TW201606522A - Fusion virtualized storage system - Google Patents

Abstract

A fusion virtualized storage sub-system is disclosed, and the system comprises a first server node, a second server node and a common storage pool including a hard disk cluster. The first server node and the second node server simultaneously access the common storage pool. The fusion virtualized storage sub-system includes an operation system, and a storage space unit is defined in the operation system. The first and second server nodes access the hard disk cluster in the common storage pool through the storage space unit, so as to synchronize the first and second server nodes for a redundant storage configuration. The operation system includes a file server and a virtual server. Based on stacked applications of the operation system, the virtual server is connected to the file server to match the virtual server and file server.

Description

Converged virtualized storage system

The present invention relates to the field of server architecture, and in particular, to a converged virtualized storage system.

In the existing enterprise user storage solution, as shown in FIG. 1 , the traditional virtualized storage system 100 is composed of a plurality of storage devices 110 (which are JBOD, collectively called Just a bunch of disks, or hard disk clusters). The file server 120 and the plurality of virtual servers 130 form a clustered three-layer architecture, and each layer is interconnected by a complicated cable (such as a SAS cable or an optical fiber) and passes through a complex software layer. The configuration completes the topology configuration of the entire system structure. As shown in FIG. 1, each of the storage devices 110 (for example, JBOD1, JBOD2, JBODn, etc.) is respectively connected to each of the file servers 120 (for example, file server 1, file server 2, file server n, etc.). Each of the file servers 120 is connected to each of the virtual servers 130 (eg, virtual server 1, virtual server 2, virtual server n, etc.), and each of the virtual servers 130 is connected to a switch 140. The 140 is connected to a plurality of client terminals 150 (for example, client terminal 1, client terminal 2, client terminal n, etc.). JBOD, as an important storage device in the storage field, can also be called a spanned disk. JBOD does not have front-end logic to manage the distribution of data on the hard disk, but each hard disk is individually addressed as a separate storage resource. The file server 120 is a type of system management function with time-sharing system and provides concurrent access for users and files. A web server that controls and secures security measures. The virtual server 130 divides itself into a plurality of hosts through software and hardware technologies, and each divided host has an independent function variable name and an IP address, but shares the CPU, RAM, operating system, and application of the real host. Relative to the user, the server being operated has the same function as the real independent host. Therefore, the user can access the material information in the storage device 110 through the client 150, the switch 140, the virtual server 130, and the file server 120. Since each of the storage devices 110 is connected to each of the file servers 120, each of the file servers 120 is connected to each of the virtual servers 130, and each of the virtual servers 130 is connected to a switch 140, thus The entire traditional virtualized storage system 100 has a better redundancy design (or redundancy design). However, such a redundant design causes the conventional virtualized storage system 100 to have the following drawbacks: First, the intricate network cable between the virtual server 130 and the file server 120 requires a software configuration topology, the file server 120 and the storage device. Intricate network cables between 110s also require a software configuration topology. Second, in the case of capacity expansion, complex cable connections and configuration interfaces can make the expansion very cumbersome and inefficient.

In view of this, it is highly desirable to provide a new converged virtualized storage system to solve the above problems.

It is an object of the present invention to provide a converged virtualized storage system that combines a three-layer architecture of a conventional virtualized storage system into a cluster system, and utilizes a virtualization solution to flatten the configuration of the entire storage system, and The complicated cable configuration between the three-layer architecture of the traditional virtualized storage system can be omitted, so that the cluster expansion of the fusion virtualized storage system of the present invention becomes simple and easy, thereby making the reliability and maintainability of the entire system Big Great enhancement.

In order to solve the above problems, the present invention provides a converged virtualized storage subsystem, comprising: at least: a first server node; a second server node; and a shared storage pool, including at least one hard disk cluster, the first The server node and the second server node can access the shared storage pool at the same time; wherein the fusion virtualized storage subsystem is installed with an operating system, and the operating system defines a storage space unit, the first server node and The second server node accesses the at least one hard disk cluster in the shared storage pool through the storage space unit, so as to synchronize the first server node and the second server node, thereby ensuring redundant storage configuration. a file server and a virtual server are disposed in the operating system, and the virtual server is connected to the file server by using an application stacking manner based on the operating system, and the virtual server and the file server are implemented. Quick match.

