WO2017067402A1

WO2017067402A1 - Computing device and computing device memory component management method and system

Info

Publication number: WO2017067402A1
Application number: PCT/CN2016/101735
Authority: WO
Inventors: 牛功彪; 李舒
Original assignee: 阿里巴巴集团控股有限公司
Priority date: 2015-10-19
Filing date: 2016-10-11
Publication date: 2017-04-27
Also published as: CN106598908A; CN106598908B

Abstract

A computing device and a computing device memory component management method and system, said computing device comprising: a computing component (20), and an interface (21) connected to an SAS switch; the computing component (20) is connected to the SAS switch by means of the interface (21), and the SAS switch also connected to a memory component; on the basis of an allocation rule, a memory component is allocated to the computing component (20) connected to the SAS switch. Separating the life cycles of a computing component and a memory component is conducive to realizing the broadest use of components having different life cycles, and by decoupling computing components and memory components, computing components and memory components can be allocated more finely and flexibly.

Description

Computing device and computing device storage component management method and system

The present application claims priority to Chinese Patent Application No. 201510681163.5, filed on Jan. 19, 2015, entitled <RTI ID=0.0>> in.

Technical field

The present application relates to the field of computers, and in particular, to a computing device and a method and system for managing storage components of a computing device.

Background technique

Currently, the storage and computing components of the data center are configured based on a single computing device, that is, the ratio of storage components to computing components for a computing device is fixed. The configuration of the storage unit and the computing unit based on the single computing device has at least the following problems:

Although the configuration of storage resources and computing resources can be implemented in a distributed manner, the refinement and flexible allocation and management of storage resources and computing resources cannot be achieved.

In addition, there is a difference between the storage component and the computing component in the life cycle, and the life cycle of the general computing component is shorter than the life cycle of the storage component. This configuration mode causes the storage component to be eliminated prematurely, resulting in waste of resources.

Therefore, the configuration scheme of the storage component and the computing component based on the single computing device needs to be improved.

Summary of the invention

One of the technical problems solved by the present application is to provide a management method and system for a computing device and a storage component of a computing device, and to realize the flexible allocation of the storage component and the computing component, and fully utilize the service time limit of each component.

According to an embodiment of an aspect of the present application, a computing device is provided, including: a computing component, further comprising:

Interface to SAS SWITCH;

The computing component is coupled to the SAS SWITCH via the interface, the SAS SWITCH being simultaneously coupled to the storage component, the storage component being allocated for the computing component connected to the SAS SWITCH based on an allocation rule.

According to an embodiment of another aspect of the present application, there is provided a management method of a computing device storage component, the computing device comprising a computing component, the computing component being connected to a SAS SWITCH, the SAS SWITCH When connected to the storage component, the method includes:

Obtaining storage component information and computing component information connected to the SAS SWITCH;

And storing the storage component for the computing component based on the obtained storage component information and the computing component information based on the allocation rule.

According to an embodiment of a further aspect of the present application, a method for managing a storage component is provided, including:

Obtaining storage component information and computing component information connected to the serial small computer system interface switch SAS SWITCH;

According to an embodiment of still another aspect of the present application, a management system for a computing device storage component is provided, comprising: a management node, at least one SAS SWITCH connected to the management node, at least one storage component connected to the SAS SWITCH, and At least one computing device computing component, wherein

The management node is connected to the SAS SWITCH through a management interface provided by the SAS SWITCH, and is configured to allocate a storage component to a computing component connected to the SAS SWITCH.

In the embodiment of the present application, the storage component in the computing device is separated from the computing device, and the separated storage component is connected to the computing component in the computing device through the SAS SWITCH, which has at least the following advantages:

The separation of computing resources and storage resources is realized, which facilitates separate maintenance and management of computing resources and storage resources;

The life cycle of the computing component and the storage component are separated to maximize the use of components of different life cycles;

Increased maintainability and reduced maintenance effort.

The decoupling between the computing component and the storage component is realized, and the storage component can be allocated according to the needs of the computing component, thereby realizing the refined and flexible allocation of the computing component and the storage component.

Those skilled in the art will appreciate that although the following detailed description is made with reference to the illustrated embodiments and drawings, the application is not limited to these embodiments. Rather, the scope of the present application is broadly defined and the scope of the application is defined by the appended claims.

DRAWINGS

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings to be used in the embodiments will be briefly described below. Obviously, the drawings in the following description are only some of the present application. Implementation For example, other drawings may be obtained from those of ordinary skill in the art in light of the inventive work.

1 is a schematic structural view of a prior art computing device.

2 is a block diagram of a computing device in accordance with an embodiment of the present application.

3 is a flow chart of a method of managing a computing device storage component in accordance with an embodiment of the present application.

4 is a schematic structural diagram of a management system of a computing device storage component according to an embodiment of the present application.

FIG. 5 is a timing diagram of powering up devices in a management system of a computing device storage component according to an embodiment of the present application.

detailed description

Before discussing the exemplary embodiments in more detail, it should be noted that some exemplary embodiments are described as a process or method depicted as a flowchart. Although the flowcharts describe various operations as a sequential process, many of the operations can be implemented in parallel, concurrently or concurrently. In addition, the order of operations can be rearranged. The process may be terminated when its operation is completed, but may also have additional steps not included in the figures. The processing may correspond to methods, functions, procedures, subroutines, subroutines, and the like.

The computer device includes a user device and a network device. The user equipment includes, but is not limited to, a computer, a smart phone, a PDA, etc.; the network device includes but is not limited to a single network server, a server group composed of multiple network servers, or a cloud computing based computer Or a cloud composed of a network server, wherein cloud computing is a type of distributed computing, a super virtual computer composed of a group of loosely coupled computers. Wherein, the computer device can be operated separately to implement the present application, and can also access the network and implement the application through interaction with other computer devices in the network. The network in which the computer device is located includes, but is not limited to, the Internet, a wide area network, a metropolitan area network, a local area network, a VPN network, and the like.

