WO2015018023A1

WO2015018023A1 - Method, device and system for scheduling resources

Info

Publication number: WO2015018023A1
Application number: PCT/CN2013/081064
Authority: WO
Inventors: 王烽
Original assignee: 华为技术有限公司
Priority date: 2013-08-08
Filing date: 2013-08-08
Publication date: 2015-02-12
Also published as: CN105359489A

Abstract

A method, device and system for scheduling resources. The method includes: determining a source physical machine and a target physical machine of a virtual machine migratory operation, according to information of physical machines in a physical machine cluster; sending a first negotiation message to the source physical machine; receiving a response message of the first negotiation message sent by the source physical machine, wherein the response message of the first negotiation message carries the information determined by the source physical machine, of the virtual machine to be migrated; migrating the virtual machine to be migrated from the source physical machine to the target physical machine, according to the information determined by the source physical machine, of the virtual machine to be migrated.

Description

Resource scheduling method, device and system

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

Background technique

In cloud computing technology, the energy consumed by the physical machine cluster is the main energy overhead of running the cluster. Because these physical machines are idle for a long time, in order to reduce the energy consumed by the physical machine cluster, the industry usually migrates and integrates the virtual machines, and the idle physical machines are powered off and powered off to save energy. When the load rises, power on the physical machine.

In the prior art, in the resource scheduling system of the physical machine cluster, the physical machine is powered on and off through a centralized global management module, thereby adjusting the energy consumed by the physical machine cluster. For example, by deploying a centralized energy management module in a physical machine cluster, powering on and off the physical machine to adjust the energy consumed by the physical machine cluster; or by DPM (Dynamic Power Module) and DRS (Dynamic Resource Scheduling) , Distributed Resource Scheduling) The module cooperates to power on and off the physical machine to adjust the energy consumed by the physical machine cluster.

However, in the resource scheduling system of the physical machine cluster, the energy management module or the DPM and the DRS module are centralized deployment modules, and the centralized deployment module assumes all scheduling tasks, and needs to run information and operations of all physical machines in the physical machine cluster. The information of the virtual machine on the physical machine is collected into the centralized deployment module, thereby making the centralized deployment module a bottleneck.

Summary of the invention

An embodiment of the present invention provides a resource scheduling method, apparatus, and system, which can negotiate resource scheduling of a physical machine cluster through a central management node and a physical machine, thereby reducing the burden of collecting a large amount of information by the central management node.

In order to achieve the above object, embodiments of the present invention use the following technical solutions: In a first aspect, an embodiment of the present invention provides a resource scheduling method, including: determining, according to information of a physical machine of a physical machine cluster, a source physical machine and a target physical machine of a virtual machine migration operation;

Sending a first negotiation cancellation to the source physical machine, 3⁄4;

Receiving a response message of the first negotiation message sent by the source physical machine, where the first information is: ', ', and according to the information of the virtual machine to be migrated determined by the source physical machine, the virtual machine to be migrated is The source physical machine migrates to the target physical machine.

In a first possible implementation of the first aspect,

Determining, according to the information of the physical machine of the physical machine cluster, the source physical machine of the virtual machine migration operation, including:

Determining a light-load physical machine in the physical machine cluster according to information of a physical machine of the physical machine cluster;

Transmitting, to the light-loading physical machine, a second negotiation message, where the second negotiation message is used to query whether the light-loading physical machine is allowed to perform a power-off operation;

A light-loading physical machine that allows a power-off operation is selected as the source physical machine.

In conjunction with the first aspect of the foregoing or the first possible implementation of the first aspect, in a second possible implementation,

Determining the overloaded physical machine in the physical machine cluster according to the information of the physical machine of the physical machine cluster;

Sending, to the overloaded physical machine, a third negotiation message, where the third negotiation message is used to query whether the overloaded physical machine is allowed to perform a subtraction operation;

A heavy physical machine that allows the offload operation is selected as the source physical machine.

In combination with the foregoing first aspect or any one of the first possible implementation to the second possible implementation of the first aspect, in a third possible implementation Before the machine, the method further includes:

Sending a fourth negotiation message to the target physical machine, where the fourth negotiation message carries information about the virtual machine to be migrated determined by the source physical machine;

Receiving a response message of the fourth negotiation message sent by the target physical machine, where the response message of the fourth negotiation message carries information about the virtual machine to be migrated accepted by the target physical machine;

Correspondingly, the virtual machine to be migrated accepted by the target physical machine is migrated from the source physical machine to the target physical machine according to the response message of the fourth negotiation message.

With reference to the foregoing first aspect or the third possible implementation manner of the foregoing aspect, in a fourth possible implementation, the determining, by the physical machine information of the physical machine cluster, the target physical machine of the virtual machine migration operation , including:

Determining, according to information of the physical machine of the physical machine cluster, a first target physical machine; sending, to the first target physical machine, a fifth negotiation message, where the fifth negotiation message is used to query whether the first target physical machine is Accept virtual machine migration;

A physical machine that accepts the virtual machine migration in the first target physical machine is selected as the target physical machine.

With the fourth possible implementation of the first aspect, in a fifth possible implementation, before the selecting the first target physical machine, the method further includes:

At least one physical machine is powered on, and the at least one physical machine that is powered on is used as the first target physical machine.

In a sixth possible implementation manner of the first aspect, the method further includes: if the virtual machine on the source physical machine is completely migrated to the target physical machine, performing power-off operation on the source physical machine .

In a seventh possible implementation manner of the first aspect, before the determining the source physical machine and the target physical machine of the virtual machine migration operation, the method further includes:

Acquiring real-time resource occupation information of the physical machine of the physical machine cluster, or acquiring a resource scheduling request message sent by the physical machine of the physical machine cluster, where the resource scheduling request message is sent by a light load or a heavy load physical machine;

The determining the virtual machine migration according to the information of the physical machine of the physical machine cluster The source physical machine and the target physical machine of the operation, including:

Determining the source physical machine and the target physical machine according to real-time resource occupation information of the physical machine of the physical machine cluster or the resource scheduling request message.

In a second aspect, the embodiment of the present invention further provides a resource scheduling method, including: receiving a first negotiation message sent by a central management node;

Sending a response message of the first negotiation message to the central management node, where the response message of the first negotiation message carries information of the virtual machine to be migrated, so that the central management node performs information according to the information of the virtual machine to be migrated. Respond to the migration of virtual machines for migration.

In a first possible implementation manner of the second aspect, after the receiving, by the receiving center, the first negotiation message sent by the node, the method further includes:

Determining information of the virtual machine to be migrated according to the first negotiation message and the first basis, where the first basis includes characteristics of the physical machine, application characteristics of the virtual machine running on the physical machine, and operation information of the virtual machine Or the scheduling constraints of the virtual machine.

With reference to the foregoing second aspect or the first possible implementation manner of the foregoing aspect, in a second possible implementation manner, before the receiving the first negotiation message sent by the central management node, the method further includes:

Sending a resource scheduling request message to the central management node; or

Sending its own resource occupation information to the central management node.

In a third aspect, an embodiment of the present invention provides a central management node, including: a determining unit, configured to determine, according to information of a physical machine of a physical machine cluster, a source physical machine and a target physical machine of a virtual machine migration operation;

a sending unit, configured to send, to the source physical machine determined by the determining unit, a first negotiation message;

a receiving unit, configured to receive a response message of the first negotiation message sent by the source physical machine, the first negotiation cancellation, the packet, the response cancellation, the packet, and the information of the virtual machine to be migrated determined by the source physical machine ;

a migration unit, configured to be determined according to the source physical machine received by the receiving unit The target physical machine.

In a first possible implementation of the third aspect,

The determining unit is specifically configured to determine, according to the information of the physical machine of the physical machine cluster, a light-load physical machine in the physical machine cluster, and select a light-load physical machine that is allowed to be powered off as the source physical machine. ;

The sending unit is further configured to send a second negotiation message to the light-loading physical machine in the physical machine cluster determined by the determining unit, where the second negotiation message is used to query the light load in the physical machine cluster Whether the physical machine is allowed to perform the power-off operation.

In conjunction with the foregoing third aspect or the first possible implementation of the third aspect, in a second possible implementation manner,

The determining unit is specifically configured to determine, according to information about a physical machine of the physical machine cluster, a heavy-duty physical machine in the physical machine cluster, and select a heavy-duty physical machine that allows a reduced-off operation as the source physical machine;

The sending unit is further configured to send a third negotiation message to the overloaded physical machine in the physical machine cluster determined by the determining unit, where the third negotiation message is used to query an overload in the physical machine cluster Whether the physical machine is allowed to perform the burden reduction operation.

In combination with any of the foregoing third aspect or the first possible implementation to the second possible implementation of the third aspect, before the third possible implementation of the target physical machine:

The sending unit is further configured to send a fourth negotiation message to the target physical machine, where the fourth negotiation message carries information about the virtual machine to be migrated determined by the source physical machine; and the receiving unit is further configured to receive a response message of the fourth negotiation message sent by the target physical machine, where the response message of the fourth negotiation message carries information about the virtual machine to be migrated accepted by the target physical machine;

Correspondingly, the migrating unit is configured to: migrate, according to the response message of the fourth negotiation message received by the receiving unit, the virtual machine to be migrated accepted by the target physical machine from the source physical machine to the Target physical machine.

In combination with the third aspect of the foregoing or the third possible implementation of the third aspect, In the fourth possible implementation,

The determining unit is specifically configured to select a first target physical machine according to information of a physical machine according to the physical machine cluster, and select a physical machine that accepts a virtual machine to move in the first target physical machine as the target Physical machine

The sending unit is further configured to send a fifth negotiation message to the first target physical machine determined by the determining unit, where the fifth negotiation message is used to query whether the first target physical machine accepts virtual machine migration .