In an embodiment of the invention, the file server respectively configures the first server node and the second server node, and the virtual server configures the first server node and the second server node respectively.

In an embodiment of the invention, the first server node and the second server node implement simultaneous access to the shared storage pool by using a storage dual domain technology.

In an embodiment of the invention, the first server node and the second server node are multiple.

In an embodiment of the invention, the converged virtualized storage subsystem is disposed in a server.

In an embodiment of the invention, the first server node and the second server node further comprise at least one controller, at least one of the controller of the first server node and At least one of the controllers of the second server node is coupled to the at least one hard disk cluster.

In an embodiment of the invention, the converged virtualization storage subsystem further includes at least one storage expansion module, and each controller is connected to the at least one hard disk cluster by the at least one storage expansion module.

In an embodiment of the present invention, at least one controller of the first server node and at least one controller of the second server node are respectively connected to the same hard disk cluster by the at least one storage expansion module. .

In an embodiment of the invention, each of the storage expansion modules is respectively connected to at least one controller of the first server node and at least one controller of the second server node.

Another object of the present invention is to provide a converged virtualized storage system, comprising: a plurality of the above-described converged virtualized storage subsystems; a switch, and each of the converged virtualized storage subsystems and a plurality of clients Each of the terminals is configured to enable a user to access a corresponding one of the converged virtualized storage subsystems through each of the user terminals.

The present invention has the advantage of utilizing a high density, high frequency wide and redundant design to fuse the connections and configurations between the original three layer architectures in a 4U (1U = 44.45 mm) storage system. Compared with the prior art, the invention has the following beneficial effects: (1) can save more than half of the space; (2) no complicated layer and layer cable connection and topology configuration process; (3) flat integrated structure It can reduce the delay and signal loss between the original 3 layers of architecture, effectively improve the efficiency and stability of the system operation; (4) Through the unified virtualization within the system, the expansion is very simple and convenient.

100‧‧‧Virtualized Storage System

110‧‧‧Storage device

120‧‧‧File Server

130‧‧‧Virtual Server

140‧‧‧Switch

150‧‧‧ Client

200‧‧‧Converged Virtual Storage Subsystem

210‧‧‧First server node

211‧‧‧Virtual Server

212‧‧‧File Server

220‧‧‧Second server node

230‧‧‧Shared storage pool

231‧‧‧ hard disk drive

240‧‧‧Storage space unit

241‧‧‧First Storage Expansion Module

242‧‧‧Second storage expansion module

243‧‧‧ Third Storage Expansion Module

244‧‧‧4th Storage Expansion Module

250‧‧‧ Controller

310‧‧‧hard disk cluster

320‧‧‧Expander

331‧‧‧First IO Module

332‧‧‧Second IO module

341‧‧‧First adapter plate

342‧‧‧Second adapter plate

343‧‧‧hard disk backplane

500‧‧‧Converged Virtual Storage System

510‧‧‧ Client

520‧‧‧Switch

1 is an architectural diagram of a prior art virtualized storage system.

2 is a diagram showing the main architecture of a converged virtualized storage subsystem in accordance with an embodiment of the present invention.

3A is a partial architecture diagram of the converged virtualized storage subsystem shown in FIG. 2, wherein the controller of the dual domain SAS is connected to the hard disk cluster by a storage expansion module of the storage space unit.

FIG. 3B is a schematic structural view of the hard disk cluster shown in FIG. 3A.

FIG. 3C is a schematic diagram of the inside of a 4U cluster server according to an embodiment of the present invention.

4A is a schematic diagram showing a connection relationship between a plurality of controllers and a plurality of memory drive modules and 70 hard disk drives in an embodiment of the present invention.