It should be noted that the user equipment, the network equipment, the network, and the like are only examples, and other existing or future computer equipment or networks may be applicable to the present application, and should also be included in the scope of the present application. It is included here by reference.

The methods discussed below, some of which are illustrated by flowcharts, can be implemented in hardware, software, firmware, middleware, microcode, hardware description languages, or any combination thereof. When implemented in software, firmware, middleware or microcode, the program code or code segments to carry out the necessary tasks can be stored in a machine or computer readable medium. (such as storage media). The processor(s) can perform the necessary tasks.

The specific structural and functional details disclosed herein are merely representative and are for purposes of describing exemplary embodiments of the present application. The present application, however, may be embodied in many alternative forms and should not be construed as being limited to the embodiments set forth herein.

It should be understood that although the terms "first," "second," etc. may be used herein to describe the various elements, these elements should not be limited by these terms. These terms are used only to distinguish one unit from another. For example, a first unit could be termed a second unit, and similarly a second unit could be termed a first unit, without departing from the scope of the exemplary embodiments. The term "and/or" used herein includes any and all combinations of one or more of the associated listed items.

It will be understood that when a unit is referred to as "connected" or "coupled" to another unit, it can be directly connected or coupled to the other unit, or an intermediate unit can be present. In contrast, when a unit is referred to as being "directly connected" or "directly coupled" to another unit, there is no intermediate unit. Other words used to describe the relationship between the units should be interpreted in a similar manner (eg "between" and "directly between" and "adjacent to" Than "directly adjacent to", etc.).

The terminology used herein is for the purpose of describing the particular embodiments, The singular forms "a", "an", It is also to be understood that the terms "comprising" and """ Other features, integers, steps, operations, units, components, and/or combinations thereof.

It should also be noted that, in some alternative implementations, the functions/acts noted may occur in a different order than that illustrated in the drawings. For example, two figures shown in succession may in fact be executed substantially concurrently or sometimes in the reverse order, depending on the function/acts involved.

The computing device described in the embodiment of the present application is a general term for a device that provides a computing service function, and includes, but is not limited to, a server or a PC.

The storage components of a computing device in the prior art are fully coupled to the computing components. As shown in FIG. 1 , the structure of the existing computing device is shown. The computing component 11 in the computing device corresponds to a fixed set of storage components 10 . The storage component and the computing component have at least the following disadvantages based on the configuration scheme of the single computing device:

1. It is impossible to refine, flexibly allocate and manage storage components and computing components.

2. There is a difference between the storage component and the computing component in the life cycle. Usually, the warranty life of the computing device is 3 years. The computing unit of the computing device should have a service life of 3 years, while the storage component has a warranty period of 5 years or longer. When the computing device reaches the warranty period, the computing device will be eliminated, and the warranty period of the storage component is not reached at this time. Therefore, the warranty period of the storage component is not fully utilized, but is prematurely eliminated, resulting in storage components. Part of the warranty time is wasted, which ultimately leads to a waste of costs.

For the problem of the above-mentioned computing device, the embodiment of the present application provides a simplified computing device and a simplified management method and system for the computing device storage component. The technical solution of the present application is further described in detail below with reference to the accompanying drawings.

2 is a block diagram of a computing device including a computing component 20 and an interface 21 coupled to a SAS SWITCH, in accordance with an embodiment of the present application. As shown in Figure 2, the computing component 20 of the computing device is still disposed within the computing device, i.e., the computing component 20 is arranged in the same manner as in the prior art. The computing component 20 is an I/O (input/output) module that retains memory and CPU (or GPU) computing functions.

The computing component 20 is connectable to the SAS SWITCH via the interface 21, which is simultaneously connected to the storage component, ie the storage component and the computing component 20 are simultaneously connected to the SAS SWITCH. The SAS SWITCH is an existing IP switch applying the SAS protocol. The computing component 20 is assigned a storage component coupled to the SAS SWITCH based on an allocation rule by a management node or hypervisor.

The embodiment of the present application is to implement flexible allocation of the computing device storage component and the computing component 20, and to enable the storage component with different lifecycles and the computing component 20 to implement the best service time, and separate the storage component of the computing device from the computing device. Only the computing component 20 is retained in the computing device. The computing component 20 retained in the computing device and the separated storage components are connected by SAS SWITCH.

The separated storage unit is connected to the SAS SWITCH through a unified interface in a storage array manner. One embodiment of the SAS SWITCH has a management interface, and the management node can be connected to the management node, and the management node can implement the connection based on the distribution rule. The storage component of the SAS SWITCH and the computing component 20 are matched, that is, the storage component connected to the SAS SWITCH is assigned a storage component connected to the SAS SWITCH. Another embodiment is to install a hypervisor in one of the computing devices connected to the SAS SWITCH, which can allocate storage components for computing components connected to the SAS SWITCH based on the allocation rules.

When the storage component 23 is allocated to the computing component 20, it may be allocated according to the actual needs of the computing component 20, that is, the number of storage components allocated for the computing component 20 varies with the actual needs of the computing component, wherein the computing component 20 The number of connected storage components can be proportional to the requirements of computing component 20. Thus, embodiments of the present application enable flexible allocation of storage components and computing components 20 of computing devices.

