WO2017124834A1

WO2017124834A1 - Resource management method, apparatus, and system

Info

Publication number: WO2017124834A1
Application number: PCT/CN2016/107277
Authority: WO
Inventors: 程茂林
Original assignee: 中兴通讯股份有限公司
Priority date: 2016-01-18
Filing date: 2016-11-25
Publication date: 2017-07-27
Also published as: CN106982241A

Abstract

Disclosed is a resource management method, said method comprising: a management node obtaining resource information mounted on control nodes; according to a received service request and said resource information, generating an assembly request; according to said assembly request, controlling said control node to assemble a logic server and processing said service request. Also disclosed are a resource management apparatus and a resource management system. In the present invention, a resource is mounted on each control node and the control nodes interact directly with the management node to manage resources; thus the complexity of software management is reduced, while the networking and wiring of the server system is simplified, thereby optimizing the management network structure of the server system, simplifying the management of the resources in the server system, and improving resource management efficiency.

Description

Resource management method, device and system

The present application is based on a Chinese patent application filed on Jan. 18, 2016, the entire disclosure of which is hereby incorporated by reference.

Technical field

The present invention relates to the field of server resource management technologies, and in particular, to a resource management method, apparatus, and system.

Background technique

With the development of cloud computing, the existing data center requires a large number of cabinet servers to meet the data processing requirements. In order to improve the resource utilization efficiency of the intensive application on the traditional cabinet server, the RSA (Rack Scale Architecture) is mainly used at present. The architecture is implemented by merging resources such as the CPU, memory, storage, and network of the server into resource pools, and selecting corresponding resources from the resource pools according to service requests to perform business processing.

The RSA architecture manages POD (Plain Old Data, a data structure) object through PODM (Plain Old Data Manager). PODM manages multiple PODML (PODM Local, local data structure management nodes); each PODML manages DMC (Drawer Management Controller) in units of RACK (each layer on the architecture), and each PODML can be Facing multiple DMCs; DMC manages BMC (Baseboard Management Controller) in units of DRAWER (drawer); BMC is based on MODULE (module). Thus, each level faces multiple sub-level units, for example, one PODML faces multiple DMCs at the same time, and is responsible for resource discovery and assembly of multiple DMCs.

The RSA system architecture identifies the location of each resource by LID (Location ID, location serial number), and sets the level of each level in the production or installation. The LID is specifically composed of POD-RACK (rack)-DRAWER-MODULE-SUBMODULE (sub-module), for example, the number of DRAWER, which can be read from the rack backplane through GPIO (General-Purpose Input/Output).

Therefore, it can be seen that the RSA system architecture is identified by the LID, the hierarchical concept is complex, and the software architecture is complex. With the continuous expansion of the current data processing scale, more and more server racks, in order to identify the location of each resource, it is necessary to strictly plan the routing to ensure the location of the resources, management difficulties, and the installation personnel number is difficult, error-prone.

Due to the complex hierarchical concept of LID, the RSA can only be extended by attaching a new management layer to the bottom layer. For example, when integrating multiple blades, only one BLADE layer can be added under MODULE to extend LID. . If a node in the LID fails, all resources under it are unavailable, resulting in redundancy and instability of the RSA hierarchy, and no targeting to the hardware application form.

Therefore, the existing server resource pool management level is complex and the openness is poor, which makes management difficult.

Summary of the invention

The main object of the present invention is to provide a resource management method, device and system, which aim to solve the technical problem that the existing server resource management is difficult.

To achieve the above objective, the present invention provides a resource management method, and the resource management method includes the following steps:

The management node obtains resource information linked by each control node;

Generating an assembly request according to the received service request and the resource information;

And controlling, according to the assembly request, the control node to assemble a logical server to process the service request.

Preferably, the step of the management node acquiring the resource information of each control node includes:

Receiving a configuration request of each control node, and assigning a dynamic address to the control node;

Obtaining registration information that is sent by the control node based on the dynamic address, and registering the control node;

Obtaining resource information that is linked by the registered control node.

Preferably, the step of generating an assembly request according to the received service request and the resource information includes:

Filtering the resource information according to the received service request;

Obtaining the service resource and obtaining the information of the service resource according to the preset weight algorithm and the filtered resource information;

An assembly request is generated based on the service resource information.

Preferably, the controlling the node to assemble the logical server according to the assembly request, and processing the service request includes:

Sending the assembly request to the control node of the service resource hook;

a control node that controls the connection of the service resource to assemble a logical server according to the assembly request;

Controlling the logical server to process the service request.

Preferably, after the step of controlling the control node to assemble the logical server and processing the service request according to the assembly request, the method further includes:

Receiving a revocation request of the service request;

Determining, by the control node that the service resource is attached, the logical server according to the revocation request;

The logical server is revoked, and the execution step is performed: the management node acquires resource information linked by each control node.

In addition, in order to achieve the above object, the present invention further provides a resource management apparatus, where the resource management apparatus includes:

a collection module, configured to acquire, by the management node, resource information linked by each control node;

An assembly module, configured to generate an assembly request according to the received service request and the resource information;

And a processing module, configured to control, according to the assembly request, the control node to assemble a logical server to process the service request.

Preferably, the collecting module comprises:

An allocating unit, configured to receive a configuration request of each control node, and allocate a dynamic address to the control node;

a registration unit, configured to acquire registration information that is sent by the control node based on the dynamic address, and register the control node;

And a collecting unit, configured to acquire resource information that is linked by the registered control node.

Preferably, the assembly module comprises:

a filtering unit, configured to filter the resource information according to the received service request;

a weighting unit, configured to select a service resource according to a preset weight algorithm and filtered resource information, and obtain information about the service resource;

And a generating unit, configured to generate an assembly request according to the service resource information.

Preferably, the processing module includes:

a requesting unit, configured to send the assembly request to a control node that is attached to the service resource;

An assembly unit, configured to control a control node of the service resource to assemble a logical server according to the assembly request;

a processing unit, configured to control the logical server to process the service request.

Preferably, the resource management device further includes:

a receiving module, configured to receive a revocation request of the service request;

a splitting module, configured to: according to the revocation request, control a control node that is connected to the service resource to split the logical server;

The undo module is used to revoke the logical server.

In addition, to achieve the above object, the present invention also provides a resource management system, where the resource management system includes a management node and a control node, where:

The management node includes a collection module, an assembly module, a processing module, a receiving module, a split module, and an undo module;

The control node is configured to register with the management node; collect the resource information of the connection node and report it to the management node; assemble the logical server according to the assembly request sent by the management node; and send according to the management node The undo request splits the logical server.

A resource management method, device and system according to an embodiment of the present invention obtains resource information linked by each control node through a management node; generates an assembly request according to the received service request and resource information; and controls the control node according to the assembly request Assemble logical servers to process business requests. The invention directly interacts with the management node for resource management by the control node, reduces the complexity of the management software, simplifies the networking and wiring of the server system, realizes the optimization of the server system management network structure, and the resources of the server system. Management is easier and improves the efficiency of resource management.