In conjunction with the fourth possible implementation of the third aspect, in a fifth possible implementation manner,

The determining unit is further configured to: power on at least one physical machine before selecting the first target physical machine, and use the at least one physical machine that is powered on as the first target physical machine.

In a sixth possible implementation manner of the third aspect, the target physical machine is further configured to perform a power-off operation on the source physical machine.

In a seventh possible implementation of the third aspect,

The receiving unit is further configured to acquire real-time resource occupation information of a physical machine of the physical machine cluster, or obtain a resource scheduling request message sent by a physical machine of the physical machine cluster, where the resource scheduling request message is lightly loaded or Reloading physical machine to send;

The determining unit is specifically configured to determine the source physical machine and the target physical machine according to the real-time resource occupation information of the physical machine of the physical machine cluster acquired by the receiving unit or the resource scheduling request message.

In a fourth aspect, an embodiment of the present invention provides a physical machine, including:

a receiving unit, configured to receive a first negotiation message sent by the central management node, and a sending unit, configured to send a response message of the first negotiation message to the central management node, where the response message of the first negotiation message carries the virtual machine to be migrated The information is such that the central management node migrates the virtual machine to be migrated according to the information of the virtual machine to be migrated.

In a first possible implementation manner of the fourth aspect, the physical machine further includes: a determining unit, configured to determine information about the virtual machine to be migrated according to the first negotiation message and the first basis received by the receiving unit, where the first basis includes characteristics of a physical machine, and running on a physical machine The application characteristics of the virtual machine, the running information of the virtual machine, or the scheduling constraints of the virtual machine.

In conjunction with the foregoing fourth aspect or the first possible implementation of the fourth aspect, in a second possible implementation manner,

The sending unit is further configured to: before the receiving unit receives the first negotiation message sent by the central management node, send the resource scheduling request cancellation packet, or the resource occupancy information of the self to the central management node.

In a fifth aspect, an embodiment of the present invention provides a central management node, including: a processor, a memory, a bus, and a communication interface;

The memory is configured to store a computer execution instruction, the processor is connected to the memory through the bus, and when the central management node is running, the processor executes the computer execution instruction stored in the memory, to The central management node is caused to perform the resource scheduling method according to any one of claims 1-8.

In a sixth aspect, an embodiment of the present invention provides a readable medium, including computer execution instructions, when the processor of a central management node executes the computer execution instruction, the central management node executes any one of claims 1-8 The resource scheduling method.

In a seventh aspect, an embodiment of the present invention provides a physical machine, including: a processor, a memory, a bus, and a communication interface;

The memory is configured to store a computer execution instruction, the processor is connected to the memory through the bus, and when the physical machine is running, the processor executes the computer execution instruction stored in the memory, so that The physical machine performs the resource scheduling method according to any one of claims 9 to 11.

According to an eighth aspect, an embodiment of the present invention provides a readable medium, including computer execution instructions, when the processor of the physical machine executes the computer to execute an instruction, the physical machine executes any one of claims 9-1 The resource scheduling method described.

According to a ninth aspect, an embodiment of the present invention provides a resource scheduling system, including: a central management node, configured to determine a virtual state according to information of a physical machine of a physical machine cluster After the source physical machine and the target physical machine of the migration operation, send a first negotiation message to the source physical machine, and receive a response message of the first negotiation message sent by the source physical machine, where the first negotiation message is Responding to the packet, carrying the information of the virtual machine to be migrated determined by the source physical machine, and transmitting the source physical machine to the receiving center management node according to the information of the virtual machine to be migrated determined by the source physical machine The first negotiation cancels, and sends a response message of the first negotiation message to the central management node;

The target physical machine is configured to receive the migration of the virtual machine to be migrated.

In a first possible implementation of the ninth aspect,

The central management node is specifically configured to determine, according to information of the physical machine of the physical machine cluster, a light-load physical machine in the physical machine cluster, and send a second to the light-load physical machine in the physical machine cluster The negotiation message is used to query whether the light-load physical machine in the physical machine cluster is allowed to perform a power-off operation, and select a light-load physical machine that allows the power-off operation to be the source physical machine.

In conjunction with the foregoing ninth aspect or the first possible implementation manner of the ninth aspect, in a second possible implementation manner,

The central management node is specifically configured to determine, according to information about a physical machine of the physical machine cluster, a heavy physical machine in the physical machine cluster, and send a third to a heavy physical machine in the physical machine cluster. Negotiating the message, the third negotiation message is used to query whether the overloaded physical machine in the physical machine cluster allows the offloading operation, and selects the overloaded physical machine that allows the offloading operation to be the source physical machine.

With reference to the foregoing ninth aspect or any one of the first possible implementation manner of the ninth aspect to the second possible implementation manner, in a third possible implementation manner, to the target physical machine The fourth negotiation message is sent to the target physical machine, where the fourth negotiation message carries information about the virtual machine to be migrated determined by the source physical machine, and receives the fourth negotiation sent by the target physical machine. a response message of the message, where the response message of the fourth negotiation message carries the virtual machine to be migrated accepted by the target physical machine Information;

Correspondingly, the central management node is configured to migrate the virtual machine to be migrated accepted by the target physical machine from the source physical machine to the target physical machine according to the response message of the fourth negotiation message.

In combination with the foregoing ninth aspect or the third possible implementation manner of the ninth aspect, in a fourth possible implementation manner,

The central management node is configured to select a first target physical machine according to the information of the physical machine of the physical machine cluster, and send a fifth negotiation message to the first target physical machine, where the fifth negotiation message is used. Querying whether the first target physical machine accepts virtual machine migration, and selecting a physical machine in the first target physical machine that accepts virtual machine migration as the target physical machine.

With reference to the fourth possible implementation manner of the ninth aspect, in a fifth possible implementation manner,

The central management node is further configured to power on at least one physical machine before selecting the first target physical machine, and use the at least one physical machine that is powered on as the first target physical machine.

In a sixth possible implementation manner of the ninth aspect, the target physical machine is further configured to perform a power-off operation on the source physical machine.

In a seventh possible implementation of the ninth aspect,

The central management node is further configured to acquire real-time resource occupation information of the physical machine of the physical machine cluster, or obtain a resource scheduling request message sent by the physical machine of the physical machine cluster, where the resource scheduling request message is lightly loaded. Or the overloaded physical machine sends the network management node, where the central management node is configured to determine the source physical machine and the target according to the real-time resource occupation information of the physical machine of the physical machine cluster or the resource scheduling request message. Physical machine

The source physical machine is further configured to send a resource scheduling request message to the central management node before receiving the first negotiation message sent by the central management node.

In an eighth possible implementation of the ninth aspect, The source physical machine is further configured to determine, according to the first negotiation message and the first basis, information about the virtual machine to be migrated, where the first basis includes a feature of the physical machine, and a virtual machine running on the physical machine. Application characteristics, operational information of the virtual machine, or scheduling constraints of the virtual machine.

The embodiment of the present invention provides a resource scheduling method, device, and system. After determining a source physical machine and a target physical machine of a virtual machine migration operation according to the information of the physical machine of the physical machine cluster, the first negotiation message is sent to the source physical machine. And receiving the response message of the first negotiation message sent by the source physical machine, where the response message of the first negotiation message carries the information of the virtual machine to be migrated determined by the source physical machine, and the information of the virtual machine to be migrated determined according to the source physical machine, Migrate the virtual machine to be migrated from the source physical machine to the target physical machine. With this solution, the central management node only needs to initially determine the source physical machine of the virtual machine migration, and then negotiate with the source physical machine to determine which virtual machines on the source physical machine need to be migrated, and finally perform resource scheduling of the physical machine cluster, thereby reducing The central management node needs to collect a large amount of virtual machine information on the physical machine.

DRAWINGS

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only It is some embodiments of the invention.

FIG. 1 is a structural diagram 1 of a resource scheduling system according to an embodiment of the present invention;

2 is a flowchart 1 of a resource scheduling method according to an embodiment of the present invention; FIG. 3 is a flowchart 2 of a resource scheduling method according to an embodiment of the present invention; FIG. 4 is a resource according to an embodiment of the present invention; FIG. 5 is a schematic structural diagram 1 of a central management node according to an embodiment of the present invention; FIG. 6 is a schematic structural diagram 1 of a physical machine according to an embodiment of the present invention;

FIG. 7 is a schematic structural diagram 2 of a physical machine according to an embodiment of the present disclosure;

8 is a schematic structural diagram 2 of a central management node according to an embodiment of the present invention; FIG. 9 is a schematic structural diagram 3 of a physical machine according to an embodiment of the present invention;

10 is a structural diagram 2 of a resource scheduling system according to an embodiment of the present invention; FIG. 1 is a third structural diagram of a resource scheduling system according to an embodiment of the present invention; FIG. 12 is a structural diagram 4 of a resource scheduling system according to an embodiment of the present invention.

detailed description

The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are only a part of the embodiments of the present invention, but not all embodiments.

Embodiment 1

As shown in FIG. 1 , a resource scheduling system architecture diagram is provided in an embodiment of the present invention. The resource scheduling system includes a physical machine cluster composed of multiple or one physical machine and a central management node, which is three in this embodiment. For example, a physical machine may be virtualized by virtualized software on each physical machine. Each physical machine is deployed with a distributed agent. The distributed agent can be used to collect attribute information and running information of the physical machine of the local machine. And the information of each virtual machine on the physical machine of the local machine, the central management node is used to collect the resource occupation information of each physical machine of the physical machine cluster that it manages, and determine the physical machine cluster according to the resource occupation information of each physical machine. a resource optimization policy, or a resource scheduling request for collecting a physical machine of the physical machine cluster that it manages, and determining a resource optimization policy of the physical machine cluster according to the collected resource scheduling request, that is, determining which physical machine can be powered off to save Resource consumption, which physical machine can accept the migration of the virtual machine, and according to the determined physics Optimizing resource cluster on a distributed agent strategy and the corresponding physical machine consultation, access to information to determine its own physical machine migration of virtual machines to be moved to accept the information and virtual machines, and migration of virtual machines based on the information obtained.