In addition, at least one storage array and computing component 20 can be simultaneously connected to the SAS SWITCH, and the storage array At least one storage component is included, so when the storage component is allocated to the computing component 20, the storage components located in different storage arrays can be allocated to the same computing component 20, that is, the storage components connected to the same computing device can be located in multiple In the storage array, to ensure that the impact on the computing device computing component 20 is reduced in the event of a failure of one of the storage arrays. For example, there are 20 storage arrays connected to a SAS SWITCH, each storage array includes 40 storage components, and 5 computing components connected to the SAS SWITCH are allocated for each of the 5 computing components. When the component is stored, one-fifth of the storage components selected from the 20 storage arrays are allocated to one computing component.

It should be noted that the computing device of the embodiment of the present application separates the storage component and connects with the computing component by using the SAS SWITCH, so that the storage component is separated from the computing component life cycle, so that the computing component reaches the specified service time limit (ie, the quality assurance is achieved). Period) does not affect the operation of the storage component, that is, the calculation component connected to the storage component can be replaced at this time, so as to maximize the utilization of the storage component and the calculation component of different life cycles, thereby avoiding the storage component being eliminated in advance The resulting resources are wasted.

The embodiment of the present application further provides a management method for a storage device of a computing device, as shown in FIG. 3 is an operation flowchart of the method, where the method mainly includes the following steps:

S310. Obtain storage component information and computing component information that are connected to the SAS SWITCH.

S320. Allocate a storage component for the computing component based on the obtained storage component information and the computing component information based on the allocation rule.

To further understand this solution, the above steps are further described in detail below.

An application scenario of the embodiment of the present application may be: the computing device is the same as the computing device described in the foregoing embodiment, that is, the computing device includes a computing component, and the computing component is connected to the SAS SWITCH, the SAS SWITCH is also connected to the storage unit. The method of this embodiment is for implementing a storage component for a computing component of the computing device.

The storage component of the embodiment of the present application is located in the storage array, and is connected to the SAS SWITCH by a unified interface of the storage array, that is, the storage array includes at least one storage component.

The storage component information acquired in step S320 includes: the number of storage components connected to the SAS SWITCH and the storage capacity, and may also include a state in which the storage component is used by the computing component and/or a service time limit of the storage component. The obtained computing component information of the computing device includes at least: a number of computing components connected to the SAS SWITCH, and may also include a computing time limit of the computing component.

The storage component information and the calculation component information can be obtained by:

Let SAS SWITCH scan each interface to get the internal signal of the cable connected to each interface of SAS SWITCH to identify whether it is connected to the interface as a computing component or a storage component (or storage array). That is, by scanning the chip connected to each interface of the SAS SWITCH, it is known whether the computing component or the storage component (or storage array) is connected to each interface of the SAS SWITCH. For example, if the chip of an interface of the SAS SWITCH is scanned, the port number of the interface is obtained, indicating that the interface is connected to the interface; if the chip of the interface of the SAS SWITCH is scanned, the controller signal is obtained, Indicates that the storage unit (or storage array) is connected to the interface. The storage unit information connected to the SAS SWITCH and the calculation part information are obtained in the case where all SAS SWITCH interfaces are scanned. The storage component information and the calculation component information can be obtained by acquiring the scan result of the SAS SWITCH.

The execution timing of this step S310 can be during or after the SAS SWITCH startup process. Each time SAS SWITCH starts, a scan operation is performed to obtain storage part information and calculation part information connected to the SAS SWITCH. It is also possible to set the SAS SWITCH to perform the scanning operation periodically after startup, so as to timely acquire the storage component information and calculate the change of the component information, for example, the change of the usage state of the storage component by the computing component.

The purpose of obtaining the storage component information connected to the SAS SWITCH and the computing component information of the computing device is that the storage component can be allocated to the computing component based on the obtained storage component information and the computing component information based on the allocation rule.

Step S320 is to allocate a storage component to the computing component based on the obtained storage component information and the computing component information based on the allocation rule. Wherein, in the case that the storage component is initially allocated for the computing component after the SAS SWITCH is started, the storage component is allocated to the computing component based on the number of the obtained storage component and the number of the calculated component based on the allocation rule. It is also possible to assign a storage component to the computing component based on the acquired storage component capacity and the number of calculated components. For the case that the updated storage component information is acquired after the SAS SWITCH is started, and the component information is calculated, the storage component is adjusted to the storage component allocated by the computing component according to the state of the obtained storage component used by the computing component, and the storage component according to the obtained storage component. The service time limit and/or the service time limit of the computing component adjusts the distribution relationship between the storage component and the computing component.

The storage component may be allocated to the computing component according to the configured allocation rule when the storage component is initially allocated to the computing component, and the distribution rule may include:

A storage unit other than the specified number of storage units is allocated to the computing unit while retaining the specified number of storage units in an unassigned state.

Wherein, in the case that the storage component is disposed in the storage array, the case of retaining the specified number of storage components includes: retaining a specified number of storage components in each storage array in an unallocated state, that is, retaining the storage array middle finger A fixed number of storage components are not assigned to any computing component, and the reserved specified number of storage components can be used as a backup of the failed storage component, ie, assigned to the computing component, in the event of a storage component failure assigned to the computing component. In the event of a storage component failure, the reserved number of storage components are used in place of the failed storage component to ensure that the normal operation of the computing component connected to the failed storage component is not affected. The number of storage components reserved in each storage array may be one or two, or the same number of storage components assigned to any of the computing components. Additionally, the number of specified number of storage components retained in each storage array may be the same or different. It will be appreciated that for ease of management, the number of specified number of storage components retained in each storage array is typically set to be the same.

As can be seen from the above-mentioned distribution rules, in the case of retaining a specified number of storage components without allocation, the other storage components are all allocated as much as possible to achieve full utilization of the storage components. For example, if one of the storage arrays connected to the SAS SWITCH includes 60 storage components, and it is determined that the specified number of storage components to be retained are in the unallocated state, the remaining 58 storage components are all allocated. The computing component that is connected to the SAS SWITCH.