DRAWINGS

1 is a schematic flowchart of a first embodiment of a resource management method according to the present invention;

2 is a schematic flowchart of a second embodiment of a resource management method according to the present invention;

3 is a schematic diagram of a management structure according to an embodiment of the present invention;

4 is a schematic flowchart of a third embodiment of a resource management method according to the present invention;

FIG. 5 is a schematic flowchart diagram of a fourth embodiment of a resource management method according to the present invention; FIG.

FIG. 6 is a schematic flowchart diagram of a fifth embodiment of a resource management method according to the present invention;

FIG. 7 is a schematic diagram of functional modules of a first embodiment of a resource management apparatus according to the present invention; FIG.

8 is a schematic diagram of functional modules of a second embodiment of a resource management apparatus according to the present invention;

9 is a schematic diagram of functional modules of a third embodiment of a resource management apparatus according to the present invention;

10 is a schematic diagram of functional modules of a fourth embodiment of a resource management apparatus according to the present invention;

FIG. 11 is a schematic diagram of functional modules of a fifth embodiment of a resource management apparatus according to the present invention; FIG.

FIG. 12 is a schematic block diagram of a first embodiment of a resource management system according to the present invention.

The implementation, functional features, and advantages of the present invention will be further described in conjunction with the embodiments.

detailed description

It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The main solution of the embodiment of the present invention is: the management node acquires the resource information of each control node; generates an assembly request according to the received service request and the resource information; and controls the control node assembly according to the assembly request. a logical server that processes the business request

Due to the complexity of the resource management architecture of the prior art server system and the poor openness, management is difficult.

The invention provides a solution for attaching resources to a control node, performing resource management through interaction between the management node and the control node, optimizing the management structure, simplifying the management level of the server resources, and improving the resource management efficiency.

Referring to FIG. 1 , a first embodiment of a resource management method of the present invention provides a resource management method, where the resource management method includes:

Step S10: The management node acquires resource information linked by each control node.

The present invention is mainly applied to resource management. In this embodiment, the resource management method of the cabinet server system is used as an example. The resource management method proposed by the present invention simplifies the networking and wiring of the existing cabinet server, optimizes the management network, and improves the cabinet server system. Practicality.

Specifically, as an implementation manner, first, a management network of the rack server system is initialized.

The management network of the cabinet server system includes a management node and a control node. The management node is a manager, such as a PODM (POD Manager, POD Data Management Node), as a management terminal, managing server resources and logical servers, such as resource discovery, allocation, and recycling; the control node is a controller, such as a DMC (Drawer Management Controller, drawer) The management controller is used as a switching node of the management network for discovering and assembling the resources it contains, and collaborating with the management node by the client (client) to complete resource management; the control node may further include a network switch. For example, a ToR switch (Top of Rack) can be used to manage and exchange network resources of a server. Of course, a ToR switch can also not use resources as a control node to manage network resources. Used for network switching.

In this embodiment, the management node is a PODM, the control node is a DMC, and the ToR switch is used only for network switching. The control node can include multiple DMC and ToR switches.

When the management network is initialized, the DMC switch module is started and the ToR switch is started. Then, the DMC switch module and the ToR switch perform election and convergence to form an acyclic tree network. Any node can be interconnected with other nodes to ensure that any node fails to reconverge to form an acyclic tree network and improve the management network. Reliability.

The management node can be attached to any node of the acyclic tree network for management.

The management node periodically collects resource information from each control node, and each control node updates its resource status in real time to ensure that the resource information acquired by the management node is accurate resource information. Alternatively, each control node actively sends local resource information to the management node, and the management node receives resource information of each control node and stores it.

It should be noted that each control node actively collects various local resource information that is connected to it. The resources connected to each control node include the processor, memory, storage, and network resources, which can be flexibly set according to actual needs. The resource information collected by each control node includes the location identifier of each resource, the current usage status, and the processing capability.

Thereby, the management node acquires resource information of each control node. The resource information acquired by the management node includes a location identifier of each resource, a connected control node, a current usage state, and a processing capability.

Then, the management node combines various resources according to the resource information to form a resource pool, for example, a CPU resource pool, a memory resource pool, a storage resource pool, and a network resource pool, and is used to process the services into a logical server according to service requirements.

The management node may store the resource information of each control node in a CMDB (Configuration Management Database), and perform data update according to the subsequently obtained control node resource information.

Step S20: Generate an assembly request according to the received service request and the resource information.

After acquiring the resource information of each control node, the management node performs resource assembly according to the service request.

Specifically, as an implementation manner, first, the management node receives a service request, for example, a telecommunication service, and may send the information to a management node through a terminal such as a remote PC (personal computer). A service request may carry information such as a service type, a resource required for processing a service, and a specification requirement for each resource, and may be flexibly set according to actual needs.

Then, the management node may obtain, according to the type of the service request, the resource specification requirements corresponding to the current service request from the preset resource allocation table, and then select, from the acquired resource information, the resources that meet the conditions; or The management node may select various resources that meet the requirements from the acquired resource information according to the requirements of the resource carried by the service request.

Therefore, the management node obtains various resources suitable for processing the current service request, and uses the obtained resources as the service resource for processing the service request. According to the selected service resource, the management node can assemble a logical server for processing the current service request.

The management node then generates an assembly request based on the selected service resource. The assembly request includes each service resource information, that is, the selected resource information required for assembling the logical server, such as the location identifier of each service resource, the connected control node, and the connection relationship.

Step S30: Control, according to the assembly request, the control node to assemble a logical server to process the service request.

After obtaining the assembly request, the management node causes the control node to assemble the logical server to process the service request according to the assembly request.

Specifically, as an implementation manner, first, the management node sends an assembly request to the control node connected to each service resource according to each service resource information in the assembly request, and sends the assembly request to the selected resource. Connected control node.

After each control node receives the assembly request, the management node controls each control node to find, according to the assembly request, the resources selected by the control nodes and the resources selected for assembling the logical server, that is, the service resources.

Then, each control node interconnects the service resources located on different control nodes according to the assembly request, and assembles the logical server. The DMC interacts with the BMC (Baseboard Management Controller) inside the subnet to access specific resources.

For example, by controlling the node's switchboard, set the SAS expander to complete the assembly of the CPU and storage. It should be noted that the interaction protocol between the DMC and the BMC may be an IPMI (Intelligent Platform Management Interface) protocol, or may be an HTTP (HyperText Transfer Protocol) or other interactive protocol. Flexible settings according to actual needs.

Then, the management node or the remote control machine deploys an operating system, cloud platform software, and other software for business processing for the assembled logical server according to the requirements of the current business processing.

After the software deployment is completed, the management node controls the logical server to process the current service request.

The logical server acts as a physical machine and processes the service through various service resources.

Subsequently, the management node or the remote control machine can control the control node to which each service resource belongs in the logical server according to the requirements of the current service processing, and perform operations such as powering on, powering off, and resetting the logical server. For example, when the logical server needs to be powered on, the management node sends a reset operation request to the control node attached to the service resource according to the information of the service resource, and after receiving the power-on operation request, the control node performs corresponding operation. Electrical operation.

