Service instance management method based on cloud technology, cloud management platform and cluster

Abstract

The invention provides a service instance management method based on a cloud technology, a cloud management platform and a cluster. The method comprises the steps that a cloud management platform receives event information sent by a first physical machine in a plurality of physical machines, the event information comprises expected execution time of operation, and the operation influences the state of a first resource node of the first physical machine; and before the expected execution time, the cloud management platform migrates the service instance in the first resource node to a second resource node of a second physical machine in the plurality of physical machines based on the QoS requirement of the service instance in the first resource node. According to the method, the QoS of the service instance can be prevented from being influenced by the operation of the resource node where the service instance is located.

Description

[0001] This application claims the priority of a Chinese patent application with an application number of 202311797508.4 and an application title of "A Fault Management System" filed with the National Intellectual Property Administration on December 25, 2023. The entire content thereof is incorporated herein by reference. Technical Field

[0002] This application relates to the field of cloud computing, and particularly to a method for managing service instances of cloud technology, a cloud management platform, and a cluster. Background Art

[0003] In the field of cloud computing, service instances such as containers are created on resource nodes such as physical machines or virtual machines (VMs) through virtualization technology. Among them, the management of resource nodes and the management of service instances are separated. For example, a virtual machine as a resource node is managed by the infrastructure as a service (IaaS) layer, and a container as a service instance is managed by the kubernetes (k8s) control plane.

[0004] Since the management of resource nodes and the management of service instances are separated, the management of service instances cannot perceive operations on resource nodes, which causes service instances in resource nodes to passively adapt to operations on resource nodes. For example, when a migration operation is performed on a resource node, the service instances in that resource node passively follow the migration of the resource node. This may affect the quality of service (QoS) of the service instances, resulting in the QoS of the service instances being lower than the QoS requirements. Summary of the Invention

[0005] This application provides a method for managing service instances based on cloud technology, a cloud management platform, and a cluster, which can avoid the QoS of service instances being affected by operations on the resource nodes where the service instances are located.

[0006] In a first aspect, a service instance management method based on cloud technology is provided. This method is applied to a cloud management platform connected to a cloud infrastructure. The cloud management platform is used to manage multiple service instances deployed in the cloud infrastructure. The cloud infrastructure includes multiple physical machines, and each physical machine in the multiple physical machines is used to deploy one or more resource nodes. Each service instance in the multiple service instances is deployed in at least one resource node in the cloud infrastructure. The quality of service (QoS) of a service instance is affected by the state of the resource node where the service instance is located. The method includes: The cloud management platform receives event information sent by a first physical machine among the multiple physical machines. The event information includes the expected execution time of an operation, where the operation affects the state of a first resource node of the first physical machine. Before the expected execution time, the cloud management platform migrates the service instance in the first resource node to a second resource node of a second physical machine among the multiple physical machines based on the QoS requirements of the service instance in the first resource node. Here, the second physical machine is a healthy physical machine, and a healthy physical machine refers to a physical machine that has not failed and has not been changed.

[0007] When a physical machine is about to perform an operation that affects the resource node of the physical machine, it can send event information. The event information includes the expected execution time of the operation. The cloud management platform can receive the event information and obtain the expected execution time of the operation from the event information. The cloud management platform can migrate the service instance to the resource node of another physical machine based on the QoS requirements of the service instance in the resource node. In this way, the execution of this operation will not affect the QoS of the service instance, that is, it avoids the QoS of the service instance being affected by the operation of the resource node where the service instance is located.

[0008] In a possible implementation, the event information further includes the type of the operation. The cloud management platform migrates the service instance in the first resource node to a second resource node of a second physical machine among the multiple physical machines based on the QoS requirements of the service instance in the first resource node, including: The cloud management platform confirms that the operation will cause the QoS of the service instance in the first resource node to be lower than the QoS requirements based on the type of the operation. The cloud management platform migrates the service instance in the first resource node to the second resource node.

[0009] When it is confirmed that the operation performed on the resource node will cause the QoS of the service instance to be lower than the QoS requirements, the service instance is migrated to the resource node of another physical machine. In this way, when the operation is executed, the QoS of the service instance can be guaranteed not to be lower than the QoS requirements, improving the user's service experience.

[0010] In a possible implementation, the cloud management platform migrates the service instances in the first resource node to the second resource node of the second physical machine among multiple physical machines based on the QoS requirements of the service instances in the first resource node, including: the cloud management platform confirms that the QoS requirement is greater than or equal to a preset QoS threshold; the cloud management platform migrates the service instances in the first resource node to the second resource node.

