Cluster Server

Abstract

The present application relates to a cluster server, comprising: a switch and at least three servers, the servers are connected to the switch; the servers include storage devices, the storage devices include hard disk controllers and disk arrays, each hard disk controller is connected to the disk array of at least one other server through a disk connector; at least three servers include a master server, the master server is used to control each server to obtain or release the control right of the disk array of the current server and / or the disk array of at least one other server. Through the present application, the problem of low storage resource utilization of cluster servers in related technologies is solved, and the storage resource utilization of cluster servers is improved.

Description

Technical Field

[0001] The present application relates to the field of server clusters, and in particular to a cluster server. Background Art

[0002] Server cluster refers to the collection of many servers to provide the same service together. To the client, it looks like there is only one server. The cluster can use multiple computers for parallel computing to achieve high computing speed, and can also use multiple computers for backup, so that if any machine fails, the entire system can still operate normally.

[0003] Existing cluster servers can usually only achieve clustering at the software system level. That is, when a server fails, the applications running on this server will be switched to other servers, and the hard disk resources on the failed machine will no longer be used. The storage link transmitted to this server will also be cut off, and the storage content on this server will not be obtained, resulting in insufficient utilization of storage resources. Summary of the invention

[0004] In this embodiment, a cluster server is provided to solve the problem of low storage resource utilization of the cluster server in the related art.

[0005] In this embodiment, a cluster server is provided, including: a switch and at least three servers, the servers are connected to the switch, the servers include storage devices, the storage devices include a hard disk controller and a disk array, the hard disk controller is connected to the disk array of at least one other server through a disk connector; the hard disk controller is used to obtain or release control of the disk array of the current server and / or the disk array of at least one other server.

[0006] In some of the embodiments, each of the hard disk controllers is connected to the disk array of the current server and the disk arrays of two other servers via a disk connector, and the storage devices of each of the servers are connected in a ring topology.

[0007] In some of the embodiments, the hard disk controller is used to monitor the operating status of the disk array of the current server, and release the control of the disk array of the current server when the operating status of the current server is abnormal; and monitor the operating status of the disk arrays of other servers, and obtain the control of the disk arrays of other servers when the operating status of other servers is abnormal.

[0008] In some embodiments, the abnormal operation status includes: no read or write operation on the disk array within a preset time interval.

[0009] In some of the embodiments, the storage device of the server is powered by a power supply independent of the computing part of the server; the abnormal operation state includes: detecting that the computing part of the server is powered off.

[0010] The server further comprises a baseboard management controller, the baseboard management controller is connected to the switch, and the baseboard management controller is also connected to a hard disk controller of the current server;

[0011] The hard disk controller is also used to monitor the operating status of the disk array of the current server, and release the control right of the disk array of the current server when the operating status of the current server is abnormal, and notify other servers of the abnormal operating status through the baseboard management controller;

[0012] The hard disk controller is also used to receive abnormal operation status notifications of other servers with abnormal operation status from the baseboard management controller of the current server, and obtain control rights over the disk arrays of other servers with abnormal operation status.

[0013] In some of these embodiments,

[0014] The baseboard management controller is also used to perform self-check and repair on the hardware of the current server after the control of the disk array of the current server is transferred to other servers;

[0015] The hard disk controller is also used to regain control of the disk array of the current server when the current server is in normal operation.

[0016] In some of the embodiments, the storage device of each server is powered by a power supply independent of the computing part of the server, and the baseboard management controller performs self-check and repair by restarting the current server.

[0017] In some embodiments, the disk connector is a Serial Attached Small Computer System Interface (Serial Attached Small Computer System Interface) connector.

[0018] In some of the embodiments, the storage device of each server is physically centralized in the server.

[0019] Compared with the related art, the cluster server provided in this embodiment includes: a switch and at least three servers, the server is connected to the switch; the server includes a storage device, the storage device includes a hard disk controller and a disk array, each hard disk controller is connected to the disk array of at least one other server through a disk connector; at least three servers include a main server, the main server is used to control each server to obtain or release the control of the disk array of the current server and / or the disk array of at least one other server, which solves the problem of low storage resource utilization of the cluster server in the related art and improves the storage resource utilization of the cluster server.