In this embodiment, the management node acquires resource information hanged by each control node; generates an assembly request according to the received service request and resource information; and then, the management node controls the control node to assemble the logical server according to the assembly request, and processes the current request. Business request. In this embodiment, the concept of the rack level of the traditional server is avoided, and resources are connected to each control node, and the control node directly interacts with the management node for resource management, which reduces the complexity of the management software and simplifies the networking of the server system. Wiring realizes the optimization of the server system management network structure, manages the resources of the server system more easily, and improves the efficiency of resource management.

Further, referring to FIG. 2, a second embodiment of the resource management method of the present invention provides a resource management method. Based on the foregoing embodiment shown in FIG. 1, the step S10 includes:

Step S11: Receive a configuration request of each control node, and allocate a dynamic address to the control node.

After the server management platform is started, first, the management network of the rack server system is initialized.

When the management network is initialized, the DMC switch module is started and the ToR switch is started. Then, the DMC switch module and the ToR switch perform election and convergence to form an acyclic tree network that communicates with each other and does not affect each other. In the formed acyclic tree network, any node can be interconnected with other nodes to ensure that any node fails, and can reconverge to form an acyclic tree network, which improves the reliability of the management network. At the same time, the acyclic tree network blocks the loop and ensures the load balancing of the nodes.

The DMC and the ToR switch run the MSTP (Multiple Spanning Tree Protocol). The DMC switch module and the ToR switch are equal nodes. The DMC node has the attached resources, but the ToR switch does not.

The generated acyclic tree network can refer to Figure 3. Each DMC or ToR switch is interconnected with at least three other nodes. After a network port or connection fails, MSTP reconverges to form a new acyclic tree network. The 3x redundancy on the 3x reliability. It should be noted that any node can be interconnected as much as possible according to the engineering design during wiring, and the number of interconnected nodes can be flexibly set according to actual needs.

The management node PODM can be attached to any node of the acyclic tree network, and the acyclic tree network Any node can continue to increase the attached DMC.

After completing the management network initialization, the management node receives the configuration request of each control node, and assigns a dynamic address to the control node.

As an implementation manner, the management node can run a DHCPD (Dynamic Host Configuration Protocol Daemon) service and receive a DHCP (Dynamic Host Configuration Protocol) request sent by the control node.

After powering on, the control node actively sends a DHCP request to the management node to request allocation of a dynamic address.

Then, the management node assigns a dynamic address, such as IP, addr (address), to the control node according to the received control node DHCP request. The dynamic address assigned by the management node is set with the lease duration. When the control node is abnormal or the lease is used, the dynamic address of the control node is recovered, and the recovered dynamic address can be assigned to other control nodes. It should be noted that the lease duration can be flexibly set according to actual needs.

The management node then sends the assigned dynamic address to the corresponding control node.

Step S12: Acquire registration information sent by the control node based on the dynamic address, and register the control node.

After the management node assigns a dynamic address, it acquires registration information of the control node and registers the control node.

Specifically, as an implementation manner, after receiving the dynamic address allocated by the management node, the control node uses the dynamic address as an address for subsequent communication with the management node.

The control node actively sends registration information to the management node for registration. The registration information of the control node includes the host name and current running status of the control node. The current running state of the control node includes information such as whether the control node is currently running normally and the resource occupancy rate.

The management node receives the registration request sent by the control node based on the dynamic address, and acquires the registration information of the control node. Thus, the management node can learn the control node information using the current dynamic address.

Then, the management node registers each control node for registration, saves the dynamic address and host name of each control node, caches the current state of each control node, and establishes a correspondence relationship for identifying different control nodes.

Step S13: Obtain resource information that is linked by the registered control node.

After the control node is powered on, it starts local resource discovery and actively collects the resources it hangs. The resources connected to each control node include the processor, memory, storage, and network resources, which can be flexibly set according to actual needs.

The resource information collected by each control node includes the location identifier of each resource, the current usage status, and the processing capability.

As an implementation manner, the resource discovered by the control node may use a control node host name and a UUID (Universally Unique Identifier) as the location identifier. The control node host name is the host name of the control node to which the resource is attached; the UUID is a pre-configured identification identifier, and the UUID can be set according to the factory information and the playing time of the resource. Of course, the UUID can also be set according to other hardware information of the resource. Flexible settings according to actual needs.

In addition, within the control node, location information such as SLED (slot) can be added, and the location identifier of the resource can be added to further identify the physical location of the resource, so that resources can be conveniently and quickly found for processing when a hardware failure occurs.

For the management node, only the host name and the UUID need to identify different resources, and it is not necessary to control the location information inside the node.

As a result, the architecture of the server system is simplified, and the hierarchical architecture is more targeted for hardware applications. When the capital occurs When the source is faulty, firstly, according to the hook relationship of the faulty resource, the connected control node is found by the host name, and then the location information inside the control node is used to quickly locate the physical location of the faulty resource, so as to quickly locate the resource location. At the same time, the complexity of the management software is simplified, and the actual operation difficulty is small, which avoids confusion and errors that are easy to occur when locating resources through strict wiring.

After transmitting the registration request for registration, each control node actively sends the local resource information to the management node, and the management node receives the resource information of each control node and stores it; or the management node periodically collects resource information from the registered control nodes. Each control node updates its resource status in real time to ensure that the resource information acquired by the management node is accurate resource information.

Thereby, the management node acquires resource information of each registered control node. The resource information acquired by the management node includes the location identifier of each resource, the connected control node, the current usage status, and the processing capability.

The management node may associate the host name of the control node with each resource information of the control node to identify resources of different control nodes. Therefore, the management node can find the resource information that is connected to each control node according to the host name, and perform data query or update of the control node resource information.

The management node stores the resource information of each control node in the CMDB. It should be noted that the CMDB storing the resource information may be configured according to the actual situation such as the amount of stored data and the size of the environment.

In this embodiment, the management node receives the configuration request of each control node, and assigns a dynamic address to the control node; then, acquires registration information sent by the control node based on the dynamic address, and registers the control node; and then acquires the registered control node Resource information. In this embodiment, the management node allocates a dynamic address to the control node, registers the control node, effectively improves the address usage efficiency and the management efficiency of the management node, and avoids the resource occupied by the failure control node; at the same time, the management node can timely acquire the resources of the effective control node. Information, data is orderly and regular, and it is convenient for subsequent management of resources.

Further, referring to FIG. 4, a third embodiment of the resource management method of the present invention provides a resource management method, which is based on the foregoing embodiment shown in FIG. 1 or FIG. 2 (this embodiment is illustrated by using FIG. 2). S20 includes:

Step S21: Filter the resource information according to the received service request.

Then, the management node filters out the resource pools required for processing the current service request from the stored resource pools according to the required resources for processing the service request.