In addition, it should be noted that, for the convenience of management, the number of storage components included in the storage array connected to the SAS SWITCH can be set to be the same, and the number of storage components initially allocated for each computing component according to the distribution rule can also be set to be the same. It can be understood that the embodiment of the present application also supports the case where the number of storage components included in each storage array connected to the SAS SWITCH is different, and the case where a different number of storage components are initially allocated for each computing component.

The allocation rule may further include:

The storage components in the storage array are divided into the same number of copies as the number of computing components assigned to the computing components.

That is, for a storage array, a plurality of storage components it contains are divided into the same number of copies as the number of computing components. Since in general, the number of storage arrays connected to the SAS SWITCH and the number of storage components included in the storage array will be greater than the number of computing components connected to the SAS SWITCH, that is, the storage components connected to the SAS SWITCH are sufficient to be allocated to the connection. The computational component on SAS SWITCH, therefore, does not appear to have fewer storage components in the storage array than the number of computational components connected to SAS SWITCH.

For example, if there are 20 storage arrays connected to a SAS SWITCH, and each storage array contains 60 storage components, and 10 computing units connected to the SAS SWITCH, according to this rule, 60 in each storage array. The storage components are divided into 10 parts, each of which is assigned to one computing unit. Among them, the storage in the storage array can be The storage unit is equally divided into the same number of parts as the number of calculation parts, and can also be randomly divided into the same number of parts as the number of calculation parts. That is, the 60 storage units in the storage array can be equally divided into 10 parts, and each storage unit includes 6 storage units, that is, 6 storage units assigned to each calculation unit. Or randomly divide 60 storage components in the storage array into 10 shares, and the number of storage components included in each share may be unequal, for example, may be 5, 10, 6, or 8, etc., and the randomly divided storage may be Parts are assigned to the calculation part.

The allocation rule does not conflict with the above-mentioned allocation rule, that is, the storage unit in the storage array can be divided into the same number of copies as the number of computing units, if the specified number of storage units in the reserved storage array are in an unallocated state. Calculating components, for example, for the above example, the storage components in each storage array can be divided into 11 parts, one of which is reserved and not allocated to any computing part, and the remaining 10 parts are allocated to the computing part; of course, the allocation rule can also be Alone.

It should be emphasized that the rule is to divide the storage components in each storage array into the same number of copies as the computing components, so that the allocation rule can allocate the storage components in the same storage array to the connected components. Each computing component on the SAS SWITCH, that is, the storage component assigned to the same computing component, is located in multiple storage arrays. Thus, in the case of a storage array failure, only a small portion of the operation of a computing component is affected, that is, the impact of the storage component failure on the computing component is effectively reduced.

The allocation rule can also be:

All storage components connected to the SAS SWITCH are allocated to the computing component in the same number of copies as the number of computing components.

The allocation rule is to assign the storage units in all the storage arrays connected to the SAS SWITCH to the computing unit by the same number of copies as the number of computing units connected to the SAS SWITCH. The allocation rule differs from the previous allocation rule in that it is possible for storage components in each storage array to be assigned to the same computing component.

For example, if there are 20 storage arrays connected to SAS SWITCH, and each storage array contains 60 storage components, and 10 computing units connected to SAS SWITCH, it is known that there are 1200 storage components connected to the SAS SWITCH ( 20 × 60), the 1200 storage components are divided into 10 shares and assigned to 10 computing components. Also, the storage unit may be equally divided into the same number of copies as the number of calculated parts, or may be randomly divided into the same number of parts as the number of calculated parts. According to the present rule, in the case of dividing 1200 storage parts into 10 shares, each of which includes 120 storage parts, 120 storage parts are assigned to one calculation part. The 120 storage elements may be selected in the order of the interface. For example, if the number of the SAS SWITCH interface connected to the storage array is the interface 1 to the interface 20, the interface number of the SAS SWITCH connected to the computing component is the interface. 21 to interface 31. Then, 60 storage components are selected from the interface 1, and then 60 storage components are selected from the interface 2 and assigned to the interface connected to the interface 21. The computing component; the 120 storage components acquired from interface 3 and interface 4 are assigned to computing components connected to interface 22, and so on.

The allocation rule can also be:

A fixed capacity storage unit is assigned to each computing component.

For example, there are 20 storage components connected to a SAS SWITCH, and each storage component has a storage capacity of 500G, and 10 computing components are connected to the SAS SWITCH. The rule stipulates that each computing component is allocated 1000G storage components. Then, a storage component of 1000 G randomly selected from the 20 storage components can be allocated to the computing component, and the allocated storage components are guaranteed not to collide. The selection method can be selected according to the interface order as described in the above rules. Of course, other methods can be used to select the storage components for the computing component, and only need to ensure that the capacity of the storage component allocated for each computing component is equal to the fixed rule. Capacity is fine.

The foregoing distribution rules are only a few examples enumerated by the inventors, and various allocation rules can be flexibly set as needed in practical applications. It can be seen that the embodiment of the present application can set various allocation rules to allocate storage components for the computing part, and effectively realize flexible allocation of the storage component and the computing component.