4B is a schematic diagram showing the connection relationship between two server nodes and a plurality of storage expansion modules in an embodiment of the present invention.

4C is a schematic diagram showing the connection relationship between two server nodes and a plurality of storage expansion modules in an embodiment of the present invention.

4D is a schematic diagram showing another connection relationship between two server nodes through a plurality of storage expansion modules and 70 hard disk drives in an embodiment of the present invention.

FIG. 5 is a block diagram of a converged virtualized storage system in accordance with an embodiment of the present invention.

The specific implementation manner of the converged virtualized storage system provided by the present invention is described in detail below with reference to the accompanying drawings.

2 is a block diagram of a converged virtualized storage subsystem in accordance with an embodiment of the present invention.

Referring to FIG. 2, according to an object of the present invention, the present invention provides a converged virtualized storage subsystem 200, and is disposed in a server (for example, a 4U cluster server). The converged virtualized storage subsystem 200 includes at least a first server node 210, a second server node 220, a storage space unit 240, and a shared storage pool 230. In other embodiments of the present invention, the first server node and the second server node may be multiple.

As shown in FIG. 2, the first server node 210 and the second server node 220 are mutually redundant configuration. That is, any one of the first server node 210 and the second server node 220 can serve as a backup for another server node. The redundant configuration described above can reduce or avoid the occurrence of failure of the converged virtualized storage subsystem 200 due to a single server node failure. Especially for storage systems used in manufacturing, medical, retail and other fields, data availability and security are extremely important for uninterrupted operation. The above-described converged virtualized storage subsystem 200 is provided with an operating system (not shown) in which a storage space unit 240 is defined. The first server node 210 and the second server node 220 respectively access the at least one hard disk cluster 310 in the shared storage pool 230 through the storage space unit 240 to implement the first server node 210 and the second The synchronization of the server node 220, thereby ensuring the configuration of redundant storage. The operating system includes at least a virtual server 211 and a file server 212. The file server 212 can be configured with the first server node 210 and the second server node 220, respectively. The virtual server 211 can also configure the first server node 210 and the second server node 220 respectively. See Figure 2. The virtual server 211 (for example, the virtual server 1 and the virtual server 2 in FIG. 2) is connected to the file server 212 by the application stacking method of the operating system (for example, the file server 1 in FIG. 2). The file server 2) is connected to achieve a quick match between the virtual server 211 and the file server 212. Compared with the prior art, through the connection of the network cable and the configuration topology of the related software, the present invention helps the performance of the storage system by stacking the application without spending too much system execution time, thereby solving the virtual The server 211 accesses the latency and complex matching issues of the file server 212. A plurality of host systems (not shown) disposed in the virtual server 211 can achieve coordinated work, unified management, and uninterrupted maintenance through a virtual layer (not shown), wherein multiple host systems can run the same / Different operating systems, the virtual layer acts as an abstraction layer that decouples the physical hardware from the application system associated with the application for greater resource utilization and flexibility. The virtual layer allows virtual machines or clients with homogeneous or heterogeneous operating systems to run on the same or different physical machines individually but in parallel. Since the server nodes 210, 220 formed by the virtual server 211 and the file server 212 belong to the same layer of flat architecture, the cable between the existing virtual server and the file server can be simplified (for example, a SAS cable or Fiber optic) management, that is, a single configuration is required to access the file server 212 through a plurality of different host systems without physically connecting to the file server 212. In addition, the flattening architecture can provide powerful scalability, so that it is easy to implement adding (or expanding) the virtual server 211 or the file server 212 in the converged virtualized storage subsystem 200, thereby greatly enhancing the entire system. Reliability and maintainability.

3A is a partial architecture diagram of the converged virtualized storage subsystem shown in FIG. 2, wherein the controller of the dual domain SAS is connected to the hard disk cluster through a storage expansion module of the storage space unit. FIG. 3B is a schematic structural view of the hard disk cluster shown in FIG. 3A. FIG. 3C is a schematic diagram of the inside of a 4U cluster server according to an embodiment of the present invention.