Then, according to the resource requirements and the resource information of the service request, the resources in each resource pool are filtered to filter out the resources that meet the requirements. For example, if the service request limits the number of cores of the multi-core CPU, the management node filters out the CPU in the CPU resource pool, the number of cores does not meet the requirements, that is, filters out CPU resource information that does not meet the requirements; the service request can also limit the amount of memory storage. , the management node according to the parameters set by the business request, Filter out the memory resources in the memory resource pool that do not meet the parameter requirements, that is, filter out the memory resource information that does not meet the requirements; the service request can also limit the network traffic, and the management node filters out the network resource pool according to the parameters set by the service request. Network resources that do not meet the parameter requirements, that is, filter out network resource information that does not meet the requirements.

Therefore, the management node filters out the resources in the resource pool that cannot meet the service processing requirements according to the service request, that is, filters out the resource information that cannot meet the service processing requirements.

The resource information obtained after filtering meets the requirements for processing the current service request.

Step S22: Select a service resource and obtain information about the service resource according to the preset weight algorithm and the filtered resource information.

After filtering the resource information, the management node selects the resource as the service resource of the assembly logical server according to the preset weight algorithm and the filtered resource information.

Specifically, as an implementation manner, first, the management node selects resources suitable for assembling the logical server according to the preset weight algorithm and the filtered resource information.

For example, if the preset weight algorithm is set with a distance parameter between the CPU and the memory, the management node selects the CPU and the memory that meet the requirements according to the distance parameter and the resource information, and avoids the data transmission abnormality caused by the CPU and the memory being too far apart.

It should be noted that the preset weight algorithm can be flexibly set according to actual conditions, and is used to set parameter conditions, so that the logical server obtained by the acquired resource combination can achieve performance balance or optimal performance such as processing speed and storage occupancy.

Thereby, the management node obtains resources for assembling the logical server, and uses each obtained resource as a service resource of the assembly logical server. According to the selected service resource, the management node can assemble a logical server for processing the current service request.

Then, the management node obtains information of each service resource. The information of each service resource includes: the location identifier of each service resource, the connected control node, the current usage status, and the processing capability, and can be flexibly set according to actual needs.

Step S23: Generate an assembly request according to the service resource information.

After obtaining the service resource information, the management node generates an assembly request according to the service resource information.

The assembly request includes the information of each service resource, that is, the selected resource information required for assembling the logical server, such as the location identifier of each service resource, the connected control node, and the connection relationship.

In this embodiment, the management node filters the resource information according to the received service request; then, according to the preset weight algorithm and the filtered resource information, the service resource is selected and the service resource information is obtained; and then the service resource information is obtained according to the service resource information. , generate an assembly request. This embodiment implements the optimization of assembling logical server resources, simplifies the process of selecting resources, and selects the obtained service resources to process service requests more efficiently.

Further, referring to FIG. 5, the fourth embodiment of the resource management method of the present invention provides a resource management method. Based on the foregoing embodiment shown in FIG. 4, the step S30 includes:

Step S31: Send the assembly request to the control node to which the service resource is attached.

Specifically, as an implementation manner, first, the management node separately sends the assembly request to the control node that is connected to each service resource according to each service resource information in the assembly request.

The assembly request sent by the management node includes resource information selected as a service resource in each control node, a connection relationship of each service resource, and the like.

Step S32: The control node that controls the connection of the service resource assembles the logical server according to the assembly request.

After the assembly request is sent to the control node to which the service resource is attached, as an implementation manner, first, the management node controls each control node to find the resources hooked by each control node according to the assembly request, and is selected for assembly. The resource of the logical server, that is, the service resource.

Then, each control node interconnects the service resources located on different control nodes according to the assembly request, and assembles them into a logical server.

The control node DMC interacts with the BMC inside the subnet to access specific resources. It should be noted that the interaction protocol between the DMC and the BMC may be the IPMI protocol, the HTTP protocol, or other interaction protocols, and may be flexibly set according to actual needs.

Thereby, the assembled logical server is obtained. A logical server consists of a connection of service resources.

Step S33: Control the logical server to process the service request.

After the logical server is assembled, the management node or the remote controller deploys an operating system, cloud platform software, and other software for business processing for the assembled logical server according to the requirements of the current business processing.

Subsequently, the management node or the remote control machine can control the control nodes connected to the service resources in the logical server according to the requirements of the current service processing, and perform operations such as powering on, powering off, and resetting the logical server. For example, when the logical server needs to be powered on, the management node sends a power-on operation request to the control node connected to the service resource according to the information of the service resource, and after receiving the power-on operation request, the control node performs corresponding Power on operation.

In this embodiment, the management node sends the assembly request to the control node to which the service resource is attached; then, the control node attached to the control service resource assembles the logical server according to the assembly request; and then controls the logical server to process the service request. In this embodiment, the assembly request is sent to the control node that is connected to the service resource, and the corresponding control node is controlled to connect to the service resource, complete the assembly of the logical server, and then process the service request. This embodiment simplifies the assembly process of resources in the logical server and improves resource management efficiency.

Further, referring to FIG. 6, the fifth embodiment of the resource management method of the present invention provides a resource management method. After the step S30, the method further includes:

Step S40: Receive a revocation request of the service request.

When the service request processing is performed, or after the business processing request is completed, if the management node receives the revocation request for the service request, the processing of the service is stopped, and the logical server that processes the service is revoked.

Specifically, as an implementation manner, the remote control machine may issue a revocation request for the service request according to actual needs, and cancel the processing for the service.

The management node receives the revocation request sent by the remote controller. The revocation request carries the business information that is revoked by the request.

Step S50: Control the node connected to the service resource to split the logical server according to the revocation request.

After receiving the revocation request, as an implementation manner, first, the management node acquires the service requested to be revoked according to the revocation request.

Then, the management node obtains the logical server information for processing the service according to the service that is requested to be revoked, and obtains the information of each service resource.

Then, the management node acquires the control node to which each service resource is attached according to the information of each service resource, and sends the revocation request to the control node to which each service resource is attached.

After receiving the revocation request, the control node performs a power-off operation on the service resource according to the revocation request, and splits the resources that constitute the logical server.

Thus, the management node implements splitting of logical server resources.

Step S60, the logical server is revoked, and the process proceeds to step S10.

After the logical server resource is split, the management node revokes the logical server and stops the business processing of the logical server.

Then, the management node sends a resource collection request to each control node, and collects the current resource information of each control node; each control node updates the resource information of the next node in real time and reports it to the management node.

After receiving the resource information of each control node, the management node updates the resource information corresponding to each control node according to the host name of each control node to implement the update of the CMDB.

It should be noted that the management node may update the resource information of all the control nodes in the current management network, or may only update the resource information of the control node of the split logical server, and may be flexibly set according to actual needs.

In this embodiment, the management node receives the revocation request for the service request; then, according to the revocation request, controls the control node to which the service resource is attached to split the logical server; and revokes the logical server, and then updates the resource information, Used for subsequent assembly of other logical servers. In this embodiment, by splitting the logical server, releasing resources and updating resource information, the resource use efficiency is improved, and the released resources can be used for processing other services, ensuring order management of resources, and simplifying management of resources. .