[0020] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0022] Figure 1 is a schematic diagram of a server of this embodiment.

[0023] Figure 2 It is a schematic diagram of the structure of the cluster server provided in this embodiment.

[0024] Figure 3 Schematic diagram of the linear topology structure of this embodiment.

[0025] Figure 4 Schematic diagram of the ring topology structure of this embodiment. DETAILED DESCRIPTION

[0026] In order to more clearly understand the purpose, technical solutions and advantages of the present application, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments.

[0027] Unless otherwise defined, the technical terms or scientific terms involved in this application shall have the general meaning understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the", "these" and the like in this application do not represent quantitative restrictions, and they can be singular or plural. The terms "include", "comprise", "have" and any variants thereof involved in this application are intended to cover non-exclusive inclusions; for example, a process, method and system, product or device comprising a series of steps or modules (units) is not limited to the listed steps or modules (units), but may include unlisted steps or modules (units), or may include other steps or modules (units) inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether directly or indirectly. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there can be three relationships. For example, "A and / or B" can mean: A exists alone, A and B exist at the same time, and B exists alone. Usually, the character " / " indicates that the objects associated with each other are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0028] This embodiment provides a cluster server, which includes at least three cluster servers. Figure 1 Schematic diagram of the server of this embodiment, the server can also be called a host, such as Figure 1 As shown, each server includes a computing part 10 and a storage part 20. The computing part 10 generally includes a central processing unit 110 (CPU, also known as a main controller or a main control); the storage part generally consists of a storage device 210.

[0029] The storage device 210 includes a hard disk controller 212 and a disk array 211. It should be noted that the disk array 211 referred to in this embodiment may include only one disk drive, or may be a disk group composed of multiple disk drives. Furthermore, the disk drives constituting the disk array are not limited to HDD disk drives or SDD disk drives, and may also be a combination of HDD disk drives and SDD disk drives in some embodiments. Furthermore, the disk array 211 may be a large-capacity disk drive formed by connecting all disk drives in series using the JBOD (Just a Bunch Of Disks) technology, or may be used by a server using technologies such as RAID (Redundant Array of Independent Disks), thereby improving the disk fault tolerance.

[0030] The interface device between the computing part 10 and the disk array 211 is called a hard disk controller 212, also known as a disk drive adapter. The hard disk controller 212 is used to interpret the commands given by the computing part 10 at the software level and send various control signals to the disk drive; the hard disk controller 212 is also used to detect the status of the disk drive, or write and read data from the disk according to the specified disk data format. From the hardware level, the hard disk controller 212 provides one or more physical interfaces for connecting to the disk array 211. The hard disk controller 212 can connect to one or more disk arrays 211 through these physical interfaces, and obtain or release the control of the disk array 211 connected to the physical interface.

[0031] Each disk array 211 may include one or more physical interfaces for connecting to a hard disk controller 212. For example, a disk array 211 based on SAS (Serial Attached Small Computer System Interface) technology may be connected to hard disk controllers 212 of multiple servers, so that multiple servers share the same disk array 211.

[0032] The computing part 10 and the storage part 20 of each server can be physically centralized, for example, in the same server chassis. The computing part 10 and the storage part 20 can be arranged on the same main circuit board, or separately. For example, the storage part 20 is arranged on the server backplane, and the computing part 10 is arranged on the main circuit board.

[0033] In addition to the storage part 20 and the computing part 10, the server usually has two core firmwares, namely BIOS (basic input and output system) (not shown) and BMC (baseboard management controller) (not shown). Among them, in the computer system, BIOS has a more basic and fundamental role than the server's operating system. It is mainly responsible for detecting, accessing and debugging the underlying hardware resources, and allocating them to the operating system to ensure the smooth and safe operation of the entire system. BMC is a small operating system independent of the server's operating system, usually integrated on the motherboard, or plugged into the motherboard through PCIe or other forms. BMC's external manifestation is usually a standard RJ45 network port, and BMC has an independent IP firmware system. Usually, the server can use BMC instructions for unattended operations, such as remote management, monitoring, installation, and restart of the server.

