration and location of the resources required for the respective storage services (eg, Local dual Active room, Active-Standby room; mainly provides the steps of the storage resource management scheduling of the computer room, and uses the automatic cross-region intelligent storage access policy to dynamically configure the user data design algorithm to be responsible for grouping. Management and scheduling of storage resources in the computer room; through the coordinated operation of each step, the purpose of automation of the operation management of the storage service of the computer room is achieved.

The design algorithm of the dynamic configuration user data of the automatic cross-region intelligent storage type access strategy of the present invention first classifies the user account into different area code and site code, and then utilizes the intelligent access mode in the analysis system/ Under the current situation of network usage, calculate the best storage location for access.

According to the above, the present invention can be summarized as follows: A cloud grouped data room dynamic storage configuration system includes at least: a single site distributed storage rack subsystem, which is a Cluster DFS (Distributed File System: Distributed File System). Software system, without the storage rack or storage server subsystem, provides a single point dynamic storage service with multiple storage rack subsystems in a single site, seamlessly connects multiple storage rack subsystem resources, and constructs a single Point local virtual storage pool (Storage Pool); a multi-point group storage storage room (storage site) subsystem, is to build multiple single site distributed storage rack subsystem to form a whole multi-point group storage The site subsystem can solve the imbalance of the development of storage service systems in the tier4/tier3 equipment room and the tier2 equipment room, as well as the power resources and storage system resources of the nationwide network (Site); and the dynamic storage configuration of a cloud group data room. Management subsystem, can be a multi-point group storage site The system resources are connected to form a virtual storage system for centralized storage configuration management; wherein the multi-point group storage site subsystem cooperates with the cloud group data room dynamic storage configuration management subsystem to support the span ( 2 points storage site above) to form a centralized storage configuration management. If there is a problem with the virtual machine (VM) service provided by the local site (local machine room) system, the cloud group data room dynamic storage The configuration management subsystem can support the virtual machine to boot from another virtual machine image (VM Image). If the whole system has problems with the network connection, the system can support the remote site storage system (remote site storage). The virtual machine image file in the system is restarted, so there will be no single failure between the network and the network, and the storage of the entire area can be effectively integrated, and the automatic cross-region will be The intelligent storage access policy function is imported into the cloud group-type data room dynamic storage configuration system, with automatic storage system, optimal scheduling configuration and cross-multiple Point to the characteristics of the site to achieve the purpose of intelligent storage configuration and data backup.

In the present invention, the storage rack subsystem is a storage rack system for dynamic storage, and the standard storage cabinet (Rack) has a built-in standard storage server subsystem and a switch to provide a cloud storage service.

In the present invention, the storage server subsystem is a dynamic storage system, which is a minimum basic unit built by a standard server and a built-in hard disk as a storage medium, and provides a cloud storage service to the outside.

In the present invention, the single Site distributed storage rack subsystem includes a load balancer, a data cache server, and a distributed storage system of each storage node installed in a plurality of cabinets, wherein the load balancing The device is responsible for understanding the load status based on the server connection status, and judging and matching the address translation (NAT) according to the port number in the network packet. Function, analyze and distribute the connection to the data cache server for the first layer of cache data processing; if the cache data is missing, then the data balance request is instantly distributed to the application server through the load balancer Data processing is performed and data is stored or written to the decentralized storage system.

In the present invention, the distributed storage system is composed of a Metadata server group module, a system monitoring module, and a data storage server group module, wherein the Metadata server group module is mainly composed of It consists of at least three servers that can manage the overall storage system space, convert file names into Metadata data, and perform high-level POSIX (Portable Operating System Interface) functions; It consists of at least three servers that can monitor the overall storage system. When the server device of the data storage server group module fails and is repaired or a new server is added to expand the storage space, the system will monitor the system. The detection and maintenance of an effective cluster mapping; and the data storage server cluster module is composed of a plurality of servers, and in turn combined with other modules into a plurality of rack architecture.