As shown in Figures 2 and 3A, the first server node 210 and the second server node 220 each further include at least one controller 250 for processing data signals and control signals and controlling inputs and outputs, each The controller 250 can be connected to the at least one hard disk cluster 310 of the shared storage pool 230 through the storage space unit 240, respectively.

As shown in FIG. 3A, the shared storage pool 230 includes at least one hard disk cluster 310; Referring to further FIGS. 3A and 3B, each hard disk cluster 310 includes a plurality of hard disk drives 231 for the first server node 210 and the second server node 220 to share the stored data. . Each of the hard disk drives 231 can be, but is not limited to, a dual domain SAS drive or a SATA drive.

Referring to FIG. 2, the first server node 210 and the second server node 220 respectively access the shared storage pool 230 by the storage space unit 240 to implement the first server node 210 and the second server node. Synchronization of 220 to ensure redundant configuration. As shown in FIG. 3A, the storage space unit 240 further includes a first storage expansion module 241 and a second storage expansion module 242. The first storage expansion module 241 and the second storage expansion module 242 are the same storage expansion module. Grouping, so that at least one of the controller 250 of the first server node 210 and at least one of the controllers 250 of the second server node 220 respectively pass the first storage expansion module 241 and the second of the storage space unit 240 The storage expansion module 242 is connected to the same hard disk cluster 310 in the shared storage pool 230. The first storage expansion module 241 and the second storage expansion module 242 are configured to transfer commands and data sent by the controller 250 to a hard disk drive 231 (such as a SAS drive or a SATA drive) of a hard disk cluster 310.

Referring to FIG. 3A and FIG. 3B, the hard disk cluster 310 of the shared storage pool 230 further includes a first IO module 331 and a second IO module 332. The first IO module 331 and the second IO module 332 each include a Expanders 320 (expanders), each of the expanders 320 are connected to the plurality of hard disk drives 231, and the two expanders 320 are respectively connected to the first server node 210 by their corresponding first storage expansion modules 241 and 210 The controller 250 is connected and connected to the controller 250 of the second server node 220 by its corresponding second storage expansion module 242. The expander 320 can adopt three routing modes: direct routing, routing table, and Subtractive routing is performed to implement mutual routing between the controller 250, the storage expansion module, the expander 320, and the hard disk drive 231. The expander 320 is an optional component. In other implementations of the present invention, the first storage expansion module 241 and the second storage expansion module 242 may be directly connected to the hard disk drive 231.

In an embodiment of the present invention, at least one controller 250 of the first server node 210 and at least one controller 250 of the second server node 220 as shown in FIGS. 3A and 3B utilize dual domain SAS (Dual Domain SAS) (SAS is short for Serial Attached SCSI) technology, the first storage expansion module 241 and the second storage expansion module 242 of the storage space unit 240 are used to simultaneously access the same hard disk drive 231. In other embodiments, the controller 250 can be replaced with a Host Bus Adapter (HBA), which can also access the same hard disk drive 231 simultaneously using dual domain SAS technology. Since the converged virtualized storage subsystem 200 implements redundant configuration by employing dual domain SAS technology, it can tolerate failure of the main bus adapter, external cable failure, expanders failure, and hard disk. Cluster (JBOD) failure, this is extremely important. The dual-domain SAS technology differs from the prior art dual-path SATA (Dual Path SATA) technology solution, which uses only a single domain to allow cable failures, while the former provides complete redundant dual-domain SAS. . In addition, dual-domain SAS provides higher reliability, higher performance, and data availability than dual-path SATA. It is important to note that the main difference between dual-domain SAS and dual-path SATA is whether to use a SAS drive or a SATA drive. Since the dual-port hard disk drive provides a full-range redundant configuration in a dual-domain SAS, and the SAS drive is a dual-port hard disk drive, a SAS drive should be used in the dual-domain SAS. In the embodiment of the present invention, the hard disk drive 231 is a SAS drive, and the SAS drive has dual ports (Dual Ports) satisfying the above conditions.