[0034] Figure 2 Schematic diagram of the structure of the cluster server provided in this embodiment. Figure 2 Five servers are used as an example for illustrative description. In other embodiments, the number of servers can be any number greater than three, which is usually set according to the demand for computing resources and storage resources of the cluster server. The number is not limited in this embodiment.

[0035] like Figure 2 The cluster server shown includes a switch 40 and five servers. Each server is connected to the switch 40. Each server in the cluster server exchanges service data through the switch 40.

[0036] The hard disk controller 212 of each server is connected to the disk array 211 of the current server and the disk array 211 of at least one other server through a disk connector (such as a SAS connector). The other servers refer to other servers in the cluster server except the current server.

[0037] In the five servers, the hard disk controller 212 of each server is used to obtain or release the control right of the disk array 211 of the current server and / or the disk array 211 of at least one other server.

[0038] The cluster server provided in this embodiment realizes the control of the disk array through the hard disk controller of each server. The hard disk controller of each server is connected to the disk array of the current server and the disk array of at least one other server through the disk connector. The control of the disk array of the current server and / or the disk array of at least one other server is obtained or released through the respective hard disk controllers, so that in the case of a failure of the current server, the control of the disk array of the current server can be transferred to other servers, or in the case of a failure of a certain other server, the control of the disk array of the other recovery period of the failure is obtained, thereby improving the utilization rate of the disk array 211. Compared with the related art that uses expensive SAS switches to realize the sharing of the disk array 211, this embodiment does not need to add any additional SAS switches, but can directly use the switch 40 used for business processing of the cluster server to meet the demand, greatly reducing the cost.

[0039] In addition, in the above embodiment, each hard disk controller controls the control right of the disk array respectively, and the hard disk controllers of the five servers are not divided into primary and secondary. This avoids the problem that when a certain server or hard disk controller is used as the primary control right of the disk array, the control right of the disk array cannot be transferred normally due to the failure of the main server or the main hard disk controller, thereby improving the robustness of the system.

[0040] In order to avoid the cost increase caused by using SAS switches to interconnect all disk arrays 211 in the cluster server, in this embodiment, each hard disk controller 212 connects the disk array 211 of the current server and the disk array 211 of at least one other server through a disk connector (SAS connector). Through such connection, the storage devices of each server can form a network such as Figure 3The linear topology shown in the figure. In a linear topology, when a server at both ends of the topology fails, the storage device can only be taken over by an adjacent server. If the computing load of this adjacent server is large, the adjacent server may fail due to the further increase in load after taking over the storage device, resulting in reduced stability of the cluster server. Or if two consecutive adjacent servers at both ends of the topology fail, the storage device of the outermost server cannot be taken over by any server. This shows that there is still room for improvement in the utilization of storage devices.

[0041] To this end, in some embodiments, each hard disk controller 212 connects the disk array 211 of the current server and the disk arrays of two other servers through a disk connector (SAS connector), and the storage devices of each server form a storage device such as Figure 4 The ring topology shown in the figure. This connection method ensures that if any server fails, there are two adjacent servers that can take over the storage device of the failed server; even if two consecutive adjacent servers fail, it can be guaranteed that one server will take over the disk array of the two failed servers; only if three consecutive adjacent servers fail, the storage device of one server cannot be taken over by any server. It can be seen that the use of a ring topology improves the stability of cluster servers and the utilization rate of storage devices.

[0042] The working process of the cluster server of this embodiment is described below.

[0043] Example 1

[0044] In this embodiment, the hard disk controller of each server directly obtains or releases the control right of the disk array of the current server and / or other servers.