According to the resource scheduling system shown in FIG. 1, as shown in FIG. 2, the embodiment of the present invention provides a resource scheduling method, which is related to the central management node side, and the method may include:

S 1 01. The central management node determines the source physical machine and the target physical machine of the virtual machine migration operation according to the information of the physical machine of the physical machine cluster.

The resource scheduling system provided by the embodiment of the present invention includes a central management node and a physical machine of a physical machine cluster managed by the central management node. The central management node can communicate with the physical machine of the physical machine cluster it manages, and obtain the information of the physical machine of the physical machine cluster, including: resource occupation information of the physical machine of the physical machine cluster, for example, the CPU of the physical machine (C en Tral P rocessi ng Un it , central processing unit), memory, network The current utilization of the resource, or the load information of the physical machine; or the resource scheduling request message sent by the physical machine of the physical machine cluster, that is, the physical machine of the physical machine cluster requests the central management node to perform resource scheduling.

Exemplarily, when some physical load of the physical machine cluster is light, the central management node can migrate the virtual machines running on the light-load physical machines, and power off the light-load physical machines to save energy. When some physical machines in a physical machine cluster are heavily loaded, the central management node can migrate virtual machines running on these overloaded physical machines to other light-loaded physical machines, thereby causing virtual machines running on physical machines in the physical machine cluster. Balanced, Qo S (Qua lity of Service) is applied to complete intelligent resource scheduling.

After the central management node obtains the information of the physical machine of the physical machine cluster, the central management node may determine the source physical machine of the virtual machine migration operation according to the information of the physical machine of the physical machine cluster. In addition, the target physical machine can also be determined. Specifically, the source physical machine is a physical machine that needs to move the virtual machine out, and the target physical machine is a physical machine that the virtual machine migrates into.

Optionally, the central management node can determine a single source physical machine and a single target physical machine; a single source physical machine and multiple target physical machines; multiple source physical machines and a single target physical machine; or multiple source physical machines and multiple The target physical machine, the invention is not limited.

51 02. The central management node sends a first negotiation message to the source physical machine.

After the central management node determines the source physical machine, the central management node sends the first negotiation message to the source physical machine.

The central management node receives the response message of the first negotiation message sent by the source physical machine, and the response message of the first negotiation message carries the information of the virtual machine to be migrated determined by the source physical machine.

After the central management node sends the first negotiation message to the source physical machine, the central management node may receive the response message of the first negotiation message sent by the source physical machine, where the response message of the first negotiation message carries the to-migration determined by the source physical machine. Information about the virtual machine. The virtual machine to be migrated is a virtual machine running on the source physical machine.

It should be noted that the central management node may send a first negotiation message to the source physical machine for negotiation, so as to obtain information about the virtual machine to be migrated determined by the source virtual machine. 5 1 04. The central management node migrates the VM to be migrated from the source physical machine to the target physical machine according to the information of the virtual machine to be migrated determined by the source physical machine.

After the central management node receives the information about the virtual machine to be migrated determined by the source physical machine, the central management node may migrate the virtual machine to be migrated from the source physical machine to the central management according to the information of the virtual machine to be migrated determined by the source physical machine. The target physical machine determined by the node completes the migration operation of the virtual machine to be migrated.

According to the resource scheduling system shown in FIG. 1, as shown in FIG. 3, the embodiment of the present invention further provides a resource scheduling method, which is related to the source physical machine side, and the method may include:

5 2 0 1. The source physical machine receives the first negotiation message sent by the central management node.

The source physical machine receives the first negotiation message from the central management node, wherein the source physical machine is determined by the central management node according to the information of the physical machine of the physical machine cluster.

S2 0 2, the source physical machine sends a response message of the first negotiation message to the central management node, and the response message of the first negotiation message carries the information of the virtual machine to be migrated, so that the central management node according to the letter of the virtual machine to be migrated, Migrate the migrated virtual machine.

After the source physical machine receives the first negotiation message sent by the central management node, the source physical machine may return a response message of the first negotiation message to the central management node, so that the central management node can determine the virtual machine to be migrated according to the source physical machine. The message is to be migrated to the virtual machine to be migrated, and the response message of the first negotiation message carries the information of the virtual machine to be migrated determined by the source physical machine.

The embodiment of the present invention provides a resource scheduling method, after determining a source physical machine and a target physical machine of a virtual machine migration operation according to the information of the physical machine of the physical machine cluster, sending a first negotiation message to the source physical machine, and receiving the source The response message of the first negotiation message sent by the physical machine, the response message of the first negotiation message carries the information of the virtual machine to be migrated determined by the source physical machine, and the information of the virtual machine to be migrated determined according to the source physical machine, and the virtual to be migrated The machine migrates from the source physical machine to the target physical machine. With this solution, the central management node only needs to initially determine the source physical machine of the virtual machine migration, and then negotiate with the source physical machine to determine which virtual machines on the source physical machine need to be migrated, and finally perform resource scheduling of the physical machine cluster, thereby reducing The central management node needs to collect a large amount of virtual machine information on the physical machine. Embodiment 2

The embodiment of the invention provides a resource scheduling method. As shown in FIG. 4, the method may include:

S 3 0 1. The central management node acquires real-time resource occupation information of the physical machine of the physical machine cluster, or acquires a resource scheduling request message sent by the physical machine of the physical machine cluster, and the resource scheduling request message is sent by the light load or the overloaded physical machine.

The resource scheduling system provided by the embodiment of the present invention includes a central management node and a physical machine of a physical machine cluster managed by the central management node. The central management node can communicate with the physical machine of the physical machine cluster that it manages, and obtain the information of the physical machine of the physical machine cluster, including: resource occupation information of the physical machine of the physical machine cluster, such as the CPU and memory of the physical machine. The current utilization of the network resource, or the load information of the physical machine; or the resource scheduling request message sent by the physical machine of the physical machine cluster, that is, the physical machine of the physical machine cluster requests the central management node to perform resource scheduling. The resource scheduling request message is sent by a light or overloaded physical machine.

Exemplarily, when some physical load of the physical machine cluster is light, the central management node can migrate the virtual machines running on the light-load physical machines, and power off the light-load physical machines to save energy. When some physical machines in a physical machine cluster are heavily loaded, the central management node can migrate virtual machines running on these overloaded physical machines to other light-loaded physical machines, thereby causing virtual machines running on physical machines in the physical machine cluster. Balance, ensure the application of Q o S, and then complete intelligent resource scheduling.

It should be noted that when the current utilization of the CPU, memory, and network resources of some physical machines of the physical machine cluster is low, the physical machines may send a resource scheduling request message to the central management node to notify the central management node of these physics. The machine is a light-loaded physical machine; correspondingly, when the current utilization of CPU, memory, and network resources of some physical machines of the physical machine cluster is high, the physical machines may send a resource scheduling request message to the central management node to tell Central Management Node These physical machines are overloaded physical machines.

Optionally, the method for the central management node to obtain the information of the physical machine of the physical machine cluster may be one of the following:

(1) The physical machine of the physical machine cluster directly reports the information of the physical machine to the central management. Node.

(2) The physical machine of the physical machine cluster reports the information of the physical machine to the monitoring module of the other physical server cluster, and then sends the monitoring module of the other physical server to the central management node.

Optionally, the central management node obtains information about the physical machine of the physical machine cluster, and may also obtain information about the virtual machine running on the physical machine of the physical machine cluster, so as to more accurately determine whether the virtual machine migration operation is required. Compared with the prior art, the embodiment of the present invention does not need to obtain the information of the virtual machine running on the physical machine of the physical machine cluster, so as to reduce the burden on the central management node to obtain information, thereby improving the calculation efficiency of the central management node for information.

S302. The central management node determines the light-load physical machine and the heavy-duty physical machine in the physical machine cluster according to the information of the physical machine of the physical machine cluster.

After the central management node obtains the information of the physical machine of the physical machine cluster, the central management node can determine the light-load physical machine and the overload in the physical machine cluster according to the physical machine information (resource occupation information or load information) of the physical machine cluster. Physical machine. For example, if the current utilization rate of the CPU of a physical machine is less than the preset light load value, it is determined that the physical machine is a light-load physical machine; if the current utilization rate of the CPU of a physical machine is greater than the preset light load value And determining that the physical machine is a heavy-duty physical machine; or receiving a resource scheduling request message of the physical machine, if the resource scheduling request message indicates that the resource needs to be added, the physical machine is a heavy-load physical machine, if the resource scheduling request The message indicates that the resource is idle, and the physical machine is a light-load physical machine.

S 303. The central management node sends a second negotiation message to the light-loading physical machine, where the second negotiation message is used to query whether the light-loading physical machine is allowed to perform the power-off operation.

After the central management node determines the light-loaded physical machine in the physical machine cluster, the central management node sends a second negotiation message to the light-loading physical machines, where the second negotiation message is used to query whether the light-loading physical machine is allowed to perform the power-off operation.

After the light-loading physical machine receives the second negotiation message sent by the central management node, the light-loading physical machine may determine whether the light-loading physical machine is allowed to be powered off according to the first basis, and return the first result to the center relative to the second negotiation message. Management node, wherein the first result is It is used to indicate whether the light load physical machine is allowed to be powered off.

Optionally, the first basis may include characteristics of the light load physical machine, application characteristics of the virtual machine running on the light load physical machine, or operation information or scheduling constraints of the virtual machine running on the light load physical machine.