A cloud group-type data room dynamic storage configuration method includes at least the following steps: 1. Performing a user service storage system service classification step, classifying and determining a user storage system service category according to a user SLA (Service Level Agreement); Second, according to the user SLA classification to provide the best storage room configuration and service assignment steps to provide the best configuration and location of the resources required for their respective storage services; Third, provide the computer room storage resource management scheduling steps, the use of a fully automatic Dynamically configuring user data design algorithms for cross-region smart storage access policies to perform management and scheduling of grouped distributed computer room storage resources; and fourth, by coordinating operations of various steps to achieve cloud grouping Data room storage service operation flow management automatically Purpose.

In the present invention, the design algorithm of the dynamic configuration user data of the automatic cross-region intelligent storage type access policy first classifies the user account into different area codes, and further correspondingly to the nearest area code (area code). The local site (local site) calculates the user's storage pre-service code (area code: site code). When the user's data is stored in the storage service system, the system first authenticates the storage pre-location. Service code, so the data will be stored in the storage resource of the primary (primary) corresponding to the main equipment room, and will be stored in the next location code of the same area code (HA secondary). : High-availability second option)) The equipment room, and finally the off-site asynchronous backup mode, using the schedule mechanism to periodically store the same number of the same level of the site code remote code backup (remote backup) room; When the user wants to access the original data, after the service is authenticated, the system will first interpret the stored pre-service code, that is, calculate the area code and site code of the corresponding data, and then The corresponding primary site will be accessed. When the primary site has a problem, the corresponding area code and site code of the corresponding HA secondary data will be calculated. Therefore, the user will continue the HA secondary site. If the HA secondary server room (site) is still invalid, the system will calculate the area code and site code of the corresponding remote backup data of the backup data, and the user will work on the remote backup room. (site) issues the last access action.

The cloud group data room dynamic storage configuration system and method of the present invention have the following advantages when compared with other conventional technologies:

1. Using the storage server storage system of the present invention, it is possible to store a large amount of distributed data, and to store the stored data in a distributed manner to reduce data corruption or because of the machine. The risk of data loss due to obstacles. And because of the decentralized data storage, the speed of accessing data can be processed in parallel to improve performance, such as the application of Amazon S3.

2. The storage rack of the present invention can uniformly place and install the storage server, and manage the storage server in a consistent manner. In the field of the distributed computer room, this is an indispensable device.

3. Using the single site distributed storage rack subsystem of the present invention, a single point local virtual storage pool (Storage Pool) can be provided, so that the user can carry out the folder without knowing the location of the server to which the resource belongs. Browse with the file and access all resources under this directory.

4. The single site distributed storage rack subsystem of the present invention can provide a single point dynamic storage service and seamlessly connect multiple storage rack subsystem resources. When the software and hardware in the system are faulty or repaired, the dynamic storage The service can continue to provide services.

5. The single-site distributed storage rack subsystem of the invention can provide parallel reading of data, unlimited expansion of storage space, and remote backup function.

6. The QoS network application service, such as a CDN application service, can be provided by using the multi-point group storage site subsystem of the present invention.

7. The multi-point group storage site subsystem of the present invention can provide the function of data backup and backup, fully integrate and utilize the site power resources and storage system resources of the whole area network.

8. Using the design algorithm of the automatic cross-region intelligent storage access strategy of the present invention, the user stored data can be dynamically configured by using the smart access mode, and the best calculation is performed under the analysis system/network usage status. After the storage location, intelligent access is performed.

To sum up, this case is not only innovative in terms of technical thinking, but also has many of the above-mentioned functions that are not in the traditional methods of the past. It has fully complied with the statutory invention patent requirements of novelty and progressiveness, and applied for it according to law. Approved this invention patent application, in order to invent invention, to the sense of virtue.