[0045] Reference Figure 4 The working process of the cluster server provided in this embodiment includes the following steps:

[0046] Step 1: The hard disk controller of server A monitors the running status of the disk array of server A.

[0047] Step 2: When the disk array of server A is in an abnormal operating state, server A releases the control right over the disk array 211 of server A.

[0048] Step 3, the hard disk controllers of server B and server E continuously monitor the operating status of the disk array of server A connected thereto, and acquire control over the disk array 211 of server A when detecting that the operating status of the disk array of server A connected thereto is abnormal.

[0049] The abnormal operation status may be, but is not limited to: no read or write operation on the disk array within a preset time interval. Generally speaking, the computing part of the server will perform read and write operations on the disk array when processing business data, and each process in the server's operating system will also frequently perform read and write operations on the disk array. Therefore, if there is no read or write operation on the disk array within a preset time interval, it can be considered that the operation status of the server is abnormal, such as the physical interface between the hard disk controller of the current server and the current disk array is loose, the server's operating system crashes abnormally, etc.

[0050] In some embodiments, the storage device 210 of each server is powered by an independent power supply relative to the computing part 10 of the server to ensure that the storage device 210 can continue to work when the computing part 10 of the server loses power. When the hard disk controller 212 detects that the computing part 10 of the server loses power, it can also be considered that the current server is in an abnormal operating state.

[0051] Through this embodiment, the hard disk controllers between different servers do not need to exchange information, but monitor the operating status of the disk array directly connected to the physical interface of the hard disk controller to achieve the acquisition and release of the control right of the disk array 211. For a certain server, there may be one or more other servers (referred to as adjacent servers) that have a connection relationship with its hard disk controller.

[0052] For example, in a ring topology, each server has two adjacent servers, and the hard disk controllers of the two adjacent servers can take over the control of the disk array 211 of the corresponding failed server, or one of the hard disk controllers can take over the control of the disk array of the failed server. When the disk array does not support being controlled by two hard disk controllers at the same time, the hard disk controllers of the two adjacent servers can obtain the control of the disk array of the failed server by competing.

[0053] Example 2

[0054] In this embodiment, when the central processing unit of each server can still work normally, the hard disk controller of each server obtains or releases the control right of the disk array of the current server and / or other servers through the central processing unit.

[0055] Reference Figure 4 The working process of the cluster server provided in this embodiment includes the following steps:

[0056] Step 1: The hard disk controller of server A monitors the running status of the disk array of server A.

[0057] Step 2: When server A detects that the operation status of the disk array of server A is abnormal, server A releases the control right to the disk array of server A.

[0058] Step 3: If the CPU of server A can still work normally, the hard disk controller of server A notifies server B and server E of the abnormal operation status through the CPU of server A.

[0059] Step 4: After receiving the abnormal operation status notification, the central processing units of server B and server E send the abnormal operation status notification to their hard disk controllers.

[0060] Step 5: The hard disk controllers of server B and server E obtain control of the disk array of server A according to the abnormal operation status notification.

[0061] The abnormal operation status notification carries identification information of the server with the abnormal operation status, or carries identification information of the disk array of the server with the abnormal operation status.

[0062] A mapping table of the physical interface of the hard disk controller 212 and the disk array 211 can be maintained in the hard disk controller to obtain the identification information of the disk array 211 connected to each physical interface, or the identification information of the server to which the disk array 211 belongs. After the hard disk controller of the current server obtains the abnormal operation status notification sent by the faulty server, the physical interface to which the disk array 211 to be taken over is connected is determined according to the identification information carried in the abnormal operation status notification, and then the control right of the disk array 211 of other servers connected to the physical interface is obtained.

[0063] It should be noted that, in the case of using a disk array 211 using SAS technology, the step of releasing the disk array of the failed server may be unnecessary, for example, the above-mentioned step 2 may be omitted.