Referring to FIG. 7, a first embodiment of the resource management apparatus of the present invention provides a resource management apparatus, where the resource management apparatus includes:

The collecting module 100 is configured to acquire, by the management node, resource information that is connected to each control node.

The present invention is mainly applied to resource management. The resource management device provided by the present invention exemplifies the networking and cabling of the existing cabinet server, optimizes the management network, and improves the cabinet server system. Practicality.

The management network of the cabinet server includes a resource management device and a control node. The resource management device is a manager, such as a PODM (POD Manager, POD Data Resource Management Device), as a management terminal, managing server resources and logical servers, such as resource discovery, allocation, and recycling; the control node is a controller, such as DMC (Drawer Management Controller) , drawer management controller), as a switching node of the management network, for discovering and assembling the resources it contains, and collaborating with the resource management device by the client (client) to complete resource management; the control node may also include A network switch, such as a ToR switch (Top of Rack), can be connected to resources to manage and exchange network resources of the server. Of course, the ToR switch can also be used as a control node to manage the network. Resources, only for network switching.

In this embodiment, the resource management device is a PODM, the control node is a DMC, and the ToR switch is used only for network switching. The control node can include multiple DMC and ToR switches.

The resource management device can be connected to any node of the acyclic tree network for management.

The collection module 100 of the resource management device periodically collects resource information from each control node, and each control node updates the state of the resource that is connected to it in real time to ensure that the resource information acquired by the collection module 100 is accurate resource information. Alternatively, each control node actively sends the local resource information to the resource management device, and the collection module 100 of the resource management device receives the resource information of each control node and stores the resource information.

Thereby, the collection module 100 acquires resource information of each control node. The resource information acquired by the collection module 100 includes a location identifier of each resource, a connected control node, a current usage state, a processing capability, and the like.

Then, the collection module 100 combines various resources according to the resource information to form a resource pool, such as a CPU resource pool, a memory resource pool, a storage resource pool, and a network resource pool, and is combined into a logical server according to service requirements when processing the service.

The collection module 100 may store the resource information of each control node in a CMDB (Configuration Management Database), and perform data update according to the subsequently acquired control node resource information.

The assembly module 200 is configured to generate an assembly request according to the received service request and the resource information.

After acquiring the resource information of each control node, the assembly module 200 performs resource assembly according to the service request.

Specifically, as an implementation manner, first, the assembly module 200 receives a service request, such as a telecommunication service, and can be sent to the assembly module 200 through a terminal such as a remote PC (personal computer). A service request may carry information such as a service type, a resource required for processing a service, and a specification requirement for each resource, and may be flexibly set according to actual needs.

Then, the assembly module 200 may obtain, according to the type of the service request, the resource specification requirements corresponding to the current service request from the preset resource allocation table, and then select, from the acquired resource information, the resources that meet the conditions; Or the assembly module 200 can select various resources that meet the requirements from the acquired resource information according to the requirements of the resources carried by the service request.

Therefore, the assembly module 200 obtains various resources suitable for processing the current service request, and uses the obtained resources as the service resource for processing the service request.

The assembly module 200 then generates an assembly request based on the selected service resource. The assembly request includes each service resource information, that is, the selected resource information required for assembling the logical server, such as the location identifier of each resource, the connected control node, and the connection relationship of each resource.

The processing module 300 is configured to control, according to the assembly request, the control node to assemble a logical server to process the service request.

After the assembly request is obtained, the processing module 300 causes the control node assembly logic server to process the service request according to the assembly request.

Specifically, as an implementation manner, first, the processing module 300 sends an assembly request to the control node that is connected to each service resource according to each service resource information in the assembly request, and sends the assembly request to the selected resource. The attached control node.

After each control node receives the assembly request, the processing module 300 controls each control node to find, according to the assembly request, the resources selected by the control nodes to be used for assembling the logical server, that is, the service resources.

Then, the processing module 300 or the remote control machine deploys an operating system, cloud platform software, and other software for business processing for the assembled logical server according to the requirements of the current business processing.

After the software deployment is completed, the processing module 300 controls the logical server to process the current service request.

Subsequently, the processing module 300 or the remote control machine can control the control node to which each service resource belongs in the logical server according to the requirements of the current service processing, and perform operations such as power-on, power-off, and reset of the logical server. For example, when the logical server needs to be powered on, the processing module 300 sends a power-on operation request to the control node that is connected to the service resource according to the information of the service resource, and after receiving the power-on operation request, the control node performs corresponding Power-on operation.

In this embodiment, the collection module 100 acquires resource information linked by each control node; the assembly module 200 generates an assembly request according to the received service request and resource information; and then, the processing module 300 controls the control node assembly logic according to the assembly request. Server, processing this business request. In this embodiment, the concept of the rack level of the traditional server is avoided, and resources are connected to each control node, and the control node directly interacts with the resource management device for resource management, which reduces the complexity of the management software and simplifies the networking of the server system. And wiring, the server system management network structure is optimized, the management of each resource in the server system is more convenient, and the efficiency of resource management is improved.

Further, referring to FIG. 8, the second embodiment of the resource management apparatus of the present invention provides a resource management apparatus. Based on the foregoing embodiment, the collection module 100 includes:

The allocating unit 110 is configured to receive a configuration request of each control node, and allocate a dynamic address to the control node.

When the management network is initialized, the DMC switch module is started and the ToR switch is started. Then, the DMC switch module and the ToR switch perform election and convergence to form an acyclic tree network that communicates with each other and does not affect each other. In In the formed acyclic tree network, any node can be interconnected with other nodes to ensure that any node fails, and can reconverge to form an acyclic tree network to improve the reliability of the management network. At the same time, the acyclic tree network blocks the loop and ensures the load balancing of the nodes.

The resource management device PODM can be attached to any node of the acyclic tree network, and any node of the acyclic tree network can continue to increase the attached DMC.

After completing the management network initialization, the allocation unit 110 of the resource management apparatus receives the configuration request of each control node, and assigns a dynamic address to the control node.

As an implementation manner, the allocating unit 110 may run a DHCPD (Dynamic Host Configuration Protocol Daemon) service to receive a DHCP (Dynamic Host Configuration Protocol) request sent by the control node.

After powering on, the control node actively sends a DHCP request to the allocating unit 110 to request allocation of a dynamic address.

Then, the allocating unit 110 assigns a dynamic address, such as IP, addr (address), to the control node according to the received control node DHCP request. The dynamic address allocated by the allocating unit 110 is set with a lease duration. When an abnormality occurs in the control node or the lease period is used, the dynamic address of the control node is recovered, and the recovered dynamic address can be allocated to other control nodes. It should be noted that the lease duration can be flexibly set according to actual needs.

The allocation unit 110 then sends the assigned dynamic address to the corresponding control node.

The registration unit 120 is configured to acquire registration information that is sent by the control node based on the dynamic address, and register the control node.