The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The invention relates to a cloud group data room dynamic storage configuration system and method, which is to configure and manage a storage body in a cloud data center in a group mode, and utilize (1) 2U/4U. Standard storage server subsystem, (2) 42U/35U standard storage rack subsystem, (3) single Site distributed storage rack subsystem 5, (4) multi-point group storage site subsystem, (5) cloud group The data storage room dynamic storage configuration management subsystem 6 and other architectures for integrated configuration and scheduling management to achieve the purpose of storage smart configuration automation and data backup.

The design of the dynamic group configuration system of the cloud group data room proposed by the invention comprises the following subsystems and functions: 2U/4U standard storage server subsystem: the dynamic storage system uses a standard 2U/4U server and a built-in hard disk as a storage body. The minimum basic unit is built to provide cloud storage services, for example: 2U server with 12 SATA HD or 4U server with 24 SATA HD; 42U/35U standard storage rack subsystem: dynamic storage storage rack subsystem to standard 42U/35U chassis (Rack) built-in 2U/4U standard storage server subsystem and switch, external Provides cloud storage service; single site distributed storage rack subsystem 5: Cluster DFS (Distributed File System) software system, does not bind storage rack or storage server subsystem, provides single point in multiple sites with multiple storage rack subsystems Dynamic storage service, seamlessly connecting multiple storage rack subsystem resources to form a single point local virtual storage pool (Storage Pool); multi-point group storage site subsystem: multiple single site distributed storage The rack subsystem is built into an overall multi-point group storage site subsystem, which can solve the imbalance of tier4/tier3 equipment room and tier2 equipment storage service system development, and integrate the site power resources and storage of the whole area network. System resources; cloud grouped data room dynamic storage configuration management subsystem 6: This subsystem needs to be able to connect the entire multi-point group storage site subsystem resources with the network management interface to form a webized concentration Store the virtual storage pool system of configuration management, and the user program (AP) of each site can be based on SLA dynamics. Virtual storage service provided by the cloud grouped data room accessed. The multi-point group storage site subsystem and the dynamic storage configuration management subsystem can support cross-region (more than 2 points) to form a comprehensive storage configuration management. If there is a problem in the near virtual machine (VM), dynamic storage configuration management The subsystem can support the virtual machine to boot from another virtual machine image (VM Image). If the entire system and the network have problems, the system can support the virtual machine image in the remote storage system. Restart, so there will be no single failure between the network and the network.

Please refer to FIG. 1 , which is a schematic diagram of the 2U storage server subsystem 1 of the present invention. The general standard size is 19 inches wide and 3.5 inches high, and can be placed in a 19 inch wide rack. Head The former mainstream CPU uses one processor (Quad-core), but can support up to two processors, providing the required speed in terms of computing power. The memory can be 1GB, 2GB, 4GB or 8GB, and provides 16 memory slots up to 128GB. As a storage server, the memory is not a critical factor in the memory, and the capacity of the hard disk is Yes, however, the size of the hard disk is important, and the capacity of the memory cannot be ignored. The larger the memory capacity, the larger the buffer space that can be used as the storage, and it plays a pivotal role in the overall storage performance. The larger the buffer space, the more data can be temporarily stored, sharing the burden of hard disk reading or writing. For the Storage Controller, it is equipped with a RAID controller card, which can achieve RAID 0/1/5/6 different levels of data backup mechanism. The hard disk is provided with at least 12 3.5-inch hot-swappable SATA/SAS HDDs. As for the capacity of one hard disk, it is 1TB in the past, and the current mainstream is 2TB. In the future, it is possible to use 3TB as the mainstream. This can greatly improve the capacity of the storage server itself, and can store more data with the same size server.

Please refer to FIG. 2, which is a schematic diagram of the 4U storage server subsystem 2 of the present invention. The general standard size is 19 inches wide and 7 inches high, and can be placed in a rack of 19 inches wide. Since the size is twice that of the 2U storage server, the default configuration of the CPU is two processors, each of which can be Quad-core. The memory can reach up to 192GB, so the memory capacity that can be used is more than that of the 2U storage server. The number of hard disks that can be accommodated is twice that of the 2U storage server, which is 24 3.5-inch Hot-Swap (SAS/SATA) hard drives. It also has a RAID controller card to provide backup mechanism, the same level of RAID 0/1/5/6 is available.