Specifically, the characteristics of the light-loading physical machine may include whether there is a BMC (Baseboard Management Controller) on the light-loaded physical machine, or whether the light-loaded physical machine can be remotely powered off; the virtual machine running on the light-loaded physical machine The application characteristics may include whether the virtual machine running on the light-load physical machine is allowed to be migrated; the scheduling constraint of the virtual machine running on the light-load physical machine may include that the light-loaded physical is not allowed if the light-loaded physical machine is powered on for less than or equal to 10 minutes. Power off the machine to avoid shocks.

Optionally, the first result may further include an estimated time required for the source physical machine to complete the power-off.

S304. The central management node sends a third negotiation message to the overloaded physical machine, where the third negotiation message is used to query whether the overloaded physical machine is allowed to perform the offloading operation.

After the central management node determines the overloaded physical machine in the physical machine cluster, the central management node sends a third negotiation message to the overloaded physical machines, wherein the third negotiation message is used to query whether the overloaded physical machine is allowed to perform the offloading operation.

After the reloading physical machine receives the third negotiation message sent by the central management node, the reloading physical machine may determine whether the overloaded physical machine allows the offloading according to the second basis, and returns a second result to the central management relative to the third negotiation message. A node, wherein the second result can be used to indicate whether the overloaded physical machine is allowed to be offloaded.

Optionally, the second basis may include the characteristics of the overloaded physical machine, the application characteristics of the virtual machine running on the overloaded physical machine, or the scheduling constraint of the virtual machine running on the overloaded physical machine.

Specifically, the characteristics of the overloaded physical machine may include whether the overloaded physical machine is overloaded; the application characteristics of the virtual machine running on the overloaded physical machine may include whether the virtual machine running on the overloaded physical machine is allowed to be migrated; The scheduling constraints of the running virtual machine may include whether the virtual machines running on the overloaded physical machine violate the mutual exclusion, such as some virtual machines cannot run on the same physical machine at the same time. It should be noted that the foregoing steps 303 and 308 are optional steps, and the system may perform any one of steps 30 3 and 304 or perform both, and the order of execution is not fixed, and step 30 3 may be performed first. After step 304 is performed, step 304 may be performed first and then step 303 is performed. In this embodiment, step 303 is performed first and then step 304 is performed as an example.

S 305. The central management node selects a light-load physical machine that allows the power-off operation and a heavy-duty physical machine that allows the power-down operation to be the source physical machine.

After receiving the first result and the second result respectively returned by the light load physical machine and the heavy load physical machine, the central management node may select the light load physical machine that is allowed to be powered down according to the first result and the second result respectively. As the source physical machine, and selecting the heavy-duty physical machine that allows the offload operation as the source physical machine.

After the central management node obtains the information of the physical machine of the physical machine cluster, the central management node may determine the source physical machine and the target physical machine of the virtual machine migration operation according to the information of the physical machine of the physical machine cluster. Specifically, the source physical machine is a physical machine to be migrated, and the target physical machine is a physical machine into which the virtual machine is moved.

Optionally, the central management node can determine a single source physical machine and a single target physical machine; a single source physical machine and multiple target physical machines; multiple source physical machines and a single target physical machine; or multiple source physical machines and multiple The target physical machine is not limited in the embodiment of the present invention.

S 306. The central management node selects the first target physical machine according to the information of the physical machine of the physical machine cluster.

The central management node may select the first target physical machine according to the information of the physical machine of the acquired physical machine cluster. The first target physical machine may be a physical machine with a lighter load of the physical machine cluster.

Optionally, the central management node may also be at least one physical machine of the above electrical physical machine cluster, and the at least one physical machine is used as the first target physical machine.

S 307. The central management node sends a fifth negotiation message to the first target physical machine, where the fifth negotiation message is used to query whether the first target physical machine accepts the virtual machine migration.

After the central management node determines the first target physical machine, the central management node may be the first The target physical machine sends a fifth negotiation message, and the fifth negotiation message can be used to query whether the first target physical machine accepts the virtual machine migration.

After the first target physical machine receives the fifth negotiation message sent by the central management node, the first target physical machine may determine, according to the third basis, whether it is allowed to accept the virtual machine to move in, and return the third result to the central management node, where The third result can be used to indicate whether the first target physical machine accepts virtual machine migration.

The third basis may include resource information of the first target physical machine or scheduling constraints of the virtual machine running on the first target physical machine.

Specifically, the resource information of the first target physical machine may include whether the resource of the first target physical machine is sufficient to run the virtual machine to be migrated on the source physical machine; the scheduling constraint of the virtual machine running on the first target physical machine may include the virtual machine Whether the mutual exclusion is violated, for example, some virtual machines cannot run on the same physical machine at the same time. The virtual machines may include a virtual machine to be migrated on the source physical machine and a virtual machine running on the first target physical machine.

S 308. The central management node selects a physical machine that accepts the virtual machine to move in the first target physical machine as the target physical machine.

The central management node may select a physical machine that allows the virtual machine to move in as the target physical machine according to the third result returned by the first target physical machine.

S 309. The central management node sends a first negotiation message to the source physical machine.

After the central management node determines the source physical machine, the central management node may send a first negotiation message to the source virtual machine to negotiate the migration of the virtual machine with the source physical machine.

S 31 0. The source physical machine determines information of the virtual machine to be migrated.

After the source physical machine receives the first negotiation message sent by the central management node, the source physical machine may determine the information of the virtual machine to be migrated on the source physical machine according to the first negotiation message and the first basis. The first basis may include characteristics of the physical machine, application characteristics of the virtual machine running on the physical machine, operational information of the virtual machine, or scheduling constraints of the virtual machine.

S 31 1. The source physical machine sends a response message of the first negotiation message to the central management node, where the response message of the first negotiation message carries the information of the virtual machine to be migrated.

After the source physical machine determines the information of the virtual machine to be migrated on the source physical machine, the source physics The information about the virtual machine to be migrated may be sent to the central management node in the response message of the first negotiation message, so that the central management node can learn which virtual machines on the source physical machine can be migrated according to the information of the virtual machine to be migrated.

Optionally, the information about the virtual machine to be migrated carried in the response message of the first negotiation message may include the specification of the virtual machine to be migrated, the scheduling constraint of the virtual machine to be migrated, and the real-time resource occupation information of the virtual machine to be migrated.

Further, the specifications of the virtual machine to be migrated may include the model of the virtual machine to be migrated, etc. The scheduling constraint of the virtual machine to be migrated may include whether the virtual machine to be migrated and other virtual machines violate the mutual exclusion, such as some virtual machines. You cannot run the same physical machine at the same time. The real-time resource usage information of the VM to be migrated can include the VM to be migrated.

CPU, memory, current utilization of network resources, etc.

It should be noted that the above steps 309 - 31 1 may also be performed before step 306 or concurrently with steps 306 - 308.

The central management node sends a fourth negotiation message to the target physical machine, and the fourth negotiation message is used to carry the information of the virtual machine to be migrated determined by the source physical machine.

After the central management node obtains the information about the virtual machine to be migrated on the source physical machine, the central management node may send a fourth negotiation message to the target physical machine to query whether the target physical machine is allowed to accept the virtual machine to be migrated, where The four negotiation messages carry information about the virtual machine to be migrated determined by the source physical machine.

5 31 3. The target physical machine sends a response message of the fourth negotiation message to the central management node, and the response message of the fourth negotiation message carries the information of the virtual machine to be migrated accepted by the target physical machine.

After the target physical machine receives the fourth negotiation message sent by the central management node, the target physical machine may determine whether it can accept the virtual machine to be migrated according to the second basis, and carry the information of the virtual machine to be migrated that is allowed to be accepted. The response message of the four negotiation messages is sent to the central management node, so that the central management node can determine, according to the response message of the fourth negotiation message, which virtual machines in the virtual machine to be migrated are allowed to be accepted by the target physical machine.

Optionally, the target physical machine carried in the response message of the fourth negotiation message is allowed to be connected. The information of the virtual machine to be migrated may include the specification of the virtual machine to be migrated that the target physical machine allows, the scheduling constraint of the virtual machine to be migrated that the target physical machine allows, and the real-time resource of the virtual machine to be migrated that the target physical machine allows to accept. Occupy information.

Further, the specification of the virtual machine to be migrated that the target physical machine allows to accept may include the model of the virtual machine to be migrated allowed by the target physical machine, and the like; the scheduling constraint of the virtual machine to be migrated that the target physical machine allows to accept may include the target physical machine allowed Whether the virtual machine to be migrated and the other virtual machines are mutually exclusive, such as some virtual machines cannot run on the same physical machine at the same time; the real-time resource occupation information of the virtual machine to be migrated that the target physical machine allows to accept may include The target physical machine allows the CPU, memory, and current utilization of network resources of the virtual machine to be migrated to be accepted.

S314. The central management node migrates the virtual machine to be migrated that the target physical machine allows to be transferred from the source physical machine to the target physical machine according to the response message of the fourth negotiation message.

After the central management node knows the information about the virtual machine to be migrated on the source physical machine and the information about the virtual machine to be migrated that the target physical machine allows, the central management node migrates the virtual machine to be migrated that the target physical machine allows to be migrated from the source virtual machine to the source virtual machine. Target physical machine.

Optionally, if the virtual machine on the source physical machine is completely migrated to the target physical machine, the central management node can power off the source physical machine to achieve energy saving.

Optionally, the central management node may send the first negotiation message, the second negotiation message, the third negotiation message, and the like by using a Hyper Text Transport Protocol (HTTP) or a WAP (Wireless Application Protocol). The fourth negotiation message or the fifth negotiation message.

In particular, the target physical machine can accept all the virtual machines to be migrated on the source physical machine, or only all the virtual machines to be migrated on the source physical machine. Specifically, the target physical machine accepts the number of the virtual machines to be migrated on the source physical machine according to the resource information of the target physical machine and the scheduling constraints of the virtual machine to be migrated, and the invention is not limited.