When the hard disk drive 231 is a dual-port hard disk drive and the first IO module 331 and the second IO module 332 are redundantly configured with each other, the dual domain SAS is made to satisfy the condition of redundant design. Therefore, the dual-domain SAS in this embodiment can allow one of the two-port type hard disk drive to fail, or allow any one of the first IO module 331 and the second IO module 332 to occur. The failure, or one of the switches of the controller 250 that supports the dual domain SAS, fails, or allows the external cable to fail, but does not affect the normal operation of the entire virtualized storage subsystem 200.

In addition, at least one controller 250 of the first server node 210 and at least one controller 250 of the second server node 220 of FIGS. 3A and 3B may use an SSP protocol and pass through a storage expansion module (ie, a first storage expansion module). The group 241, the second storage expansion module 242) and the expander 320 are connected to a hard disk drive 231 belonging to the SAS drive. Of course, the controller 250 can also be connected to the hard disk drive 231 belonging to the SATA drive through the storage expansion module and expander 320 using the STP protocol. In an embodiment of the invention, the controller 250 is coupled to a hard disk drive 231 belonging to the SAS drive by a storage expansion module and expander 320 using an SSP protocol.

Referring to FIGS. 3A and 3C, each of the controllers 250 may be specifically disposed on a first interposer 341 (see FIG. 3C, reference numeral 341). For example, the first interposer 341 is an Interposer Board for processing. The data signal and the control signal, and the control input and output, the first storage expansion module 241 and the second storage expansion module 242 of the storage space unit 240 may be specifically disposed in a second adapter plate 342 (see FIG. 3C). Reference numeral 342), a second riser board 342 (also referred to as MID Plane) is respectively connected to the first riser board 341 and the hard disk back board 343 for transferring data signals and control signals.

4A is a diagram showing multiple controllers expanding through multiple storage expansion modules in accordance with an embodiment of the present invention; Schematic diagram of the connection relationship between a group and 70 hard disk drives. Referring to FIG. 4A, the converged virtualized storage subsystem includes four controllers 250, each of which has eight ports, each of which has a storage expansion mode corresponding to one of the storage space units 240. The storage expansion modules are respectively connected to the hard disk drive 231 of the corresponding hard disk cluster 310 in the shared storage pool 230. As shown in FIG. 4A, the first storage expansion module 241 is respectively associated with the first group of hard disk drives 231 (0~9 hard disk drives), the second set of hard disk drives 231 (10~19 hard disk drives), and Three sets of hard disk drives 231 (20-29 hard disk drives) and a fourth set of hard disk drives 231 (30-39 hard disk drives) are connected. The third storage expansion module 243 is respectively connected to the first group of hard disk drives 231 (0~9 hard disk drives), the second set of hard disk drives 231 (10~19 hard disk drives), and the third set of hard disk drives. The 231 (20~29 hard disk drive) and the fourth group of hard disk drives 231 (30~39 hard disk drives) are connected. The second storage expansion module 242 and the fourth group of hard disk drives 231 (30~39 hard disk drives), the fifth group of hard disk drives 231 (40~49 hard disk drives), and the sixth group of hard disk drives The 231 (50~59 hard disk drive) and the seventh group of hard disk drive 231 (60~69 hard disk drive) are connected. The fourth storage expansion module 244 and the fourth group of hard disk drives 231 (30~39 hard disk drives), the fifth group of hard disk drives 231 (40~49 hard disk drives), and the sixth group of hard disk drives The 231 (50~59 hard disk drive) and the seventh group of hard disk drive 231 (60~69 hard disk drive) are connected.