Please refer to FIG. 3, which is a schematic diagram of the 42U standard storage rack subsystem 3 of the present invention. The 42U standard rack is about 40 inches long, about 24 inches wide, and about 78 inches high. The rack itself is a detachable combination that includes Lockable Doors, Removable Side Panels, and Perforated Doors, Easy Leveling Castors, Adjustable Mounting Posts, Square Mounting Tabs, and Cable Access, you can adjust the height of the partition to place the storage server, provide power, vents and network holes, and place different labels or different designs according to your needs. A high-level storage server with rack stabilizers, backplanes, and grounding kits for added protection for installation or rack mounting.

Referring to Figure 4, there is shown a schematic diagram of a 35U standard storage rack subsystem 4 of the present invention. This is a 35U standard rack having a length of about 40 inches, a width of about 24 inches, and a height of about 68 inches. The rack itself, like the 42U rack, is a detachable combination that makes it easy to assemble or disassemble the Side Panel without the need for tools. Components such as fans, base rollers, and the like are provided to make installation easier.

Please refer to FIG. 1 , which is a schematic diagram of a single Site distributed storage rack subsystem 5 of the present invention, including a load balancer 51 , a data cache server 52 , and distributed storage of storage nodes installed in a plurality of cabinets. System 53 has three parts. The load balancer 51 is responsible for understanding the load status according to the server connection status, and judging and matching the address conversion (NAT) function according to the port number in the network packet, analyzing and distributing the connection to the data. The cache server 52 performs the first layer cache data processing; if the cache data is missed, the data request is immediately distributed to the application server AP 54 through the load balancer for data processing, and the data is stored or written. Into the decentralized storage system 53. The single Site distributed storage rack subsystem of the present invention can not only reduce the burden on the storage device, but also increase the speed of data access.

The distributed storage system 53 of the present invention comprises three main modules: a Metadata server group module 531, a system monitoring module 532, and a data storage server group module 533, wherein:

1. Metadata server group module 531: Metadata server group module can be mainly 3~5 It consists of a server that manages the overall storage system space, converts file names to Metadata data, and performs high-level POSIX functions.

2. System monitoring module 532: The system monitoring module can be mainly composed of 3~5 servers, which can monitor the overall storage system, when the server of the data storage server group mode fails and repairs or the new server is added. To extend storage space, the system monitoring module detects and maintains a valid cluster map.

3. Data Storage Server Group Module 533: The data storage server group module is composed of a plurality of servers, and is combined with other modules to form a plurality of rack structures. This module is mainly to assist in the reading and writing of data. The data of user operations are managed and stored by the object storage cluster.

Referring to FIG. 2, the multi-point group storage site subsystem architecture diagram of the present invention is a multi-point group storage site subsystem in which a plurality of single site distributed storage rack subsystems 5 are constructed as a whole. Each single-site distributed storage rack subsystem 5 includes a Firewall/IPS security defense device to prevent intrusion and destruction of abnormal traffic.

The multi-point group storage site subsystem must be connected via the cloud group data room dynamic storage configuration management subsystem 6 to support the cross-region (2 points) in conjunction with the cloud group data room dynamic storage configuration management subsystem 6. The above) management forms a storage configuration management map integrated in the whole area. If there is a problem in the near virtual machine (VM), the dynamic storage configuration management subsystem 6 can support the virtual machine from another virtual machine image file (VM Image). Startup, if there is a problem with the entire system and the network, the system can support the virtual machine image in the remote storage system and restart it, so there will be no single failure between the near network and the network (Single Failure). happened.