[0064] Example 3

[0065] As an independent third party in the server, BMC can monitor the hardware information of the entire server, such as system temperature, power supply voltage, fan speed, etc., and can also monitor the working status of the system network module, user interaction module (such as USB module, display module) or other modules. In particular, BMC can achieve interconnection through the RJ45 network interface and can exchange information with the hard disk controller 212 of the current server.

[0066] In this embodiment, when the central processing unit of each server cannot work normally, the hard disk controller of each server can obtain or release the control right of the disk array of the current server and / or other servers through the BMC.

[0067] Reference Figure 4 The working process of the cluster server provided in this embodiment includes the following steps:

[0068] Step 1: The hard disk controller of server A monitors the running status of the disk array of server A.

[0069] Step 2: When the running state of server A is abnormal, server A releases the control right of the disk array of server A.

[0070] Step 3: The hard disk controller of server A notifies the BMCs of server B and server E of the abnormal operation status through the BMC of server A.

[0071] Step 4: After receiving the abnormal operation status notification, the BMCs of server B and server E send the abnormal operation status notification to their hard disk controllers.

[0072] Step 5: The hard disk controllers of server B and server E obtain control of the disk array of server A according to the abnormal operation status notification.

[0073] Among them, the hard disk controllers 212 of server B and server E can jointly obtain control of the disk array 211 of server A; server B and server E can also determine the hard disk controller 212 of one of the servers to obtain control of the disk array 211 of server A through a negotiation mechanism or a competition mechanism.

[0074] It should be noted that, in the case of using a disk array 211 using SAS technology, the step of releasing the disk array of the failed server may be unnecessary, for example, the above-mentioned steps 2 and 3 may be omitted.

[0075] The abnormal operation status notification carries identification information of the server with the abnormal operation status, or carries identification information of the disk array of the server with the abnormal operation status.

[0076] A mapping table of the physical interface of the hard disk controller 212 and the disk array 211 can be maintained in the hard disk controller to obtain the identification information of the disk array 211 connected to each physical interface, or the identification information of the server to which the disk array 211 belongs. After the hard disk controller of the current server obtains the abnormal operation status notification sent by the faulty server, the physical interface to which the disk array 211 to be taken over is connected is determined according to the identification information carried in the abnormal operation status notification, and then the control right of the disk array 211 of other servers connected to the physical interface is obtained.

[0077] In addition, in this embodiment, after the control right of the disk array of the current server is transferred to other servers, the BMC can also perform self-check and repair on the hardware of the current server. After the self-check and repair of the current server is successful, the hard disk controller of the current server can regain the control right of the disk array of the current server.

[0078] For example, when the hard disk controller of the current server regains control of the disk array 211 of the current server, a normal operation status notification may be sent to the server currently taking over the disk array 211 through the BMC. After receiving the normal operation status notification, the server currently taking over the disk array 211 releases the control right of the corresponding disk array according to the identification information carried in the normal operation status notification, and returns a confirmation message to the current server through the BMC. After receiving the confirmation message, the current server regains control of the disk array of the current server.

[0079] The above method realizes self-checking and self-repair of the faulty server.

[0080] Among them, the disk array 211 of each server is powered by a power supply independent of the computing part 10. The server can restart the current server through the central processing unit or BMC to perform self-check and repair, and ensure that the disk array 211 of the current server is not powered off and can be taken over and used by other servers.

[0081] In some embodiments, the cluster server may further include a control node, which is connected to the switch 40 and is used to configure each server, such as configuring a control program of each server, or identification information of each server, or a mapping table stored in each server. In addition, the BMC of each server may be controlled by the control node to implement remote unattended functions, such as remote restart.

[0082] In summary, traditional cluster services usually treat abnormal nodes by disconnecting the node service and making it impossible to call the storage part. This embodiment realizes the completion of cluster services from the hardware aspect, effectively utilizing the storage part of the abnormal device for multiplexing and retrieving the content of the storage part. This embodiment uses a disk connector to interconnect the disk arrays of multiple servers, so that the storage parts of multiple servers become a whole that can transfer control rights. The hard disk controller of each server participates in cluster control separately, which greatly improves the stability and security of the cluster solution. Once an abnormality occurs, a quick decision can be made to transfer control rights of the disk array, which greatly improves the stability of the cluster solution.