4B is a schematic diagram showing the connection relationship between two server nodes and a plurality of storage expansion modules in an embodiment of the present invention. As shown in FIG. 4B, in the embodiment, a controller 250 of the first server node 210 and a controller 250 of the second server node 220 respectively pass through two storage expansion modules of the storage space unit 240. Connected to the same hard disk cluster 310 in the shared storage pool 230. It should be noted that in other embodiments, the controller in the first server node 210 The number of 250s and the number of controllers 250 in the second server node 220 are not limited to one, and may be plural. The first server node 210 is directly connected to the first storage expansion module 241 and the third storage expansion module 243 by the controller 250, and the second server node 220 is respectively connected to the second storage expansion module by the controller 250. 242 and the fourth storage expansion module 244 are directly connected. The first storage expansion module 241 and the second storage expansion module 242 are connected in common to the same hard disk cluster 310. The third storage expansion module 243 and the fourth storage expansion module 244 are connected to the same another. Hard disk cluster 310. It can be seen from the above configuration that the first storage expansion module 241 and the second storage expansion module 242 are redundant with each other. The third storage expansion module 243 and the fourth storage expansion module 244 are redundant with each other. Thus, the first server node 210 and the second server node 220, with the redundancy settings described above, can reduce or avoid the occurrence of failure of the converged virtualized storage subsystem 200 due to a single server node failure. In addition, as can be seen from the foregoing, the first server node 210 and the second server node 220 can simultaneously access multiple hard disk drives 231 in the same hard disk cluster 310 by using the dual domain technology, that is, the first servo. The second storage expansion module 242 can simultaneously access the plurality of hard disk drives 231 in the same one of the hard disk clusters 310 by the first storage expansion module 241 and the second server node 220. Similarly, the first server node 210 can also simultaneously access multiple of the same other hard disk clusters 310 by the fourth storage expansion module 244 by using the third storage expansion module 243 and the second server node 220. Hard disk drive 231.

4C is a schematic diagram showing the connection relationship between two server nodes and a plurality of storage expansion modules in an embodiment of the present invention. As shown in FIG. 4C, in the embodiment, each of the storage expansion modules 241 is respectively connected to at least one controller 250 of the first server node 210 and at least one controller 250 of the second server node 220. . The first server node 210 and the second server node 220 are respectively connected to the plurality of storage expansion modules through the controller 250 to connect to more The hard disk drive of the plurality of hard disk clusters 310 in the shared storage pool 230 can be connected to the corresponding one of the first server node 210 and the second server node 220 to create a larger-scale SAS domain. In this embodiment, the first server node 210 includes a controller 250, and the second server node 220 includes a controller 250, a controller 250 of the first server node 210, and a second server node 220 The controllers 250 each have eight HD SAS ports. The controller 250 of the first server node 210 is connected to the first storage expansion module 241 and the second storage expansion module 242 disposed on the second riser board 342 in a double solid mode (as a redundant configuration). . The controller 250 of the first server node 210 is connected to the third storage expansion module 243 and the fourth storage expansion module 244 disposed on the second riser board 342 in a double dashed manner (as a redundant configuration). Similarly, the controller 250 of the second server node 220 is connected to the third storage expansion module 243 and the fourth storage expansion module 244 disposed on the second riser board 342 in a double dashed manner. The controller 250 of the second server node 220 is connected to the first storage expansion module 241 and the second storage expansion module 242 disposed on the second riser board 342 in a double solid line manner. It can be seen that the first storage expansion module 241 and the second storage expansion module 242 are redundant with each other, and the third storage expansion module 243 and the fourth storage expansion module 244 are redundant with each other, thereby reducing Or avoiding the situation where the converged virtualized storage subsystem is not functioning properly due to a single controller 250 failure.