Wherein, the specific representation of the storage component for the computing component is that the computing component can access and use the storage component assigned to it. That is, the process of allocating storage components for computing components is the process of opening the access and usage rights of the corresponding storage components for the computing components, and also the process of setting the matching relationship between the computing components and the storage components. After the storage component is allocated to the computing component according to the distribution rule, the matching relationship between the computing component and the storage component can be recorded. Since each interface of the SAS SWITCH has an address (or number), and each interface is connected to a computing component or a storage array, the address of the computing component or storage array connected to the interface is the address of the interface of the SAS SWITCH. If the interface to a SAS SWITCH is connected to a storage array, the storage array contains a plurality of storage components having a unique address (or number) for each storage component in the storage array. Then, after the storage component is allocated to the computing component, the correspondence between the computing component and the address of the allocated storage component can be recorded when the matching relationship between the computing component and the storage component is recorded. For example, the interface 1 of the SAS SWITCH is connected to the storage array. The address of the interface 1 is assumed to be address 1, and the address of the storage array 1 is the address 1; the interface 5 is connected to the computing component, and the address of the interface 5 is assumed to be the address 5, then The address of the computing component is the address 5; if the storage component X in the storage array is assigned to the computing component, and the address of the storage component X is the address X, the matching relationship between the computing component and the storage component is: address 5 Corresponding to the address X in address 1, that is, if the computing component wants to access the storage component X allocated to it, it needs to find the address of the storage array where the storage component X is located, that is, the address 1, and then find the address X in the address 1 to find the address X. The storage part X assigned to it. It can be understood that since the storage component allocated for the computing component may be located in multiple storage arrays, when the computing component and the storage component are matched, the same computing component can be simultaneously Storage component matching. The embodiment of the present application may save the matching relationship between the computing component and the storage component by matching the relationship table.

The matching relationship between the calculation part and the storage part recorded in the matching relation table. The matching relationship table can be synchronously updated when the matching relationship between the computing component and the storage component changes.

Since the foregoing allocation rule is for the initial allocation process, the initial allocation process is the process of allocating storage components for the computing component for the first time after the SAS SWITCH is started, and after the initial allocation is completed, the matching relationship table is generated, and the subsequent possible The matching relationship is adjusted, and each time the matching relationship is adjusted, the matching relationship table is updated correspondingly. The storage component information and the calculation component information may be obtained periodically, and the matching relationship needs to be adjusted when the acquired storage component information and the calculation component information change. The specific scenarios include:

The acquired storage component information includes a state in which the storage component is used by the computing component; the storage component assigned to the computing component is adjusted according to a state in which the storage component is used by the computing component.

That is, the embodiment of the present application acquires the state in which the storage component allocated to the computing component is used by the computing component in real time or periodically after initially allocating the storage component for the computing component. The use state is whether the storage capacity of the storage unit is used, the ratio of use, and the like.

The purpose of obtaining the usage state is to adjust the storage component allocated to the computing component according to the usage state, and mainly adjust the number of storage components allocated to the computing component. Specific adjustment methods include:

When the usage ratio of the storage component to the computing component is lower than the set lower limit value, the storage component allocated to the computing component is reduced, that is, a part of the unused storage component is removed from the storage component allocated to the computing component or All. For example, if the usage rate of the calculated component of the storage component is less than 60%, the unused 40% of the storage component is canceled, or 30% of the unused storage component is canceled, and the like. or,

When the usage rate of the storage component by the computing component is higher than the set upper limit value, the storage component allocated for the computing component is added. For example, if the usage rate of the calculated component of the storage component is higher than 95%, then the computing component is assigned one or two storage components, etc., or all or part of the storage component that is canceled from other computing components is assigned to The calculation unit whose usage rate is higher than the upper limit value.

The usage rate of the obtained storage component by the calculation component may be an average usage rate within a predetermined time length range.

The scenario in which the matching relationship needs to be adjusted further includes:

The acquired storage component information includes a service time limit of the storage component, the obtained information of the computing component includes a service time limit of the computing component, a service time limit of the storage component, and a service time limit of the computing component are reached respectively, and a predetermined time limit is reached; The service time limit reaches the specified time limit and/or the service time limit of the computing component reaches the specified time limit. In the case of adjusting the matching relationship, in particular, providing a prompt to replace the storage component or the computing component that reaches the specified time limit, so that the worker adjusts according to the replacement result according to the storage component or the computing component according to the prompt according to the specified time limit. The matching relationship.

The above-described operation of determining the storage means and calculating whether or not the time limit of the component reaches the respective predetermined time limit can be periodically performed. The warranty time is reached when the specified time limit is reached. The predetermined time limit of each computing component and the storage component is saved, and whether the current time is a specified time limit of the saved computing component and the storage component, and if the predetermined time limit of any of the computing components is specified, the computing component is prompted to be replaced. . At this point, a new computing component can be created to replace the computing component that reaches the specified time limit without affecting the use of the storage component, ie the storage component can continue to be used. Both the computing component and the storage component are utilized to the maximum extent.

The embodiment of the present application further provides a management system for a computing device storage component, as shown in FIG. 4 , which is a schematic structural diagram of the system. The system mainly includes the following: a management node 40, at least one SAS SWITCH 41 connected to the management node 40, At least one storage component 420 coupled to the SAS SWITCH 41 and computing component 43 of at least one computing device. As shown in FIG. 4, storage component 420 can be located in storage array 42, with at least one storage component 420 included in each storage array.

The storage array 42 is a combination of storage components 420 separated from the computing device, and each storage array 42 is connected to the SAS SWITCH 41 through a unified interface.

The computing component 43 is an I/O (input/output) module that retains memory and CPU (or GPU) computing functions in the computing device, which is internal to the computing device.

The storage array 42 and the computing component 43 can be located in the same cabinet, that is, the storage array 42 and the computing component 43 are included in the same cabinet. Of course, the storage arrays 42 can be separately disposed in one cabinet, and the computing component 43 can be disposed in another cabinet. This arrangement is more advantageous for storing different storages in different storage arrays 42. Component 420 is assigned to the same computing component 43, providing the benefit of redundancy.