It should be noted that if multiple light-loaded physical machines are allowed to be powered off, that is, if there are multiple source physical machines, the central management node can migrate the virtual machines running on the source physical machine and the next one by selecting one source physical machine at a time. The power physical machine can also migrate the virtual machine running on the source physical machine and the physical machine under the power source in a way that all the source physical machines are selected at one time. The invention is not limited.

Correspondingly, if there are multiple heavy-duty physical machines that allow for the reduction of the load, that is, there are multiple source physical machines, the central management node can migrate the virtual machines running on the source physical machine in a manner of selecting one source physical machine at a time, or once. The virtual machine running on the source physical machine is migrated in a manner of selecting all the source physical machines, and the present invention is not limited.

According to the resource scheduling method provided by the embodiment of the present invention, the central management node negotiates resource scheduling by negotiating with the physical machine, thereby reducing the burden on the central management node to collect information of the virtual machine on the physical machine, and avoiding the central management node in the prior art. Bottlenecks, which in turn support large-scale physical machine clusters.

The embodiment of the present invention provides a resource scheduling method, after determining a source physical machine and a target physical machine of a virtual machine migration operation according to the information of the physical machine of the physical machine cluster, sending a first negotiation message to the source physical machine, and receiving the source The response message of the first negotiation message sent by the physical machine, the response message of the first negotiation message carries the information of the virtual machine to be migrated determined by the source physical machine, and the information of the virtual machine to be migrated determined according to the source physical machine, and the virtual to be migrated The machine migrates from the source physical machine to the target physical machine. Through this scheme, the central management node can determine the source physical machine and the target physical machine of the virtual machine migration, and negotiate with the source physical machine and the target physical machine respectively to determine which virtual machines on the source physical machine need to be migrated, and which targets The physical machine can accept the virtual machine to be migrated on the source virtual machine, and finally perform resource scheduling of the physical machine cluster through the result of negotiation with the source physical machine and the target physical machine, thereby alleviating the need of the central management node to collect virtual on a large number of physical machines. The burden of machine information.

Embodiment 3

As shown in FIG. 5, an embodiment of the present invention provides a central management node 1 , which may include:

The determining unit 1 0 is configured to determine the source physical machine and the target physical machine of the virtual machine migration operation according to the information of the physical machine of the physical machine cluster.

The sending unit 1 1 is configured to send a first negotiation message to the source physical machine determined by the determining unit 10.

The receiving unit 1 2 is configured to receive the first negotiation message sent by the source physical machine The response message of the first negotiation message carries the information of the virtual machine to be migrated determined by the source physical machine.

The migration unit 13 is configured to migrate the virtual machine to be migrated from the source physical machine to the target physical machine according to the information about the virtual machine to be migrated determined by the source physical machine received by the receiving unit 12 .

Optionally, the determining unit 10 is configured to determine, according to information about the physical machine of the physical machine cluster, a light-load physical machine in the physical machine cluster, and select a light-load physical machine that is allowed to be powered off. The source physical machine, and the sending unit 1 1 is further configured to send, to the light-loading physical machine in the physical machine cluster determined by the determining unit 10, a second negotiation message, where the second negotiation message is sent It is used to query whether the light-load physical machine in the physical machine cluster allows the power-off operation. .

Optionally, the determining unit 10 is specifically configured to determine, according to information about a physical machine of the physical machine cluster, a heavy-duty physical machine in the physical machine cluster, and select a heavy-duty physical machine that allows a reduced-off operation as The source physical machine, and the sending unit 1 1 is further configured to send a third negotiation message to the reloading physical machine in the physical machine cluster determined by the determining unit 10, where the third negotiation message is used. It is asked whether the heavy-duty physical machine in the physical machine cluster is allowed to perform the off-duty operation.

Optionally, before the migration unit 13 migrates the virtual machine to be migrated from the source physical machine to the target physical machine:

The sending unit 1 1 is further configured to send a fourth negotiation message to the target physical machine, where the fourth negotiation message carries information about the virtual machine to be migrated determined by the source physical machine, and the receiving unit 12 And a response message of the fourth negotiation message sent by the target physical machine, where the response message of the fourth negotiation message carries information about the virtual machine to be migrated accepted by the target physical machine; correspondingly, the migration The unit 1 3 is specifically configured to migrate the virtual machine to be migrated accepted by the target physical machine from the source physical machine to the target physics according to the response message of the fourth negotiation message received by the receiving unit 12 machine.

Optionally, the determining unit 10 is specifically configured to: select, according to information about a physical machine of the physical machine cluster, a first target physical machine, and select the first target physical And receiving, by the sending unit, the physical machine that the virtual machine is in the virtual machine, and the sending unit 11 is further configured to send a fifth negotiation message to the first target physical machine that is determined by the determining unit 10, where The fifth negotiation message is used to query whether the first target physical machine accepts virtual machine migration.

The determining unit 10 is further configured to: power on at least one physical machine before selecting the first target physical machine, and use the at least one physical machine that is powered on as the first target physical machine. The target physical machine is further configured to perform a power-off operation on the source physical machine.

The receiving unit 12 is further configured to acquire real-time resource occupation information of a physical machine of the physical machine cluster, or obtain a resource scheduling request message sent by a physical machine of the physical machine cluster, where the resource scheduling request message is lightly loaded. Or the overloaded physical machine is sent, and the determining unit 10 is configured to determine the source according to the real-time resource occupation information of the physical machine of the physical machine cluster acquired by the receiving unit 12 or the resource scheduling request message. a physical machine and the target physical machine.

Optionally, the central management node provided by the embodiment of the present invention may include a physical machine, a virtual machine, an ARM (Advanced reduced instruction set computer machines), a DSP (Digital Signal Processing) platform, Embedded systems and microcontrollers.

The embodiment of the present invention provides a central management node. After the central management node determines the source physical machine and the target physical machine of the virtual machine migration operation according to the information of the physical machine of the physical machine cluster, the central management node sends the first physical management node to the source physical machine. a negotiation message, and receiving a response message of the first negotiation message sent by the source physical machine, where the response message of the first negotiation message carries information about the virtual machine to be migrated determined by the source physical machine, and the central management node determines according to the source physical machine The information of the virtual machine to be migrated migrates the virtual machine to be migrated from the source physical machine to the target physical machine. Through this scheme, the central management node can determine the source physical machine and the target physical machine of the virtual machine migration, and negotiate with the source physical machine and the target physical machine respectively to determine which virtual machines on the source physical machine need to be migrated, and which targets The physical machine can accept the virtual machine to be migrated on the source virtual machine, and finally pass The result of negotiation with the source physical machine and the target physical machine performs resource scheduling of the physical machine cluster, thereby reducing the burden on the central management node to collect virtual machine information on a large number of physical machines.

As shown in FIG. 6, an embodiment of the present invention provides a physical machine 2, which may include:

The receiving unit 2 0 is configured to receive a first negotiation message sent by the central management node. a sending unit 2 1 , configured to send a response message of the first negotiation message to the central management node, where the response message of the first negotiation message carries information of the virtual machine to be migrated, so that the central management node is configured according to the The information of the migrated virtual machine migrates the virtual machine to be migrated.

Optionally, as shown in FIG. 7, the physical machine 2 provided by the embodiment of the present invention further includes: a determining unit 2 2, configured to determine, according to the first negotiation message and the first basis received by the receiving unit 20 The information about the virtual machine to be migrated, wherein the first basis includes characteristics of the physical machine, application characteristics of the virtual machine running on the physical machine, operation information of the virtual machine, or scheduling constraints of the virtual machine.

Optionally, the sending unit 2 1 is further configured to: before the receiving unit 20 receives the first negotiation message sent by the central management node, send a resource scheduling request message or its own resource occupation information to the central management node. .

Optionally, the physical machine 2 provided by the embodiment of the present invention may include a physical computer. An embodiment of the present invention provides a physical machine, where the physical machine receives a first negotiation message sent by a central management node, and sends a response message of the first negotiation message to the central management node, where the response message of the first negotiation message carries the virtual machine to be migrated. The information is such that the central management node migrates the migrated virtual machine according to the letter of the virtual machine to be migrated. With the solution, the physical machine can return the response message of the first negotiation message after receiving the first negotiation message sent by the central management node, and carry the information of the virtual machine to be migrated on the physical machine in the response message of the first negotiation message. Therefore, the central management node can perform resource scheduling of the physical machine cluster according to the response message of the first negotiation cancellation returned by the physical machine, thereby reducing the burden of the central management node for collecting virtual machine information on a large number of physical machines. Embodiment 4

As shown in FIG. 8, a schematic structural diagram of a central management node provided by an embodiment of the present invention may include a processor 14, a memory 15, a system bus 16, and a communication interface 17. The processor 14, the memory 15 and the communication interface 17 are connected by the system bus 16 and communicate with each other.

Processor 14 may be a single core or multi-core central processing unit, or a particular integrated circuit, or one or more integrated circuits configured to implement embodiments of the present invention.

The memory 15 may be a high speed RAM memory or a non-volatile memory such as at least one disk memory.

The memory 15 is used to store computer execution instructions. Specifically, the program code may be included in the computer execution instruction.

When the central management node is running, the processor 14 executes the computer execution instruction, and may execute the method flow described in any one of FIG. 2 or FIG. 4, specifically including:

The processor 14 is configured to determine, according to information of a physical machine of the physical machine cluster, a source physical machine and a target physical machine of the virtual machine migration operation, and send the source physical to the source through the system bus 16 and the communication interface 17 Receiving a first negotiation message, and receiving, by the system bus 16 and the communication interface 17, a response message of the first negotiation message sent by the source physical machine, where the response message of the first negotiation message carries the source physics The information about the virtual machine to be migrated is determined by the machine, and the processor 14 is configured to migrate the virtual machine to be migrated from the source physical machine to the information according to the information of the virtual machine to be migrated determined by the source physical machine a target physical machine, the memory 15, a program code for storing information of a physical machine of a physical machine cluster, a program code of a first negotiation message, a program code of a response message of a first negotiation message, and controlling the processor 14 computer instructions to complete the above steps, thereby causing the processor 14 to execute the above computer instructions and invoke the various program generations described above The above steps.