4D is a schematic diagram showing another connection relationship between two server nodes through a plurality of storage expansion modules and 70 hard disk drives in an embodiment of the present invention. Referring to FIG. 4D, the first server node 210 and the second server node 220 in the fused virtualization storage subsystem 200 are respectively connected to the second riser board 342 through the controller 250 (see the label in FIG. 3C). The first storage expansion module 241, the second storage expansion module 242, the third storage expansion module 243, and the fourth storage expansion module 244 on the 342). The first server node 210 and the second adapter board 342 respectively A storage expansion module 241, a second storage expansion module 242, a third storage expansion module 243, and a fourth storage expansion module 244 are connected in the same manner as in FIG. 4C. Similarly, the second server node 220 is also respectively connected to the first storage expansion module 241, the second storage expansion module 242, the third storage expansion module 243, and the fourth storage expansion module on the second adapter board 342. The connection 244 is connected in the same manner as in Fig. 4C. The first storage expansion module 241, the second storage expansion module 242, the third storage expansion module 243, and the fourth storage expansion module 244 are respectively connected to the hard disks of the plurality of hard disk clusters in the shared storage pool 230. The driver 231, for example, as shown in FIG. 4D, the first storage expansion module 241 is respectively hard with the first group of hard disk drives 231 (0-9 hard disk drives) and the second group of hard disk drives 231 (10-19) The disc drive), the third group of hard disk drives 231 (20-29 hard disk drives) and the fourth group of hard disk drives 231 (30~39 hard disk drives) are connected. The third storage expansion module 243 is respectively connected to the first group of hard disk drives 231 (0~9 hard disk drives), the second set of hard disk drives 231 (10~19 hard disk drives), and the third set of hard disk drives. The 231 (20~29 hard disk drive) and the fourth group of hard disk drives 231 (30~39 hard disk drives) are connected. The second storage expansion module 242 and the fourth group of hard disk drives 231 (30~39 hard disk drives), the fifth group of hard disk drives 231 (40~49 hard disk drives), and the sixth group of hard disk drives The 231 (50~59 hard disk drive) and the seventh group of hard disk drive 231 (60~69 hard disk drive) are connected. The fourth storage expansion module 244 and the fourth group of hard disk drives 231 (30~39 hard disk drives), the fifth group of hard disk drives 231 (40~49 hard disk drives), and the sixth group of hard disk drives The 231 (50~59 hard disk drive) and the seventh group of hard disk drive 231 (60~69 hard disk drive) are connected. Referring to FIG. 4C, the first controller 250 and the second controller 250 of the first server node 210 and the second server node 220 are redundantly arranged with each other, and the first controller 250 and The second controllers 250 are cluster interconnects with each other so that when the controller 250 or the cable fails For redundancy or redundancy.

In the embodiment of the present invention, the number of the plurality of hard disk drives 231 in the shared storage pool 230 is 70. Since the fusion virtualized storage subsystem 200 of the present invention adopts the Dual Domain SAS mode, the first server node 210 and the second server node 220 can simultaneously access all 70 hard disk drives 231. The first server node 210 and the second server node 220 respectively access the 70 hard disk drives 231 of the shared storage pool 230 through the storage expansion module of the storage space unit 240 and utilize the storage dual domain technology (here) The hard disk drive 231 is a SAS drive), thereby implementing synchronization of the first server node 210 and the second server node 220, and ensuring redundant storage configuration and symmetric path, thus serving as an enterprise-class converged virtualized storage sub-server. System 200 is able to guarantee the availability and security of its data.

Referring to FIG. 5, in accordance with another aspect of the present invention, the present invention further provides a converged virtualized storage system 500, including: a plurality of the above-described converged virtualized storage subsystems 200; and a switch 520, respectively The virtualized storage subsystem 200 is connected to a plurality of clients 510. Each of the clients 510 is configured to enable a user to access a corresponding virtualized storage subsystem 200 through each of the clients 510.

Since each of the converged virtualized storage subsystems 200 fuses the virtual server 211, the file server 212, and the storage device (for example, a hard disk drive belonging to a SAS drive), all connections and configurations are uniformly integrated at design time, thereby Can provide users with a complete solution. More importantly, the present invention combines the prior art three-layer architecture into a layered architecture of the converged virtualized storage subsystem 200, and utilizes a virtualization manner to flatten the configuration of the entire converged virtualized storage system 500. Omit the complex cable configuration between traditional three-tier architectures.