The AP for each site can be dynamically configured according to the SLA mode, and the CDN should be provided. Use the service. In addition, this subsystem can also solve the imbalance of tier4/tier3 room and tier2 room storage service system development, and can dynamically store more important and frequently used data in tier3/tier4 site, which will be less important or less. The data used is stored in the tier2 site, or the backup data is stored in the tier2 site to achieve the purpose of data backup, and the site power resources and storage system resources of the entire network are fully integrated and utilized.

Referring to FIG. 3, the dynamic storage configuration management method of the present invention includes at least the following steps: mainly performing the steps of the user demand storage system service classification 61, classifying and discriminating the user storage system service category according to the user SLA; classifying according to the user SLA The steps of providing the optimal storage room configuration and service assignment 62 to provide the best configuration and location of the resources required for the respective storage services (such as the local dual Active room and the Active-Standby room); The step of using a fully automatic cross-region intelligent storage access policy to dynamically configure the user data design algorithm to manage and schedule the distributed storage room storage resources of the group; The purpose of computer room storage service operation process management automation.

For the design algorithm of the dynamic configuration user data of the automatic cross-region intelligent storage access policy of the present invention, as shown in FIG. 5, the algorithm first classifies the user accounts into different area codes, such as userA and userB. Assigned to area code:02, userC, userD is assigned to area code:07. Further, the user is corresponding to the computer room of the nearest local site, for example, userA and userB are assigned to site code: 01 and 02 in area code: 02, and userC and userD are assigned to site code: 01 and 02 in area code: 07. Therefore, the storage pre-service codes of userA, userB, userC, and userD are 02:01, 02:02, 07:01, and 07:02, respectively.

When the dataA data of userA is stored in the service system, the system will first verify that the storage pre-service code is 02:01, so it will save the resources of area code:02, site code:01, and It will immediately use the remote synchronous backup mode and then deposit it into the next code site code room of the same area code (here is the site code: 02 machine room), and finally use the off-site asynchronous backup mode, using the schedule mechanism to periodically save Go to the next number of site code (here is area code: 07) the same number of site code remote data backup machine room (here is site code: 01 machine room). The other userB's dB, userC's dC, and userD's dD data access methods are the same as those of userA. Please refer to Figure 5.

Then, when userA wants to access the original dataA data, after the service authentication, the system will first calculate the corresponding primary data of the area code: site code is 02:01, then userA will serve the 02:01 site service. Access. When there is a problem with the 02:01 site (such as network interruption, line congestion, host failure, storage abnormality, etc.), the system automatically configures the user profile of the automatic cross-region intelligent storage access policy. The design algorithm will calculate the corresponding area code:site code of the corresponding High Availability secondary data of 02:02, so userA will continue to issue access operations to the 02:02 site. If the site of 02:02 is still invalid, the design algorithm of the dynamic configuration user data of the automatic cross-region intelligent storage access strategy of the system will calculate the final step (that is, the backup data is used), and the corresponding remote backup data Area code:site code is 07:01, userA will send the last access action to the site of 07:01 (since 07:01 is the site of remote backup, the data will have the problem of non-synchronization time difference, so the action should match Snapshot to make file corrections). Other userB's dB, userC's dC, userD's dD intelligent data access mode and data backup access mode are the same as userA's algorithm, please refer to Figure 5.

The virtual storage system of the present invention has high reliability of storage services and data, scalability of online capacity, support for multiple data copies (more than one copy), and local access of cache data (Cache Data). Features. The cloud group-type data room storage system service has a dynamic storage configuration mechanism, which enables users to dynamically configure according to the SLA mode, and can be based on the instant use situation of the local site storage system, including bandwidth, remaining storage space, power, etc. The fully automatic cross-region intelligent storage access strategy of the present invention dynamically configures the actual storage location of the user data to balance the resources of the storage system. The data frequently accessed by the AP is configured in the local location (in the dual active mode), and the backed up data is configured in the remote site. The virtual storage system can support the multi-path transmission file mode, and users can obtain data from multiple group storage site subsystems simultaneously to improve the overall efficiency of the storage system. The cloud group data room dynamic storage configuration management subsystem 6 can set data protection parameters and dynamically adjust the security, reliability and storage economy of the specified storage materials. The cloud grouped data room dynamic storage configuration management subsystem 6 also provides a QoS hierarchical control interface for customers in different regions to dynamically configure the virtual storage space, and also provides a control interface (GUI) to the user to observe the customer's Store the current usage of the data.