After the allocation unit 110 assigns the dynamic address, the registration unit 120 acquires the registration information of the control node and registers the control node.

Specifically, as an implementation manner, after receiving the dynamic address allocated by the allocating unit 110, the control node uses the dynamic address as an address for subsequent communication with the resource management apparatus.

The control node actively sends registration information to the registration unit 120 for registration. The registration information of the control node includes the host name and current running status of the control node. The current running state of the control node includes information such as whether the control node is currently running normally and the resource occupancy rate.

The registration unit 120 receives the registration request sent by the control node based on the dynamic address, and acquires registration information of the control node. Thus, the registration unit 120 can learn the control node information using the current dynamic address.

Then, the registration unit 120 registers each control node that performs registration, saves the dynamic address and host name of each control node, caches the current state of each control node, and establishes a correspondence relationship for identifying different control nodes.

The collecting unit 130 is configured to acquire resource information that is linked by the registered control node.

For the resource management device, only different resources need to be identified by the host name and the UUID, and it is not necessary to control the location information inside the node.

As a result, the architecture of the server system is simplified, and the hierarchical architecture is more targeted for hardware applications. When a resource fault occurs, firstly, according to the hook relationship of the faulty resource, the mounted control node is found by the host name, and then the location information inside the control node is used to quickly locate the physical location of the faulty resource, thereby realizing the rapid location of the resource. Positioning, while simplifying the complexity of the management software, the practical operation is less difficult, avoiding the confusion and errors that are easy to occur when locating resources through strict wiring.

After sending the registration request for registration, each control node actively sends the local resource information to the collecting unit 130, and the collecting unit 130 receives the resource information of each control node and stores it; or the collecting unit 130 periodically collects the collected control nodes. The resource information is updated in real time by each control node to ensure that the resource information acquired by the collecting unit 130 is accurate resource information.

Thereby, the collecting unit 130 acquires resource information of each registered control node. The resource information acquired by the collecting unit 130 includes a location identifier of each resource, a connected control node, a current usage state, a processing capability, and the like.

The collecting unit 130 may establish a correspondence between the host name of the control node and each resource information that is connected to the control node, and is used to identify resources of different control nodes. Therefore, the collecting unit 130 can search for resource information linked to each control node according to the host name, and perform data query or update of the control node resource information.

Then, the collecting unit 130 combines the various types of resources according to the resource information to form a resource pool, for example, a CPU resource pool, a memory resource pool, a storage resource pool, and a network resource pool, and is used to process the services into a logical server according to the business requirements.

The collecting unit 130 stores the resource information of each control node in the CMDB. It should be noted that the CMDB storing the resource information may be configured according to the actual situation such as the amount of stored data and the size of the environment.

In this embodiment, the allocating unit 110 receives the configuration request of each control node, and allocates a dynamic address to the control node; then, the registration unit 120 acquires the registration information sent by the control node based on the dynamic address, and registers the control node; then, the collecting unit 130 Obtain the resource information of the registered control node. In this embodiment, the resource management device allocates a dynamic address to the control node, registers the control node, effectively improves the address usage efficiency and the management efficiency of the resource management device, and avoids the resource occupied by the failure control node; at the same time, the resource management device can acquire the effective control node in time. The resource information, the data is orderly and regular, and it is convenient for subsequent management of resources.

Further, referring to FIG. 9, the third embodiment of the resource management apparatus of the present invention provides a resource management apparatus, which is based on the embodiment shown in FIG. 7 or FIG. 8 (this embodiment is illustrated by FIG. 8). Module 200 includes:

The filtering unit 210 is configured to filter the resource information according to the received service request.

Specifically, as an implementation manner, first, the filtering unit 210 receives a service request, such as a telecommunication service, and may send the information to the resource management device through a terminal such as a remote PC (personal computer). A service request may carry information such as a service type, a resource required for processing a service, and a specification requirement for each resource, and may be flexibly set according to actual needs.

Then, the filtering unit 210 filters out the resource pools required for processing the current service request from the stored resource pools according to the required resources for processing the service request.

Then, the filtering unit 210 filters the resources in each resource pool according to the resource requirements and the resource information according to the service request, and filters out the resources that meet the requirements. For example, if the service request limits the number of cores of the multi-core CPU, the filtering unit 210 filters out the CPU in the CPU resource pool, and the number of cores does not meet the requirements, that is, filters out CPU resource information that does not meet the requirements; the service request may also limit the memory storage. The filtering unit 210 filters out the memory resources in the memory resource pool that do not meet the parameter requirements according to the parameters set by the service request, that is, filters out the memory resource information that does not meet the requirement; the service request can also limit the network traffic, and the filtering unit The network resource of the network resource pool that does not meet the parameter requirement is filtered out according to the parameter set by the service request, that is, the network resource information that does not meet the requirement is filtered out.

Therefore, the filtering unit 210 filters out resources in the resource pool that cannot meet the service processing requirements according to the service request, that is, filters out resource information that cannot meet the service processing requirements.

The weighting unit 220 is configured to select a service resource and obtain information about the service resource according to the preset weight algorithm and the filtered resource information.

After the resource information is filtered, the weighting unit 220 selects the resource as the service resource of the assembly logical server according to the preset weight algorithm and the filtered resource information.

Specifically, as an implementation manner, first, the weight unit 220 selects resources suitable for assembling the logical server according to the preset weight algorithm and the filtered resource information.

For example, if the preset weight algorithm is configured with a distance parameter between the CPU and the memory, the weight unit 220 selects the CPU and the memory that meet the requirements according to the distance parameter and the resource information, and avoids the data transmission abnormality caused by the CPU and the memory being too far apart.

Thereby, the weight unit 220 obtains resources for assembling the logical server, and uses each obtained resource as a service resource of the assembly logical server. According to the selected service resources, a logical server can be assembled to process the current service request.

Then, the weight unit 220 acquires information of each service resource. The information of each service resource includes: the location identifier of each service resource, the connected control node, the current usage status, and the processing capability, and can be flexibly set according to actual needs.

The generating unit 230 is configured to generate an assembly request according to the service resource information.

After acquiring the service resource information, the generating unit 230 generates an assembly request based on the service resource information.

The assembly request includes information of each service resource, that is, selected resource information required for assembling the logical server, such as location identifier of each service resource, connected control node, connection relationship of each resource, and the like. .

In this embodiment, the filtering unit 210 filters the resource information according to the received service request. Then, the weighting unit 220 selects the service resource and obtains the information of the service resource according to the preset weight algorithm and the filtered resource information. Then, the generating unit 230 generates an assembly request based on the service resource information. This embodiment implements the optimization of assembling logical server resources, simplifies the process of selecting resources, and selects the obtained service resources to process service requests more efficiently.

Further, referring to FIG. 10, a fourth embodiment of the resource management apparatus of the present invention provides a resource management apparatus. The processing module 300 includes:

The requesting unit 310 is configured to send the assembly request to the control node that is attached to the service resource.