Optionally, the processor 14 is specifically configured to determine, according to information about a physical machine of the physical machine cluster, a light-load physical machine in the physical machine cluster, and pass the system bus 16 and the communication interface. 17 sending a second negotiation message to the light-loading physical machine in the physical machine cluster, where the second negotiation message is used to query the light in the physical machine cluster Whether the physical carrier is allowed to perform the power-off operation, and the light-load physical machine that allows the power-off operation to be selected as the source physical machine.

Optionally, the processor 14 is specifically configured to determine, according to information about a physical machine of the physical machine cluster, a heavy-duty physical machine in the physical machine cluster, and pass the system bus 16 and the communication interface. 17 sending a third negotiation message to the overloaded physical machine in the physical machine cluster, where the third negotiation message is used to query whether the overloaded physical machine in the physical machine cluster is allowed to perform the offload operation, and select the allowable offload operation. The overloaded physical machine acts as the source physical machine. Before the standard physical machine:

The processor 14 is further configured to send, by using the system bus 16 and the communication interface 17, a fourth negotiation message to the target physical machine, where the fourth negotiation message carries the virtual to be migrated determined by the source physical machine. Receiving, by the system bus 16 and the communication interface 17, a response message of the fourth negotiation message sent by the target physical machine, where the response message of the fourth negotiation message carries the target physical machine Correspondingly, the processor 14 is configured to: migrate, according to the response message of the fourth negotiation message, the virtual machine to be migrated accepted by the target physical machine from the source physical machine to The target physical machine.

Optionally, the processor is specifically configured to: select, according to information about a physical machine according to the physical machine cluster, a first target physical machine, and send the first to the first through the system bus 16 and the communication interface 17 The target physical machine sends a fifth negotiation message, where the fifth negotiation message is used to query whether the first target physical machine accepts virtual machine migration, and selects a physical machine that accepts the virtual machine to move in the first target physical machine. As the target physical machine.

Optionally, the processor 14 is further configured to: power on at least one physical machine before selecting the first target physical machine, and use the at least one physical machine that is powered on as the first target physical machine .

Optionally, the processor 14 is configured to perform a power-off operation on the source physical machine if the virtual machine on the source physical machine is completely migrated to the target physical machine. Optionally, the processor 14 is further configured to acquire real-time resource occupation information of the physical machine of the physical machine cluster by using the system bus 16 and the communication interface 17, or obtain the physical machine cluster. a resource scheduling request message sent by the physical machine, where the resource scheduling request message is sent by a light load or a reloading physical machine; and the processor 14 is specifically configured to use real-time resource occupation of the physical machine according to the physical machine cluster The information or the resource scheduling request message determines the source physical machine and the target physical machine.

Optionally, the central management node provided by the embodiment of the present invention may include a physical machine, a virtual machine, an ARM, a DS P platform, an embedded system, and a single chip.

The embodiment of the present invention provides a central management node. After the central management node determines the source physical machine and the target physical machine of the virtual machine migration operation according to the information of the physical machine of the physical machine cluster, the central management node sends the first physical management node to the source physical machine. a negotiation message, and receiving a response message of the first negotiation message sent by the source physical machine, where the response message of the first negotiation message carries information about the virtual machine to be migrated determined by the source physical machine, and the central management node determines according to the source physical machine The information of the virtual machine to be migrated migrates the virtual machine to be migrated from the source physical machine to the target physical machine. Through this scheme, the central management node can determine the source physical machine and the target physical machine of the virtual machine migration, and negotiate with the source physical machine and the target physical machine respectively to determine which virtual machines on the source physical machine need to be migrated, and which targets The physical machine can accept the virtual machine to be migrated on the source virtual machine, and finally perform resource scheduling of the physical machine cluster through the result of negotiation with the source physical machine and the target physical machine, thereby alleviating the need of the central management node to collect virtual on a large number of physical machines. The burden of machine information.

FIG. 9 is a schematic structural diagram of a physical machine according to an embodiment of the present invention. The physical machine in the embodiment of the present invention may include a processor 23, a memory 24, a system bus 25, and a communication interface 26. The processor 2 3, the memory 24 and the communication interface 26 are connected by the system bus 25 and communicate with each other.

Processor 23 may be a single core or multi-core central processing unit, or a particular integrated circuit, or one or more integrated circuits configured to implement embodiments of the present invention.

The memory 24 can be a high speed RAM memory or a non-volatile memory such as at least one disk memory. The memory 24 is used to store computer execution instructions. Specifically, the program code may be included in the computer execution instruction.

When the physical machine is running, the processor 2 runs the computer to execute the instruction, and the method flow described in any one of FIG. 3 or FIG. 4 may be performed, including:

The processor 2 3 is configured to receive, by using the system bus 25 and the communication interface 26, a first negotiation message sent by a central management node, and to the center through the system bus 25 and the communication interface 26 The management node sends a response message of the first negotiation message, where the response message of the first negotiation message carries the information of the virtual machine to be migrated, so that the central management node performs the virtual to be migrated according to the information of the virtual machine to be migrated. Performing migration, the memory 24 may be configured to store program code of the first negotiation message, program code of the response message of the first negotiation message, and computer instructions for controlling the processor 23 to complete the above steps, thereby causing the processing The unit 2 3 completes the above steps by executing the above computer instructions and calling the respective program codes described above.

Optionally, the processor 2 3 is configured to determine information about the virtual machine to be migrated according to the first negotiation message and the first basis stored in the memory 24, where the first basis includes The characteristics of the physical machine, the application characteristics of the virtual machine running on the physical machine, the running information of the virtual machine, or the scheduling constraints of the virtual machine.

Optionally, the processor 23 is further configured to send a resource scheduling request to the central management node by using the system bus 25 and the communication interface 26 before receiving the first negotiation message sent by the central management node. The message or its own resource occupancy information.

Optionally, the physical machine 2 provided by the embodiment of the present invention may include a physical computer. An embodiment of the present invention provides a physical machine, where the physical machine receives a first negotiation message sent by a central management node, and sends a response message of the first negotiation message to the central management node, where the response message of the first negotiation message carries the virtual machine to be migrated. The information is such that the central management node migrates the migrated virtual machine according to the letter of the virtual machine to be migrated. With the solution, the physical machine can return the response message of the first negotiation message after receiving the first negotiation message sent by the central management node, and carry the information of the virtual machine to be migrated on the physical machine in the response message of the first negotiation message. So that the central management node can perform physical machine clustering according to the first negotiation cancellation, packet, and response message returned by the physical machine. Resource scheduling, which in turn reduces the burden on the central management node to collect virtual machine information on a large number of physical machines.

Embodiment 5

As shown in FIG. 10, an embodiment of the present invention provides a resource scheduling system, including a central management node, a source physical machine, and a target physical machine, where

The central management node is configured to: after determining the source physical machine and the target physical machine of the virtual machine migration operation, according to the information of the physical machine of the physical machine cluster, send a first negotiation message to the source physical machine, and receive the source The first negotiation cancellation, the packet, the response message sent by the physical machine, the first negotiation cancellation, the packet, the response cancellation, the packet, the information of the virtual machine to be migrated determined by the source physical machine, and according to the source The source physical machine determined by the physical machine to be the virtual machine to be migrated is used to receive the first negotiation cancellation sent by the central management node, and send a response message of the first negotiation message to the central management node;

Optionally, the central management node is specifically configured to determine, according to the information of the physical machine of the physical machine cluster, the light-load physical machine in the physical machine cluster, and to the light-loaded physical medium in the physical machine cluster The machine sends a second negotiation message, where the second negotiation message is used to query whether the light-loading physical machine in the physical machine cluster is allowed to perform a power-off operation, and select a light-load physical machine that allows the power-off operation to be the source physics. machine.

Optionally, the central management node is specifically configured to determine, according to information about the physical machine of the physical machine cluster, the overloaded physical machine in the physical machine cluster, and to the overloaded physical in the physical machine cluster The machine sends a third negotiation message, the third negotiation message is used to query whether the overloaded physical machine in the physical machine cluster allows the offloading operation, and selects the overloaded physical machine that allows the offloading operation to be the source physical machine.

Optionally, the central management node is configured to send a fourth negotiation message to the target physical machine before the virtual machine to be migrated is migrated from the source physical machine to the target physical machine, where the The fourth negotiation message carries the information of the virtual machine to be migrated determined by the source physical machine, and receives a response message of the fourth negotiation cancellation sent by the target physical machine, where the response message of the fourth negotiation message carries the target physical Accepted by the machine Moving the information of the virtual machine; correspondingly, the central management node is configured to: migrate the virtual machine to be migrated accepted by the target physical machine from the source physical machine to the location according to the response message of the fourth negotiation message The target physical machine.

Optionally, the central management node is configured to: select, according to information about a physical machine of the physical machine cluster, a first target physical machine, and send a fifth negotiation message to the first target physical machine, where the The five negotiation message is used to query whether the first target physical machine accepts virtual machine migration, and selects a physical machine in the first target physical machine that accepts the virtual machine to move in as the target physical machine.

Optionally, the central management node is configured to power on at least one physical machine before selecting the first target physical machine, and use the at least one physical machine that is powered on as the first target physical machine. .

Optionally, the central management node is configured to perform a power-off operation on the source physical machine if the virtual machine on the source physical machine is completely migrated to the target physical machine.