The above is only the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention; the modifications and equivalents of the present invention are intended to be included in the scope of the present invention without departing from the spirit and scope of the invention. Within the scope of protection.

1‧‧‧2U storage server subsystem

2‧‧‧4U storage server subsystem

3‧‧‧42U standard storage rack subsystem

4‧‧‧35U standard storage rack subsystem

5‧‧‧Single Site Distributed Storage rack Subsystem

51‧‧‧Load balancer

52‧‧‧Data cache server

53‧‧‧Distributed storage system

531‧‧‧Metadata Server Group Module

532‧‧‧System Monitoring Module

533‧‧‧Data Storage Server Group Module

54‧‧‧AP

55‧‧‧DB

6‧‧‧Cloud Group Data Room Dynamic Storage Configuration Management Subsystem

61‧‧‧User Demand Storage System Service Classification

62‧‧‧Best storage room configuration and service assignment

63‧‧‧Computer room storage resource management scheduling

The following is a detailed description of the present invention and its accompanying drawings, which will further understand the technical contents of the present invention and its effects; the related drawings are: FIG. 1 is a single Site distributed storage rack subsystem architecture diagram; The system is a multi-point group storage site subsystem architecture diagram; FIG. 3 is a flowchart of the dynamic storage configuration management method of the present invention; 1 is a schematic diagram of a 2U storage server subsystem of the present invention; FIG. 2 is a schematic diagram of a 4U storage server subsystem of the present invention; and FIG. 3 is a schematic diagram of a 42U standard storage rack subsystem of the present invention; The schematic diagram of the 35U standard storage rack subsystem of the present invention; FIG. 5 is a design algorithm for dynamically configuring user data of the automatic cross-region intelligent storage access strategy.

51‧‧‧Load balancer

52‧‧‧Data cache server

53‧‧‧Distributed storage system

531‧‧‧Metadata Server Group Module

532‧‧‧System Monitoring Module

533‧‧‧Data Storage Server Group Module

54‧‧‧AP

55‧‧‧DB

Claims (7)