After the assembly request is obtained, the processing module 300 causes the control node to assemble the logical server to process the service request according to the assembly request.

Specifically, as an implementation manner, first, the requesting unit 310 separately sends an assembly request to the control node to which each service resource is attached according to each service resource information in the assembly request.

The assembly request sent by the requesting unit 310 includes resource information selected as a service resource in each control node, a connection relationship of each service resource, and the like.

The assembling unit 320, the control node for controlling the service resource hooking, assembles the logical server according to the assembly request.

After the assembly request is sent to the control node to which the service resource is attached, as an implementation manner, first, the assembly unit 320 controls each control node to find, according to the assembly request, the resources that are connected to each control node, and is selected to be used for Assemble the resources of the logical server, that is, find the service resources.

The processing unit 330 is configured to control the logical server to process the service request.

After the logical server is assembled, the processing unit 330 or the remote control machine deploys an operating system, cloud platform software, and other software for business processing for the assembled logical server according to the requirements of the current business processing.

After the software deployment is completed, the processing unit 330 controls the logical server to process the current service request.

Subsequently, the processing unit 330 or the remote control machine can control the control nodes connected to the service resources in the logical server according to the requirements of the current service processing, and perform operations such as power-on, power-off, and reset of the logical server. For example, when the logical server needs to be powered on, the processing unit 330 sends a power-on operation request to the control node that is connected to the service resource according to the information of the service resource, and after receiving the power-on operation request, the control node performs corresponding Power-on operation.

In this embodiment, the requesting unit 310 sends an assembly request to the control node to which the service resource is attached; then, The assembly unit 320 controls the control node to which the service resource is attached to assemble the logical server according to the assembly request; then, the processing unit 330 controls the logical server to process the service request. In this embodiment, the assembly request is sent to the control node that is connected to the service resource, and the corresponding control node is controlled to connect to the service resource, complete the assembly of the logical server, and then process the service request. This embodiment simplifies the assembly process of resources in the logical server and improves resource management efficiency.

Further, referring to FIG. 11, the fifth embodiment of the resource management apparatus of the present invention provides a resource management apparatus. The resource management apparatus further includes:

The receiving module 400 is configured to receive a revocation request of the service request.

When the service request processing is performed, or after the business processing request is completed, if the receiving module 400 receives the revocation request for the service request, the processing of the service is stopped, and the logical server that processes the service is revoked.

The receiving module 400 receives the revocation request sent by the remote controller. The revocation request carries the business information that is revoked by the request.

The splitting module 500 is configured to, according to the revocation request, control a control node that is connected to the service resource to split the logical server.

After receiving the revocation request, as an implementation manner, first, the splitting module 500 acquires the service requested to be revoked according to the revocation request.

Then, the splitting module 500 obtains the logical server information for processing the service according to the service that is requested to be revoked, and obtains the information of each service resource.

Then, the splitting module 500 acquires the control node to which each service resource is attached according to the information of each service resource, and sends the revocation request to the control node to which each service resource is attached.

Thus, the split module 500 implements splitting of logical server resources.

The revocation module 600 is configured to revoke the logical server.

After completing the splitting of the logical server resources, the undo module 600 revokes the logical server and stops the business processing of the logical server.

Then, the collection module 100 sends a resource collection request to each control node, and collects the current resource information of each control node; each control node updates the resource information of the control node in real time, and reports the information to the collection module 100.

After receiving the resource information of each control node, the collection module 100 updates the resource information corresponding to each control node according to the host name of each control node to update the CMDB.

It should be noted that the collection module 100 may update the resource information of all the control nodes in the current management network, or may only update the resource information of the control node of the split logical server, and may be flexibly set according to actual needs.

In this embodiment, the receiving module 400 receives the revocation request for the service request; then, the splitting module 500 controls the control node connected to the service resource to split the logical server according to the revocation request; and the revocation module 600 revokes the The logical server then collects the module 100 to update the resource information for subsequent assembly of other logical servers. In this embodiment, by splitting the logical server, releasing the resource and updating the resource information, the resource use efficiency is improved, and the released resource can be used for processing other services, thereby ensuring the orderly management of the resource. The management of resources has been reduced.

Referring to FIG. 12, a first embodiment of the resource management system of the present invention provides a resource management system, where the resource management system includes a management node A and a control node B, where:

The management node A includes, for example, a collection module 100, an assembly module 200, a processing module 300, a receiving module 400, a split module 500, and an undo module 600.

The management network includes a management node A and a control node B. Management node A is a manager, such as PODM (POD Manager, POD Data Management Node), as a management terminal, management server resources and logical servers, such as resource discovery, allocation, and recycling; control node B is a controller, such as DMC (Drawer Management Controller) , drawer management controller), as a switching node of the management network, for discovering and assembling the resources it contains, and collaborating with the management node A by the client (client) to complete resource management; the control node B can also Including network switches, such as ToR switches (Top of Rack), ToR switches can be used to manage and exchange network resources of servers. Of course, ToR switches can also be used as control nodes. Network resources, only for network switching.

In this embodiment, the management node A is a PODM, the control node B is a DMC, and the ToR switch is used only for network switching. Control Node B may include multiple DMC and ToR switches.

Management node A can be attached to any node of the acyclic tree network for management.

The management node A collection module 100 periodically collects resource information from each control node, and each control node B updates its resource status in real time to ensure that the resource information acquired by the collection module 100 is accurate resource information. Or the collection module 100 of the management node A receives the resource information of each control node and stores it. It should be noted that each control node B actively collects various local resource information that is connected to it, and actively sends the local resource information to the management node A.

Thereby, the collection module 100 acquires resource information of each control node. The resource information acquired by the collection module 100 includes a location identifier of each resource, a connected control node B, a current usage state, a processing capability, and the like.

Then, the assembly module 200 receives the service request, and obtains the resource specification requirements corresponding to the current service request from the preset resource allocation table according to the type of the service request, and then selects the resource information from the acquired resource information. Each resource that meets the requirements; or the assembly module 200 may select various resources that meet the requirements from the acquired resource information according to the requirements of the resource carried by the service request.

Therefore, the assembly module 200 obtains various resources suitable for processing the current service request, and uses the obtained resources as the service resource for processing the service request. Then, the assembly module 200 generates an assembly request according to the selected service resource information. The assembly request includes each service resource information, that is, the selected resource information required for assembling the logical server, such as the location identifier of each service resource, the connected control node, and the connection relationship of each service resource.

Then, the processing module 300 sends the assembly request to the control node B to which each service resource is attached according to each service resource information in the assembly request.

After each control node B receives the assembly request, the processing module 300 controls each control node B to find, according to the assembly request, the resources selected by the control nodes and the resources selected for assembling the logical server, that is, the service resources.

Then, each control node B connects the service resources located on different control nodes to each other according to the assembly request, and assembles the logical server.

After the business process is completed, if the remote control machine issues a revocation request, the receiving module 400 receives the revocation request.