[0083] It should be understood that the specific embodiments described herein are only used to explain the application, rather than to limit it. Based on the embodiments provided in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the protection scope of this application.

[0084] Obviously, the drawings are only some examples or embodiments of the present application. For ordinary technicians in the field, the present application can also be applied to other similar situations based on these drawings without creative work. In addition, it is understandable that although the work done in this development process may be complicated and lengthy, for ordinary technicians in the field, certain changes in design, manufacturing or production based on the technical content disclosed in this application are only conventional technical means and should not be regarded as insufficient content disclosed in this application.

[0085] The term "embodiment" in this application refers to a specific feature, structure or characteristic described in conjunction with the embodiment that can be included in at least one embodiment of the present application. The appearance of this phrase in various locations in the specification does not necessarily mean the same embodiment, nor does it mean that it is mutually exclusive with other embodiments and is independent or optional. It is clearly or implicitly understood by those of ordinary skill in the art that the embodiments described in this application can be combined with other embodiments without conflict.

[0086] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of patent protection. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the scope of protection of the present application. Therefore, the scope of protection of the present application shall be subject to the attached claims.

Claims

1. A cluster server, characterized in that include: A switch and at least three servers, wherein the server is connected to the switch, the server includes a storage device and a central processor, the storage device includes a hard disk controller and a disk array, the hard disk controller is connected to the disk array of at least one other server through a disk connector; the hard disk controller is used to obtain or release control of the disk array of the current server and / or the disk array of at least one other server; When the running status of the current server is abnormal, but the central processor of the current server is working normally, the hard disk controller of the current server notifies other servers of the abnormal running status through the central processor; The server further comprises a baseboard management controller, the baseboard management controller is connected to the switch, and the baseboard management controller is also connected to a hard disk controller of the current server; When the central processor of each server cannot work normally, the hard disk controller is also used to monitor the operation status of the disk array of the current server, and release the control right of the disk array of the current server when the operation status of the current server is abnormal, and notify the abnormal operation status to other servers through the baseboard management controller; The hard disk controller is also used to receive abnormal operation status notifications of other servers with abnormal operation status from the baseboard management controller of the current server, and obtain control rights over the disk arrays of other servers with abnormal operation status.

2. The cluster server according to claim 1, characterized in that: Each of the hard disk controllers is connected to the disk array of the current server and the disk arrays of two other servers via a disk connector, and the storage devices of each of the servers are connected in a ring topology.

3. The cluster server according to claim 1, characterized in that: The hard disk controller is used to monitor the operating status of the disk array of the current server, and release the control right of the disk array of the current server when the operating status of the current server is abnormal; and monitor the operating status of the disk arrays of other servers, and obtain the control right of the disk arrays of other servers when the operating status of other servers is abnormal.

4. The cluster server according to claim 3, characterized in that: The abnormal operation state includes: there is no read or write operation on the disk array within a preset time interval.

5. The cluster server according to claim 3, characterized in that: The storage device of the server is powered by a power supply independent of the computing part of the server; The abnormal operation state includes: detecting that a computing part of the server loses power.

6. The cluster server according to claim 5, characterized in that: The baseboard management controller is also used to perform self-check and repair on the hardware of the current server after the control of the disk array of the current server is transferred to other servers; The hard disk controller is also used to regain control of the disk array of the current server when the current server is in normal operation.

7. The cluster server according to claim 5, characterized in that: The storage device of each server is powered by a power supply independent of the computing part of the server, and the baseboard management controller performs self-check and repair by restarting the current server.

8. The cluster server according to any one of claims 1 to 7, characterized in that: The disk connector is a SAS connector.

9. The cluster server according to any one of claims 1 to 7, characterized in that: The storage device of each server is physically centrally arranged in the server.

Images