Through the use of multiple converged virtualized storage subsystems 200, multiple client terminals 510 (eg, PC terminals), and switches 520, the user can conveniently access the desired converged virtualized storage subsystem 200 through the client 510. The data in the hard disk drive 231. In addition, according to the above, the stacking technology between the file server 212 and the virtual server 211 and the storage dual-domain SAS technology make the cluster expansion of the entire converged virtual storage system 500 easier, thereby enabling the entire fusion. The reliability and maintainability of the virtualized storage system 500 is greatly enhanced.

It can be seen from the above that the fusion virtualized storage system 500 of the present invention utilizes a high density, high frequency width and redundancy design to fuse the original three in a 4U (1U=44.45 mm) storage system compared to the prior art storage system. Connections and configurations between layer architectures. By flattening the existing three-layer architecture into a layer to embody the high density of the architecture of the present invention, and the plurality of storage expansion modules merging the storage space unit 240 become the interface of the same shared storage pool 230, the efficient use thereof can Achieve high frequency width. That is, the present invention combines the prior art three-layer architecture into one layer architecture, thereby achieving more than half of the space, without complicated layer-to-layer cable connection and topology configuration flow, and utilizing a flat integrated structure. It can reduce the delay and signal loss between the original three-tier architecture, effectively improving the efficiency and stability of the system operation. Furthermore, through the unified virtualization within the system, it is very simple and convenient to expand and expand the cluster, thereby enhancing the reliability and maintainability of the entire system.

The above is only a preferred embodiment of the present invention, and it should be noted that those skilled in the art can also make several improvements and refinements without departing from the principles of the present invention. These improvements and retouchings should also be considered as The scope of protection of the present invention.

200‧‧‧Converged Virtual Storage Subsystem

210‧‧‧First server node

211‧‧‧Virtual Server

212‧‧‧File Server

220‧‧‧Second server node

230‧‧‧Shared storage pool

240‧‧‧Storage space unit

Claims (10)

A converged virtualized storage subsystem includes at least: a first server node; a second server node; and a shared storage pool including at least one hard disk cluster, allowing the first server node and the second The server node accesses the shared storage pool at the same time; wherein the fusion virtualized storage subsystem is installed with an operating system, and a storage space unit is defined in the operating system, and the first server node and the second server node respectively borrow Accessing the at least one hard disk cluster in the shared storage pool by the storage space unit to synchronize the first server node and the second server node, thereby ensuring redundant storage configuration, and setting a The file server and a virtual server are connected to the file server by means of an application stacking manner based on the operating system, so that the virtual server and the file server are quickly matched. According to the fusion virtualized storage subsystem of claim 1, wherein the file server respectively configures the first server node and the second server node, and the virtual server configures the first server node and The second server node. The converged virtualized storage subsystem of claim 1, wherein the first server node and the second server node simultaneously access the shared storage pool by using a storage dual domain technology. The converged virtualized storage subsystem of claim 1, wherein the first server node and the second server node are plural. The converged virtualized storage subsystem of claim 1 wherein the converged virtualized storage subsystem is disposed in a server. The fusion virtualized storage subsystem of claim 1, wherein the first server node and the second server node further comprise at least one controller, the at least one controller of the first server node and The at least one controller of the second server node is coupled to the at least one hard disk cluster. The converged virtualization storage subsystem of claim 6 further includes at least one storage expansion module, each controller being connected to the at least one hard disk cluster by the at least one storage expansion module. The fusion virtualization storage subsystem of claim 7, wherein the at least one controller of the first server node and the at least one controller of the second server node respectively use the at least one storage expansion mode The group is connected to the same hard disk cluster. A converged virtualized storage subsystem according to claim 8 wherein each of the storage expansion modules is respectively connected to at least one controller of the first server node and at least one controller of the second server node . A converged virtualized storage system includes: a plurality of converged virtualized storage subsystems according to claim 1; a switch respectively connected to each of the converged virtualized storage subsystems and a plurality of clients; as well as Each of the clients is configured to allow a user to access the corresponding converged virtualized storage subsystem by each of the clients.