Then, the splitting module 500 acquires the service that is requested to be revoked according to the revocation request, and obtains the logical server information for processing the service according to the service that is revoked by the current request, and obtains the information of each service resource.

Then, the splitting module 500 acquires the control node B to which each service resource is attached according to the information of each service resource, and sends the revocation request to the control node B to which each service resource is attached.

After receiving the revocation request, the control node B performs a power-off operation on the service resource according to the revocation request, and splits the resources that constitute the logical server.

Then, the undo module 600 revokes the logical server and stops the processing service of the logical server.

The control node B is configured to register with the management node A; collect the resource information of the connection node B and report it to the management node A; assemble the logical server according to the assembly request sent by the management node A; The revocation request sent by the management node A splits the logical server.

After power-on, the control node B actively sends a DHCP request to the management node A to request allocation of a dynamic address. Upon receiving the dynamic address assigned by the management node A, the control node B sends a registration request to the management node A to register.

At the same time, the control node B actively collects various local resource information linked thereto. The resources that are connected to each control node B include the processor, memory, storage, network resources, etc., and can be flexibly set according to actual needs. The resource information collected by each control node B includes location identifiers, current usage states, processing capabilities, and the like of the resources.

After the registration is completed, each control node B actively sends local resource information to the management node A, and the management node A receives the resource information of each control node and stores it; or the management node A periodically collects resource information from each control node B, each The control node B updates the state of the resource under its control in real time to ensure that the resource information acquired by the management node A is accurate resource information.

If the control node B receives the assembly request sent by the management node A, it searches for the resources of the attached logical server, that is, the service resource, according to the assembly request. Then, the control node B connects the service resources located on the different control nodes B to each other according to the assembly request, and assembles them into a logical server.

If the control node B receives the revocation request from the management node A, it searches for the service resource in the resource to which it is attached according to the revocation request, and then performs a power-off operation on the service resource, the logical server.

In this embodiment, the resource management system includes a management node A and a control node B. The control node B discovers and collects the information of the resources that are hanged and reports the information to the management node A. The management node A hangs the control node B. The resources are merged to form a resource pool, and then the service resources are filtered out from the resource pool according to the service request, and the control node B assembles the service resources into a logical server for service processing. After the service processing is completed, the management node A can split the logical server and release the resources according to the revocation request. In this embodiment, the complexity of the management software is reduced, and the networking and wiring of the server system are simplified, and the resource management is simplified by the management node and the control node, and the plane structure of the server resource management is optimized, and each of the server systems is optimized. The management of resources is easier and the efficiency of resource management is improved.

The modules or units in the apparatus provided by the embodiments of the present application may pass through one or more digital signal processors (DSPs), application specific integrated circuits (ASICs), processors, microprocessors, controllers, microcontrollers, and on-site Implemented by a programming array (FPGA), programmable logic device, or other electronic unit, or any combination thereof. Some of the functions or processes described in this application embodiment may also be implemented by software executing on a processor.

For example, the embodiment of the present invention further provides a resource management device, for example, the device can be applied to a management terminal, including:

processor;

a memory for storing processor executable instructions;

Wherein the processor is configured to:

Obtaining resource information linked by each control node;

Industrial applicability

The invention can be applied to the field of server resource management technology. The embodiment of the present invention obtains the resource information of each control node by using the management node; generates an assembly request according to the received service request and the resource information; and according to the assembly request, the control node assembles the logical server to process the service request. In this way, the control node can directly interact with the management node for resource management, which reduces the complexity of the management software, and simplifies the networking and wiring of the server system, realizes the optimization of the server system management network structure, and makes the resources in the server system. The management is easier and the efficiency of resource management is improved.

The above are only the preferred embodiments of the present invention, and are not intended to limit the scope of the invention, and the equivalent structure or equivalent process transformations made by the description of the present invention and the drawings are directly or indirectly applied to other related technical fields. The same is included in the scope of patent protection of the present invention.

Claims

A resource management method, wherein the resource management method comprises the following steps:

The management node obtains resource information linked by each control node;

Generating an assembly request according to the received service request and the resource information;

And controlling, according to the assembly request, the control node to assemble a logical server to process the service request.
The resource management method according to claim 1, wherein the step of the management node acquiring the resource information of each control node comprises:

Receiving a configuration request of each control node, and assigning a dynamic address to the control node;

Obtaining registration information that is sent by the control node based on the dynamic address, and registering the control node;

Obtaining resource information that is linked by the registered control node.
The resource management method according to claim 1 or 2, wherein the step of generating an assembly request according to the received service request and the resource information comprises:

Filtering the resource information according to the received service request;

Obtaining the service resource and obtaining the information of the service resource according to the preset weight algorithm and the filtered resource information;

An assembly request is generated based on the service resource information.
The resource management method according to claim 3, wherein the controlling the control node to assemble the logical server according to the assembly request, and processing the service request comprises:

Sending the assembly request to the control node of the service resource hook;

a control node that controls the connection of the service resource to assemble a logical server according to the assembly request;

Controlling the logical server to process the service request.
The resource management method according to claim 4, wherein the step of controlling the control node to assemble the logical server and processing the service request according to the assembly request further comprises:

Receiving a revocation request of the service request;

Determining, by the control node that the service resource is attached, the logical server according to the revocation request;

The logical server is revoked, and the execution step is performed: the management node acquires resource information linked by each control node.
A resource management device, wherein the resource management device comprises:

The collection module is configured to obtain the resource information of each control node by the management node;

An assembly module, configured to generate an assembly request according to the received service request and the resource information;

The processing module is configured to control the control node to assemble the logical server to process the service request according to the assembly request.
The resource management device of claim 6, wherein the collection module comprises:

An allocating unit, configured to receive a configuration request of each control node, and allocate a dynamic address to the control node;

a registration unit, configured to acquire registration information sent by the control node based on the dynamic address, and register the control node;

The collecting unit is configured to obtain resource information that is linked by the registered control node.
The resource management device according to claim 6 or 7, wherein the assembly module comprises:

a filtering unit, configured to filter the resource information according to the received service request;

a weight unit, configured to select a service resource according to a preset weight algorithm and filtered resource information, and obtain information about the service resource;

The generating unit is configured to generate an assembly request according to the service resource information.
The resource management device of claim 8, wherein the processing module comprises:

a requesting unit, configured to send the assembly request to a control node that is attached to the service resource;

An assembly unit, the control node configured to control the service resource hooking, assembling the logical server according to the assembly request;

a processing unit configured to control the logical server to process the service request.
The resource management device of claim 9, wherein the resource management device further comprises:

a receiving module, configured to receive a revocation request of the service request;

a splitting module, configured to: according to the revocation request, control a control node that is connected to the service resource to split the logical server;

The undo module is set to revoke the logical server.
A resource management system, wherein the resource management system includes a management node and a control node, wherein:

The management node includes the apparatus of any one of claims 6-10;

The control node is configured to register with the management node; collect the resource information of the connection node and report it to the management node; assemble the logical server according to the assembly request sent by the management node; and send according to the management node The undo request splits the logical server.