Optionally, the central management node is further configured to acquire real-time resource occupation information of a physical machine of the physical machine cluster, or obtain a resource scheduling request message sent by a physical machine of the physical machine cluster, where the resource scheduling request is The message is sent by the light load or the overloaded physical machine. The central management node is specifically configured to determine the source physical machine according to the real-time resource occupation information of the physical machine of the physical machine cluster or the resource scheduling request message. And the source physical machine, and the source physical machine, before receiving the first negotiation message sent by the central management node, is further configured to send a resource scheduling request message to the central management node.

Optionally, the source physical machine is further configured to determine, according to the first negotiation message and the first basis, information about the virtual machine to be migrated, where the first basis includes the feature of the physical machine, and runs on the physical machine. The application characteristics of the virtual machine, the running information of the virtual machine, or the scheduling constraints of the virtual machine.

The resource scheduling system provided by the embodiment of the present invention is taken as an example, and the resource scheduling method provided by the embodiment of the present invention is further described in two scenarios (Scenario 1 and Scene 2). Specifically, the resource scheduling system provided by the embodiment of the present invention may include a central management node and a physical machine, where the virtual machine runs on the physical machine. Scene 1: Powering off the physical machine

As shown in FIG. 11 , the resource scheduling system includes a physical machine cluster composed of a physical machine 01, a physical machine 02, and a physical machine 03. The physical machine cluster may include multiple physical machines. Here, the physical machine cluster includes three physical machines as an example. ; virtual machine 01 running on physical machine 01; virtual machine 02 and virtual machine 03 running on physical machine 02; virtual machine 04 running on physical machine 03; and central management node. Among them, the physical machine 01, the physical machine 02 and the physical machine 03 are all deployed with software modules, such as distributed agents.

Exemplarily, since there are fewer virtual machines running on the physical machine 01 and the physical machine 03, in actual use, the following electrophysical machine 01 or physical machine 03 can be saved by rescheduling the virtual machines on the respective physical machines. Energy consumption. Specific steps can include:

(1) The central management node determines the light-loaded physical machine based on the physical machine information reported by the three physical machines. In this example, the central management node finds that the physical machine 01 uses the smallest memory and the lowest migration cost. Therefore, the central management node decides to power off the physical machine 01, that is, the physical machine 01 is used as a light-load physical machine; the central management node can determine The physical machine other than the physical machine 01 serves as the first target physical machine, and the physical machine 02 and the physical machine 03 can be determined as the first target physical machine.

(2) After the central management node determines that the physical machine 01 is a light-loaded physical machine, the central management node sends a second negotiation message to the physical machine 01 for inquiring whether the physical machine 01 is allowed to perform the power-off operation, for example, if the physical machine 01 When the power-off operation is allowed, the central management node selects the physical machine 01 as the source physical machine.

(3) The central management node sends the first negotiation information to the physical machine 01; and receives the response message of the first negotiation message returned by the physical machine 01, where the response message of the first negotiation message carries the virtual machine to be migrated determined by the physical machine 01 Information, such as information about virtual machine 01.

(4) The central management node sends a fifth negotiation message to the physical machine 02 and the physical machine 03 for inquiring whether the physical machine 02 and the physical machine 03 are allowed to receive the virtual machine to move in, for example, if the physical machine 02 refuses to accept the virtual machine migration. If the physical machine 03 agrees to accept the virtual machine migration, the central management node selects the physical machine 03 as the target physical machine.

(5) The central management node sends a fourth negotiation message to the physical machine 03, the fourth negotiation The message carries the information of the virtual machine to be migrated determined by the physical machine 01, that is, the information of the virtual machine 01, and receives the response message of the fourth negotiation message sent by the physical machine 03, and the response message of the fourth negotiation message carries the physical machine 0 3 accepted. Information about the virtual machine to be migrated, such as information about virtual machine 01.

(6) The central management node migrates the virtual machine 01 running on the physical machine 01 to the physical machine 03 according to the response message of the fourth negotiation message.

(7) The central management node powers down the physical machine 01 to save energy.

In the resource scheduling system provided by the embodiment of the present invention, if the light-loaded physical machine needs to be powered off to save energy, the central management node can negotiate the resource scheduling by negotiating with the physical machine, thereby reducing the central management node to collect the physical machine. The burden of the information of the virtual machine avoids the bottleneck problem of the central management node in the prior art, and further supports the large-scale physical machine cluster.

Scenario 2: Powering on the physical machine

As shown in FIG. 12, the resource scheduling system includes a physical machine cluster composed of a physical machine 01, a physical machine 02, and a physical machine 03. The physical machine cluster may include multiple physical machines. Here, the physical machine cluster includes three physical machines as an example. ;); virtual machine 02 and virtual machine 03 running on physical machine 02; virtual machine 01 and virtual machine 04 running on physical machine 03; and central management node. The physical machine 01 is a physical machine that has been powered off, and the physical machine 01, the physical machine 02, and the physical machine 0 3 are all deployed with software modules, such as distributed agents.

Exemplarily, since there are many virtual machines running on the physical machine 03, and the resources of the physical machine 03 cannot meet the resource requirements of the virtual machines, in actual use, the physical machine 01 can be powered on, and the physical machine can be powered on. 03 Some VMs running on the machine are migrated to the physical machine 01 to balance the virtual machines and ensure the application of Qo S. Specific steps can include:

(1) The central management node determines the overloaded physical machine based on the physical machine information reported by the three physical machines. In this example, the central management node finds that the average resource utilization of physical machine 02 and physical machine 03 is higher than the preset threshold. Therefore, it is decided to perform migration balancing on the virtual machines running on physical machine 02 and physical machine 0 3, that is, the physical Machine 02 and physical machine 03 act as heavy-duty physical machines; the central management node can power on the powered-off physical machine 01, and the physical machine 01 as the first target physical machine.

(2) After the central management node determines that the physical machine 02 and the physical machine 03 are overloaded physical machines, the central management node sends a third negotiation message to the physical machine 02 and the physical machine 03, respectively, for inquiring the physical machine 02 and the physical machine 03. Whether the off-duty operation is allowed, for example, if both the physical machine 02 and the physical machine 03 allow the off-duty operation, the central management node selects the physical machine 02 and the physical machine 03 as the source physical machine.

(3) The central management node sends the first negotiation information to the physical machine 02 and the physical machine 03 respectively; and receives the response message of the first negotiation message respectively returned by the physical machine 02 and the physical machine 03, wherein the response message of the first negotiation message The information of the virtual machine to be migrated determined by the physical machine 02 and the physical machine 03, such as the information of the virtual machine 02 running on the physical machine 02 and the information of the virtual machine 01 running on the physical machine 03.

(4) The central management node sends a fifth negotiation message to the physical machine 01 for inquiring whether the physical machine 01 is allowed to receive the virtual machine to move in. For example, if the physical machine 01 agrees to accept the virtual machine to move in, the central management node selects the physical Machine 01 serves as the target physical machine.

Optionally, if the physical machine cluster includes other physical machines, the central management node may also select other lightly loaded physical machines as the target physical machine.

(5) The central management node sends a fourth negotiation message to the physical machine 01, where the fourth negotiation message carries the information of the virtual machine to be migrated determined by the physical machine 02 and the physical machine 03, that is, the information of the virtual machine 02 and the information of the virtual machine 01. And receiving the response message of the fourth negotiation message sent by the physical machine 01, and the response message of the fourth negotiation message carries the information of the virtual machine to be migrated accepted by the physical machine 01, such as the information of the virtual machine 02 and the information of the virtual machine 01.

(6) The central management node migrates the virtual machine 02 running on the physical machine 02 and the virtual machine 01 running on the physical machine 0 to the physical machine 01 according to the response message of the fourth negotiation message.

In the resource scheduling system provided by the embodiment of the present invention, if the heavy-duty physical machine needs to be reduced, that is, the virtual machine running on the overloaded physical machine is balanced to ensure the application of Qo S, and the central management node can negotiate with the physical machine. The resource scheduling can reduce the burden of the central management node collecting the information of the virtual machine on the physical machine, avoiding the bottleneck problem of the central management node in the prior art, and supporting the large-scale physical machine cluster. It will be clearly understood by those skilled in the art that for the convenience and brevity of the description, only the division of each functional module described above is exemplified. In practical applications, the above functional assignments may be completed by different functional modules as needed. The internal structure of the device is divided into different functional modules to perform all or part of the functions described above. For the specific working process of the system, the device and the unit described above, refer to the corresponding process in the foregoing method embodiment, and details are not described herein again.

In the several embodiments provided by the present application, it should be understood that the disclosed system, apparatus, and method may be implemented in other manners. For example, the device embodiments described above are merely illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be used. Combined or can be integrated into another system, or some features can be ignored, or not executed. In addition, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be electrical, mechanical or otherwise.

The units described as separate components may or may not be physically separated, and the components displayed as the units may or may not be physical units, and may be located in one place or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the embodiment of the present embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a standalone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may contribute to the prior art or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium. , including a number of instructions to make a computer device (either a personal computer, a server, or a network device) Or a processor or processor performs all or part of the steps of the method of the various embodiments of the present invention. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and the like, which can store program codes. .

The above is only the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily think of changes or substitutions within the technical scope of the present invention. It should be covered by the scope of the present invention. Therefore, the scope of the invention should be determined by the scope of the claims.

Claims

A resource scheduling method, comprising:

Determining the source physical machine and the target physical machine of the virtual machine migration operation according to the information of the physical machine of the physical machine cluster;

Sending a first negotiation message to the source physical machine;

Receiving, by the source physical machine, a response message of the first negotiation message, the first negotiation cancellation, the packet, the response cancellation, the packet, and the information of the virtual machine to be migrated determined by the source physical machine; The information about the virtual machine to be migrated determined by the physical machine, and the virtual machine to be migrated is migrated from the source physical machine to the target physical machine.