The storage array 42 and the computing component 43 are connected by a SAS SWITCH 41, which is an existing IP switch applying the SAS protocol. The SAS SWITCH 41 has a plurality of interfaces including a management interface for connecting to the management node 40, and a plurality of interfaces connecting the storage array 42 and the computing component 43. The SAS SWITCH 41 can obtain the storage array 42 or the computing component 43 connected to each interface by scanning each interface after booting, thereby obtaining the number of storage arrays 42 connected to the SAS SWITCH 41 and the number of computing components 43 for connecting the storage array 42. The interface can be further scanned to obtain the number of storage components 420 included in the storage array 42 and the storage space size of each storage component 420.

The management node 40 is configured to allocate, to the computing component 43, a storage component 42 connected to the computing component 43 on the SAS SWITCH 41. It should be noted that at least one SAS SWITCH 41 is connected to one management node 40 in the management system. In FIG. 4, only the management node 40 is connected to one SAS SWITCH 41 as an example. That is, the same management node 40 can simultaneously manage a plurality of storage arrays 42 and computing components 43 connected by SAS SWITCH 41. The management node 40 is connected to the SAS SWITCH 41 through a management interface provided by the SAS SWITCH41. The number of computing components 43 and storage arrays 42 connected to the plurality of SAS SWITCHs 41 connected to the same management node 40 may be the same or different. In the case where the number of connections is the same, the same management policy (i.e., the policy of allocating the storage unit 420 to the computing component 43) may be employed for the plurality of SAS SWITCHs 41 for ease of management.

The management node 40 may be an existing computing device, or may be one of the computing devices connected to the SAS SWITCH 41 provided by the embodiment of the present application, or a management program in the computing device.

To improve the reliability, the management node 40 of the embodiment of the present application includes two management devices (dual management nodes) that are mutually active and standby, and two management devices that are mutually active and standby can synchronize the heartbeat through a USB (Universal Serial Bus) signal. The two management devices that are mutually active and standby synchronize the saved information in real time or periodically to ensure that the standby management device can replace the management of the failed management device in the event of a failure of the primary management device.

In addition, in order to further improve the reliability, the storage component included in each management device may be set to be a primary-standby relationship, that is, the storage components included in each management device are divided into two groups, one of which is a group. As a primary storage component, another set serves as a backup of the primary storage component to ensure that in the event of a primary storage component failure, the alternate storage component can replace the primary storage component's operation to reduce the impact of the management node failure on the operation of the computing device.

The reliability of the system is ensured by the above-mentioned management devices that are active and standby, and the storage components that are active and standby in the management device. This reduces the probability that the computing device cannot work normally due to the failure of the management device.

The system described in this embodiment can control the power-on sequence of each device. The power-on sequence diagram of each device is as shown in FIG. 5. As can be seen from FIG. 5, the management node 40 is powered on first in the system. That is, the bypass power supply is powered. During power-on of the management node 40, the main power is ready, and the main power is the power of other devices in the system except the management node 40. After the power-on of the management node 40 is completed, the management node 40 controls the power-on sequence of other devices. That is, after the power-on startup of the management node 40 is completed, the power-on sequence of other devices including the SAS SWITCH 41, the storage array 42, and the computing component 43 can be controlled. The power-on sequence of the other devices can be:

The SAS SWITCH 41 is controlled to power up first, after which the storage array 42 and the computing component 43 can be powered up simultaneously.

It can be understood that the storage array 42 is powered on to power up the storage component 420. In the system, the work of allocating the storage unit 42 to the computing component 43 of the computing device can be performed only after the management node 40 is powered on. If the power-on sequence of the foregoing devices is disordered, the storage array 42 and the computing component 43 are powered on before the SAS SWITCH 41, or before the management node is powered on, or the SAS SWITCH 41 is powered on before the management node 40, and the devices after the power-on are required. Waiting for the management node 40 to be powered on after the boot is completed, this process will increase the power consumption of the system. It can be seen that the power-on sequence of each device provided in this embodiment of the present application can effectively reduce power consumption.

The specific representation of the management node 40 assigning the storage component 420 to the computing component 43 is that the computing component 43 can access and use the storage component 420 assigned thereto. That is, the process in which the management node 40 allocates the storage component 420 to the computing component 43 is the process of opening the access and usage rights of the corresponding storage component 420 for the computing component 43, and also the process of setting the matching relationship between the computing component 43 and the storage component 420. After the storage unit 420 is allocated to the calculation unit 43 in accordance with the distribution rule, the matching relationship between the calculation unit 43 and the storage unit 420 can be recorded. Since each interface of the SAS SWITCH 41 has an address (or number), and each interface is connected to the computing component 43 or the storage array 42, the address of the computing component 43 or the storage array 42 connected to the interface is the interface of the SAS SWITCH. the address of. If the interface to a SAS SWITCH 41 is connected to a storage array 42, the storage array 42 includes a plurality of storage components 420 having a unique address (or number) for each storage component 420 in the storage array. Then, after the storage unit 420 is allocated to the calculation unit 43, the correspondence relationship between the calculation unit 43 and the address of the allocated storage unit 420 can be recorded when the matching relationship between the calculation unit 43 and the storage unit 420 is recorded.

Wherein, the management node 40 is configured to allocate the storage component 42 to the computing component 43 as:

The storage unit 420 other than the specified number of storage units 420 is assigned to the calculation unit 43 while retaining the specified number of storage units 420 in an unallocated state.

Wherein, in the case where the storage component 420 is located in the storage array 42, the reserved number of storage components 420 is reserved for each storage array 42 to reserve a specified number of storage components 420 in an unallocated state, ie, the specified number of reserved storage arrays 42 The storage component 420 is not assigned to any of the computing components 43. The reserved number of reserved storage components 420 can be used as a backup for the failed storage component in the event of a storage component failure assigned to computing component 43. That is, the management node 40 is configured to replace the failed storage component 420 with the specified number of storage components 420 retained in the event of a failure of the storage component 420 assigned to the computing component 43 to ensure that the failure is not affected. The normal operation of the computing unit 43 to which the storage unit 420 is connected.