A cloud group-type data room dynamic storage configuration system includes at least: a single site distributed storage rack subsystem, which is a software system of a Cluster DFS (Distributed File System), which is not bound to a storage system. The rack or storage server subsystem provides a single point dynamic storage service with multiple storage rack subsystems in a single site, seamlessly connecting multiple storage rack subsystem resources, and constructing a single point local virtual storage pool (Storage) Pool); a multi-point group storage site subsystem, which is constructed by constructing a single multi-site distributed storage rack subsystem to form a holistic multi-point group storage site subsystem, which can solve tier4/tier3 The imbalance between the development of the storage service system in the equipment room and the tier2 room, as well as the integration of the nationwide network (Site) power resources and storage system resources; and the cloud group-type data room dynamic storage configuration management subsystem, which can The point group storage site subsystem resources are connected to form a virtual storage system for centralized storage configuration management; The multi-point group storage site subsystem cooperates with the cloud group data room dynamic storage configuration management subsystem to support cross-region (above 2 storage sites) to form a comprehensive storage configuration management, if near (local site: local machine room) When the virtual machine (VM) service provided by the system has a problem, the cloud grouped data room dynamic storage configuration management subsystem can support the virtual machine from another virtual machine image file ( VM Image) starts. If there is a problem with the connection between the entire system and the network, the system can support the virtual machine image file in the remote site storage system to restart it, so it will not be near. Single failure between the station and the network, and the storage of the entire area can be given With effective integration, the automatic cross-region intelligent storage access policy function is imported and integrated into the cloud group-type data room dynamic storage configuration system, with automatic storage system, optimal scheduling configuration and characteristics across multiple sites. To achieve the purpose of intelligent storage configuration and data backup for storage. For example, the cloud group type data room dynamic storage configuration system described in claim 1 wherein the storage rack subsystem is a dynamic storage storage rack system, and the standard chassis (Rack) has a built-in standard storage server subsystem. With the switch, the cloud storage service is provided externally. The cloud group-type data room dynamic storage configuration system according to the first aspect of the patent application, wherein the storage server subsystem is a dynamic storage system, and the minimum basic configuration is implemented by using a standard server and a built-in hard disk as a storage medium. Unit, providing cloud storage services to the outside world. The cloud group-type data room dynamic storage configuration system described in claim 1, wherein the single-site distributed storage rack subsystem includes a load balancer, a data cache server, and a plurality of cabinets. Three parts of the distributed storage system of each storage node, wherein the load balancer is responsible for understanding the load status according to the connection status of the server, and judging and matching the address conversion according to the port number in the network packet ( NAT) function, analyze and distribute the connection to the data cache server for the first layer cache data processing; if the cache data is missing, then the data balance request is instantly distributed to the application through the load balancer The server performs data processing and stores or writes data to the distributed storage system. The cloud group-type data room dynamic storage configuration system described in claim 4, wherein the distributed storage system is controlled by a metadata (Metadata) server group module and system. The module and the data storage server group module, wherein the Metadata server group module is mainly composed of at least three servers, can manage the overall storage system space, convert file names into Metadata data, and execute high. POSIX (Portable Operating System Interface) function; the system monitoring module is mainly composed of at least three servers, which can monitor the overall storage system, when the data storage server group mode servo When the device fails and is repaired or a new server is added to expand the storage space, the system monitors and maintains an effective cluster mapping; and the data storage server group module is composed of multiple servers. It is combined with other modules to form multiple rack architectures. A cloud group-type data room dynamic storage configuration method includes at least the following steps: 1. Performing a user service storage system service classification step, classifying and determining a user storage system service category according to a user SLA (Service Level Agreement); Second, according to the user SLA classification to provide the best storage room configuration and service assignment steps to provide the best configuration and location of the resources required for their respective storage services; Third, provide the computer room storage resource management scheduling steps, the use of a fully automatic Dynamically configuring user data design algorithms for cross-region smart storage access policies to perform management and scheduling of grouped distributed computer room storage resources; and fourth, by coordinating operations of various steps to achieve cloud grouping Data room storage service operation process management automation purposes. For example, the dynamic grouping configuration method of the cloud group data room according to the sixth aspect of the patent application scope, wherein the design algorithm of the dynamic configuration user data of the automatic cross-region intelligent storage type access policy first classifies the user account into Different area codes, and further corresponding to them To the nearest local site, the user's storage pre-service code (area code: site code) is calculated; when the user's data is stored in the storage service system, the system first authenticates the storage. The pre-service code, so the data will be stored in the storage resource of the primary (primary) corresponding to the main equipment room, and will be immediately stored in the same location code of the same area code (HA secondary (synchronous backup mode). High Availability: High-availability second option)) The equipment room, and finally the off-site asynchronous backup mode, using the schedule mechanism to periodically store the same number of the next level of the same site code remote data backup (remote backup) room When the user wants to access the original data, after the service is authenticated, the system will first interpret the stored pre-service code, that is, calculate the area code and site code of the corresponding data, and then the corresponding primary Site access service; when there is a problem with the primary site, it will calculate the corresponding area code and site code of the corresponding HA secondary data, so the user will HA seconda The site of ry continues to issue access actions; if the site of the HA secondary is still invalid, the system will calculate the area code and site code of the corresponding remote backup data of the backup data. The remote backup's site sends the last access action.