The resource scheduling method according to claim 1, wherein the source physical machine that determines the virtual machine migration operation according to the information of the physical machine of the physical machine cluster includes:

Determining, according to information of the physical machine of the physical machine cluster, a light-load physical machine in the physical machine cluster;

The resource scheduling method according to claim 1 or 2, wherein the source physical machine that determines the virtual machine migration operation according to the information of the physical machine of the physical machine cluster includes:

Determining a heavy-duty physical machine in the physical machine cluster according to information of a physical machine of the physical machine cluster;

The resource scheduling method according to any one of claims 1 to 3, wherein the method further comprises:

Sending a fourth negotiation message to the target physical machine, where the fourth negotiation message carries Decoding the information of the virtual machine to be migrated determined by the source physical machine;

The resource scheduling method according to claim 1 or 4, wherein the determining the target physical machine of the virtual machine migration operation according to the information of the physical machine of the physical machine cluster comprises:

The resource scheduling method according to claim 5, wherein before the selecting the first target physical machine, the method further includes:

The resource scheduling method according to claim 1, wherein the method further comprises:

If the virtual machine on the source physical machine is completely migrated to the target physical machine, the source physical machine is powered off.

The resource scheduling method according to claim 1, wherein before the determining the source physical machine and the target physical machine of the virtual machine migration operation, the method further comprises: acquiring the physical machine of the physical machine cluster Realizing resource occupation information, or acquiring a resource scheduling request message sent by a physical machine of the physical machine cluster, where the resource scheduling request message is sent by a light load or a heavy load physical machine;

The source physical machine and the target physical machine that determine the virtual machine migration operation according to the information of the physical machine of the physical machine cluster include: Determining the source physical machine and the target physical machine according to real-time resource occupation information of the physical machine of the physical machine cluster or the resource scheduling request message.

9. A resource scheduling method, characterized in that:

Receiving a first negotiation message sent by the central management node;

Sending, by the central management node, a response message of the first negotiation message, the first negotiation cancellation, the packet, the response cancellation, the packet, and the information of the virtual machine to be migrated, so that the central management node is to be migrated according to the to-be-migrated The information of the virtual machine migrates the virtual machine to be migrated.

The resource scheduling method according to claim 9, wherein after the first negotiation message sent by the receiving center management node, the method further includes:

The resource scheduling method according to claim 9 or 10, wherein before the receiving the first negotiation message sent by the central management node, the method further comprises: sending a resource scheduling to the central management node. Request message; or,

Sending its own resource occupation information to the central management node.

1 2. A central management node, which is characterized by comprising:

a determining unit, configured to determine, according to information of a physical machine of the physical machine cluster, a source physical machine and a target physical machine of the virtual machine migration operation;

a sending unit, configured to send a first negotiation message to the source physical machine determined by the determining unit;

a receiving unit, configured to receive a first negotiation cancellation packet, a response message sent by the source physical machine, where the response message of the first negotiation message carries information about the virtual machine to be migrated determined by the source physical machine;

And a migration unit, configured to determine a target physical machine according to the source physical machine received by the receiving unit.

The central management node according to claim 12, wherein the determining unit is specifically configured to use information about a physical machine of the physical machine cluster, Determining a light-load physical machine in the physical machine cluster, and selecting a light-load physical machine that allows a power-off operation as the source physical machine;

The central management node according to claim 12 or 13, wherein the determining unit is specifically configured to determine, according to information of the physical machine of the physical machine cluster, the physical machine cluster Reloading the physical machine, and selecting a heavy-duty physical machine that allows the offload operation to be the source physical machine;

The sending unit is further configured to send a third negotiation message to the reloading physical machine in the physical machine cluster determined by the determining unit, where the third negotiation message is used to query an overload in the physical machine cluster Whether the physical machine is allowed to perform the burden reduction operation.

The central management node according to any one of claims 1 2 to 14 wherein the target physical machine is:

The sending unit is further configured to send a fourth negotiation message to the target physical machine, where the fourth negotiation message carries information about the virtual machine to be migrated determined by the source physical machine;

The receiving unit is further configured to receive a response message of the fourth negotiation message sent by the target physical machine, where the response message of the fourth negotiation message carries information about the virtual machine to be migrated accepted by the target physical machine;

The central management node according to claim 12 or 15, wherein the determining unit is specifically configured to select the first target physical machine according to the information of the physical machine according to the physical machine cluster. And selecting, as the target physical machine, a physical machine that accepts a virtual machine to move in the first target physical machine;

The sending unit is further configured to send a fifth negotiation message to the first target physical machine determined by the determining unit, where the fifth negotiation message is used to query the first target physical Whether the machine accepts virtual machine migration.

The central management node according to claim 16, wherein the determining unit is further configured to: power on at least one physical machine before selecting the first target physical machine, and power on the device At least one physical machine is described as the first target physical machine.

The central management node according to claim 12, wherein the migration unit is further configured to: if the virtual machine on the source physical machine is completely migrated to the target physical machine The physical machine is powered off.

The central management node according to claim 12, wherein the receiving unit is further configured to acquire real-time resource occupation information of a physical machine of the physical machine cluster, or acquire the physical machine cluster. a resource scheduling request message sent by the physical machine, where the resource scheduling request message is sent by a light load or a heavy load physical machine;

2 0. A physical machine, comprising:

The physical machine according to claim 20, wherein the physical machine further comprises:

a determining unit, configured to determine, according to the first negotiation message and the first basis received by the receiving unit, information about the virtual machine to be migrated, where the first basis includes a characteristic of a physical machine, and is running on a physical machine The application characteristics of the virtual machine, the running information of the virtual machine, or the scheduling constraints of the virtual machine.

The physical machine according to claim 20 or 21, wherein the sending unit is further configured to receive, by the receiving unit, a central management node, Before the first negotiation message, the resource scheduling request message or its own resource occupation information is sent to the central management node.

2 3. A central management node, comprising: a processor, a memory, a bus, and a communication interface;

A readable medium, comprising: a computer executing instruction, when the processor of the central management node executes the computer execution instruction, the central management node performs the method of any one of claims 1-8 Resource scheduling method.

2 5. A physical machine, comprising: a processor, a memory, a bus, and a communication interface;

The memory is configured to store a computer execution instruction, the processor is connected to the memory through the bus, and when the physical machine is running, the processor executes the computer execution instruction stored in the memory, so that The physical machine performs the resource scheduling method according to any one of claims 9-1.

A readable medium, comprising: a computer executing instructions, when the processor of the physical machine executes the computer to execute an instruction, the physical machine executes the resource according to any one of claims 9-1 Scheduling method.

2 7. A resource scheduling system, comprising:

a central management node, configured to send a first negotiation message to the source physical machine after receiving the source physical machine and the target physical machine of the virtual machine migration operation according to the information of the physical machine of the physical machine cluster, and receive the source physical machine And the response message of the first negotiation message sent by the source physical machine, and the information about the virtual machine to be migrated determined according to the source physical machine And the source physical machine to be used to receive the first negotiation message sent by the central management node, and send a response message of the first negotiation message to the central management node; The target physical machine is configured to receive the migration of the virtual machine to be migrated.

The resource scheduling system according to claim 27, wherein the central management node is specifically configured to determine a light-load physical machine in the physical machine cluster according to information of a physical machine of the physical machine cluster And sending a second negotiation message to the light-loading physical machine in the physical machine cluster, where the second negotiation message is used to query whether the light-load physical machine in the physical machine cluster is allowed to perform a power-off operation, and select permission A light-loaded physical machine that is powered down operates as the source physical machine.

The resource scheduling system according to claim 27 or 28, wherein the central management node is specifically configured to determine an overload in the physical machine cluster according to information of a physical machine of the physical machine cluster. a physical machine, and sending a third negotiation message to the overloaded physical machine in the physical machine cluster, where the third negotiation message is used to query whether the overloaded physical machine in the physical machine cluster is allowed to perform the subtraction operation, and select A heavy physical machine that allows the offload operation is used as the source physical machine.

The resource scheduling system according to any one of claims 27 to 29, wherein the target physical machine is further configured to send a fourth negotiation message to the target physical machine, where the fourth negotiation message is And carrying the information about the virtual machine to be migrated determined by the source physical machine, and receiving a response message of the fourth negotiation message sent by the target physical machine, where the response message of the fourth negotiation message carries the The information of the virtual machine is migrated; correspondingly, the central management node is configured to: migrate the virtual machine to be migrated accepted by the target physical machine from the source physical machine to the location according to the response message of the fourth negotiation message The target physical machine.

The resource scheduling system according to claim 27 or 30, wherein the central management node is specifically configured to select a first target physical machine according to information of a physical machine of the physical machine cluster, and The first target physical machine sends a fifth negotiation message, where the fifth negotiation message is used to query whether the first target physical machine accepts virtual machine migration, and selects the first target physical machine to accept virtual machine migration The physical machine acts as the target physical machine. The resource scheduling system according to claim 31, wherein the central management node is further configured to power on at least one physical machine before selecting the first target physical machine, and to power on the At least one physical machine acts as the first target physical machine.

The resource scheduling system according to claim 27, wherein the central management node is further configured to: if the virtual machine on the source physical machine is completely migrated to the target physical machine The physical machine is powered off.

The resource scheduling system according to claim 27, wherein the central management node is further configured to acquire real-time resource occupation information of a physical machine of the physical machine cluster, or acquire physicality of the physical machine cluster. a resource scheduling request message sent by the machine, where the resource scheduling request message is sent by a light load or a heavy load physical machine;

The central management node is specifically configured to determine the source physical machine and the target physical machine according to real-time resource occupation information of the physical machine of the physical machine cluster or the resource scheduling request message;

The resource scheduling system according to claim 27, wherein the source physical machine is further configured to determine information about the virtual machine to be migrated according to the first negotiation message and the first basis, where The first basis includes the characteristics of the physical machine, the application characteristics of the virtual machine running on the physical machine, the running information of the virtual machine, or the scheduling constraint of the virtual machine.