The number of storage components 420 reserved in each storage array 42 may be one or two, or the same number as the storage components 420 assigned to any of the computing components. Additionally, the number of designated number of storage components 420 remaining in each storage array 42 may be the same or different. It will be appreciated that the number of designated number of storage components 420 retained in each storage array is typically set to be the same for ease of management.

The management node 40 may also be configured to: when allocating the storage component 42 to the computing component 43:

The plurality of storage sections 420 in the storage array 42 are divided into the same number of copies as the calculation section 43 and assigned to the calculation section 43.

That is, for one memory array 42, a plurality of storage sections 420 contained therein are divided into the same number of copies as the number of calculation sections 43 and assigned to the calculation section 43. Since in general, the number of storage arrays connected to the SAS SWITCH and the number of storage components included in the storage array will be greater than the number of computing components connected to the SAS SWITCH, that is, the storage components connected to the SAS SWITCH are sufficient to be allocated to the connection. The computational component on SAS SWITCH, therefore, does not appear to have fewer storage components in the storage array than the number of computational components connected to SAS SWITCH.

For example, if there are 20 storage arrays 42 connected to the SAS SWITCH 41, and each storage array 42 includes 60 storage components 420, and the number of computing components 43 connected to the SAS SWITCH 41 is 10, each storage array is configured according to the present rule. The 60 storage units 420 of 42 are divided into 10 shares, one for each of which is assigned to one calculation unit 43. Here, the storage unit 420 in the storage array 42 may be equally divided into the same number of copies as the number of the calculation unit 43, or may be randomly divided into the same number of copies as the number of calculation units. That is, the 60 storage units 420 in the storage array 42 can be equally divided into 10 shares, and each storage unit includes six storage units, that is, six storage units for each calculation unit. Or randomly divide 60 storage components 420 in the storage array 42 into 10 shares, and the number of storage components 420 included in each share may be unequal, for example, may be 5, 10, 6, or 8, etc., and will be randomly divided. The subsequent storage unit 420 is assigned to the calculation unit 43.

The rule is that the storage unit 420 in each storage array 42 is divided into the same number of copies as the calculation unit 43 and assigned to the calculation unit 43, so that the allocation rule can assign the storage unit 420 in the same storage array to the connection unit. Each computing component 43 on the SAS SWITCH 41, thus, in the event of a failure of the storage array 42, affects only a small portion of the operation of the computing component 430, i.e., effectively reduces the impact of the storage component 420 failure on the computing component.

The management node 40 can also be configured to:

The state in which the storage unit 420 is used by the calculation unit 43 is acquired; the storage unit 420 assigned to the calculation unit 43 is adjusted in accordance with the state in which the storage unit 420 is used by the calculation unit 43.

That is, the embodiment of the present application acquires the state in which the storage component allocated to the computing component is used by the computing component in real time or periodically after initially allocating the storage component for the computing component. The use state is whether the storage space of the storage component is used, the ratio of use, and the like.

The purpose of obtaining the state in which the storage component is used by the computing component is that the number of storage components allocated to the computing component can be adjusted according to the state of the storage component computing component, including:

When the storage component is used by the computing component to be lower than the set lower limit value, the storage component allocated to the computing component is reduced, that is, part or all of the unoccupied storage component is removed from the storage component allocated to the computing component . For example, if the acquired storage component is used by the computing component below 60%, the unused 40% storage component is cancelled, or 30% of the unused storage component is cancelled, and the like. or,

When the storage component is used by the computing component above the set upper limit, the storage component allocated for the computing component is added. For example, if the acquired storage component is used by the computing component higher than 95%, then the computing component is assigned one or two storage components, etc., or all or part of the storage components that are canceled from other computing components are allocated for use. The rate is higher than the calculated part of the upper limit.

The usage rate of the obtained storage component may be an average occupancy rate within a predetermined length of time.

The management node 40 can also be configured to:

Determining whether the service time limit of the storage component and the service time limit of the computing component reach the respective specified time limit; and when the service time limit of the storage component reaches a predetermined time limit and/or the service time limit of the computing component reaches a predetermined time limit, providing a replacement to reach the specified time limit A reminder for a storage component or a computing component.

The determining operation may be performed periodically, and it is determined whether the storage time limit of the storage component and the computing component reaches a predetermined time limit, that is, whether the storage component and the computing component meet the warranty time. The predetermined time limit of each computing component and the storage component is saved, and whether the current time is a specified time limit of the saved computing component and the storage component is determined. If the predetermined time limit of any of the computing component or the storage component is reached, the prompting is replaced. The computing component or storage component. For example, where the service time limit of one of the computing components reaches a specified time limit, a new computing component can be created to replace the computing component that reaches the specified time limit without affecting the use of the storage component, ie, the storage component can continue to be used. Both the computing component and the storage component are utilized to the maximum extent.

In summary, the embodiment of the present application separates the storage component in the computing device from the computing device, and connects the separated storage component to the computing component in the computing device through SAS SWITCH, which has at least the following advantages:

1. Separation of computing resources and storage resources is realized, which facilitates separate maintenance and management of computing resources and storage resources;

2. The life cycle of the computing component and the storage component are separated, so that the components of different life cycles can be utilized to the maximum extent.

3. The decoupling between the computing component and the storage component is realized, and the storage component can be allocated according to the needs of the computing component, thereby realizing the refined and flexible allocation of the computing component and the storage component;

4, maintainability is enhanced, and the maintenance workload is effectively reduced.

It should be noted that the present application can be implemented in software and/or a combination of software and hardware, for example, using an application specific integrated circuit (ASIC), a general purpose computer, or any other similar hardware device. In one embodiment, the software program of the present application can be executed by a processor to implement the steps or functions described above. Likewise, the software programs (including related data structures) of the present application can be stored in a computer readable recording medium such as a RAM memory, a magnetic or optical drive or a floppy disk and the like. In addition, some of the steps or functions of the present application may be implemented in hardware, for example, as a circuit that cooperates with a processor to perform various steps or functions.

In addition, a portion of the present application can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide a method and/or technical solution in accordance with the present application. The program instructions for invoking the method of the present application may be stored in a fixed or removable recording medium, and/or transmitted by a data stream in a broadcast or other signal bearing medium, and/or stored in a The working memory of the computer device in which the program instructions are run. Herein, an embodiment in accordance with the present application includes a device including a memory for storing computer program instructions and a processor for executing program instructions, wherein when the computer program instructions are executed by the processor, triggering The apparatus operates based on the aforementioned methods and/or technical solutions in accordance with various embodiments of the present application.

It is obvious to those skilled in the art that the present application is not limited to the details of the above-described exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application. Therefore, the present embodiments are to be considered as illustrative and not restrictive, and the scope of the invention is defined by the appended claims instead All changes in the meaning and scope of equivalent elements are included in this application. Any reference signs in the claims should not be construed as limiting the claim. In addition, it is to be understood that the word "comprising" does not exclude other elements or steps. A plurality of units or devices recited in the system claims can also be implemented by a unit or device by software or hardware. The first, second, etc. words are used to denote names and do not denote any particular order.

Claims

A computing device, comprising: a computing component, further comprising:

An interface to a serial small computer system interface switch SAS SWITCH;

The computing component is connected to the serial small computer system interface switch SAS SWITCH through the interface, and the serial small computer system interface switch SAS SWITCH is simultaneously connected with the storage component, and the small computer connected to the serial based on the distribution rule The computing component on the system interface switch SAS SWITCH allocates storage components.
The computing device of claim 1 wherein said storage component is located in a storage array and is coupled to said serial small computer system interface switch SAS SWITCH by a unified interface of said storage array.
The computing device of claim 1 wherein the storage component allocated for the computing component is located in a plurality of storage arrays.
The computing device of claim 1 wherein the number of storage components coupled to said computing component is proportional to the demand of the computing component.
A management method of a computing device storage component, the computing device comprising a computing component coupled to a serial small computer system interface switch SAS SWITCH, the serial small computer system interface switch SAS SWITCH simultaneous storage Component connection, characterized in that the method comprises:

Obtaining storage component information and computing component information connected to the serial small computer system interface switch SAS SWITCH;

And storing the storage component for the computing component based on the obtained storage component information and the computing component information based on the allocation rule.
The method of claim 5 wherein the storage component coupled to the serial small computer system interface switch SAS SWITCH is located in the storage array with the unified interface of the storage array and the serial small computer system Interface switch SAS SWITCH connection.
The method of claim 6 wherein the storage component allocated for the computing component is located in a plurality of storage arrays.
The method according to claim 6 or 7, wherein the storage part information comprises a storage part number, the calculation part information comprises a calculation part quantity, and the storage part information and the calculation part information are based on the distribution rule according to the acquired Allocating storage components for computing components includes:

The specified number of storage units will be removed if the specified number of storage units are left unallocated The external storage component is assigned to the computing component.
The method of claim 8 wherein the method further comprises:

In the event of a storage component failure assigned to the computing component, the failed storage component is replaced with the specified number of storage components retained.
The method according to claim 5, wherein the storage component information comprises a state in which the storage component is used by the computing component, and the storing component is allocated to the computing component based on the obtained storage component information and the computing component information based on the allocation rule. :

The storage component assigned to the computing component is adjusted based on the state in which the storage component is used by the computing component.
The method of claim 5, wherein the storage component information comprises a service time limit of the storage component, the computing component information comprises a service time limit of the computing component, the method further comprising:

Determining whether the service time limit of the storage component and the service time limit of the computing component reach a respective predetermined time limit;

In the event that the service time limit of the storage component reaches a specified time limit and/or the service time limit of the computing component reaches a specified time limit, a prompt to replace the storage component and/or the computing component that reaches the specified time limit is provided.
A method for managing a storage component of a computing device, comprising:

Obtaining storage component information and computing component information connected to the serial small computer system interface switch SAS SWITCH;

And storing the storage component for the computing component based on the obtained storage component information and the computing component information based on the allocation rule.
A management system for computing device storage components, comprising: a management node, at least one serial small computer system interface switch SAS SWITCH connected to the management node, and a serial small computer system interface switch SAS SWITCH Connected at least one storage component and at least one computing component of the computing device, wherein

The management node is connected to the serial small computer system interface switch SAS SWITCH through a management interface provided by the serial small computer system interface switch SAS SWITCH for connecting to a small computer system interface connected to the serial The computing component on the switch SAS SWITCH allocates storage components.
The system according to claim 13, wherein said management node comprises two management devices that are mutually active and standby.
The system of claim 14 wherein said management device comprises storage components that are mutually active.
The system of claim 13 wherein said management node is configured to:

Controls the power-on sequence of the serial small computer system interface switch SAS SWITCH, storage components, and computing components.
The system of claim 16 wherein said powering sequence is:

After the serial small computer system interface switch SAS SWITCH is powered on, the storage component and the computing component are powered on.
The system of claim 13 wherein the storage components and computing components on the serial small computer system interface switch SAS SWITCH are located in different cabinets.
The system of claim 13 wherein the storage component coupled to the serial small computer system interface switch SAS SWITCH is located in the storage array, and the serial small computer system interface switch SAS is unified by the storage array SWITCH connection.
The system of claim 19 wherein said management node is configured to:

The storage components allocated for the compute component are located in multiple storage arrays.