[0011] This QoS threshold is a threshold for judging whether a service instance can tolerate operations that affect the resource node. If the QoS requirement of the service instance is greater than or equal to this QoS threshold, it means that the QoS requirement of service instance A1 cannot tolerate any operations that affect the resource node. In this case, operations that affect the resource node are defaulted to cause the QoS of the service instance to be not lower than the QoS requirement. Therefore, there is no need to further judge whether the operations that affect the resource node will cause the QoS of the service instance to be not lower than the QoS requirement, and directly migrate the service instance to the resource node of other physical machines. In this way, while ensuring that the QoS of the service instance is not lower than the QoS requirement and improving the user's service experience, it can also save the computing resources of the cloud management platform.

[0012] In a possible implementation, the method further includes: the cloud management platform prohibits deploying new service instances in the first resource node.

[0013] Through the event information, the cloud management platform can confirm that an operation that affects the first resource node will be performed on the first resource node. In this case, the cloud management platform no longer deploys new service instances in the first resource node to avoid the impact of the operation that affects the first resource node on the QoS of the newly deployed service instances.

[0014] In a possible implementation, the operation is used to handle hardware failures or changes, where the hardware is used for the first physical machine to run the first resource node; or the operation is used to handle software failures or changes; where the software includes at least one of the operating system of the first physical machine and the management software of the first resource node.

[0015] That is to say, failures or changes in the software and hardware related to the resource node in the physical machine trigger the physical machine to perform an operation to handle the failure or change on the resource node and send out event information. The cloud management platform can, through this event information, before the operation is executed, migrate the services in the resource node to the resource nodes of other physical machines to avoid the impact of the operation performed on the resource node on the QoS of the service instances.

[0016] In a possible implementation, the service instances in the first resource node include any one or more of containers, database instances, and data warehouse instances.

[0017] In a possible implementation, the first resource node is a virtual machine or a bare-metal server, or the first resource node is the first physical machine itself.

[0018] In a second aspect, a cloud management platform is provided. The cloud management platform is connected to a cloud infrastructure and is used to manage multiple service instances deployed in the cloud infrastructure. The cloud infrastructure includes multiple physical machines, and each physical machine in the multiple physical machines is used to deploy one or more resource nodes. Each service instance in the multiple service instances is deployed in at least one resource node in the cloud infrastructure, and the quality of service (QoS) of the service instance is affected by the state of the resource node where the service instance is located. The cloud management platform includes: a receiving module, configured to receive event information sent by a first physical machine among the multiple physical machines, where the event information includes the expected execution time of an operation, and the operation affects the state of a first resource node of the first physical machine; and a migration module, configured to, before the expected execution time, based on the QoS requirements of the service instances in the first resource node, migrate the service instances in the first resource node to a second resource node of a second physical machine among the multiple physical machines.

[0019] In a possible implementation, the event information further includes the type of the operation; the migration module is configured to: based on the type of the operation, confirm that the operation will cause the QoS of the service instances in the first resource node to be lower than the QoS requirements; and migrate the service instances in the first resource node to the second resource node.

[0020] In a possible implementation, the migration module is configured to: confirm that the QoS requirements are greater than or equal to a preset QoS threshold; and migrate the service instances in the first resource node to the second resource node.

[0021] In a possible implementation, the cloud management platform further includes: a prohibition module, configured to prohibit deploying new service instances in the first resource node.

[0022] In a possible implementation, the operation is used to handle hardware failures or changes, where the hardware is used for the first physical machine to run the first resource node; or the operation is used to handle software failures or changes; where the software includes at least one of the operating system of the first physical machine and the management software of the first resource node.

[0023] In a possible implementation, the service instances in the first resource node include any one or more of containers, database instances, and data warehouse instances.

[0024] In a possible implementation, the first resource node is a virtual machine or a bare-metal server, or the first resource node is the first physical machine itself.

[0025] In a third aspect, a computing device cluster is provided, including at least one computing device, and each computing device includes a processor and a memory; the processor of at least one computing device is configured to execute instructions stored in the memory of at least one computing device, so that the computing device cluster executes the method provided in the first aspect.

[0026] In a fourth aspect, a computer-readable storage medium is provided, including computer program instructions, and when the computer program instructions are executed by the computing device cluster, the computing device cluster executes the method provided in the first aspect.

[0027] In a fifth aspect, a computer program product including instructions is provided, and when the instructions are run by a computer device cluster, the computer device cluster is caused to execute the method provided in the first aspect.

[0028] For the beneficial effects of the second aspect to the fifth aspect, reference may be made to the introduction of the beneficial effects of the first aspect above, and details are not described herein again. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 A schematic diagram of a system architecture provided by an embodiment of the present application;

[0030] Figure 2 A schematic diagram of the structure of an anomaly perception instance provided by an embodiment of the present application;

[0031] Figure 3 A schematic diagram of the structure of an event generation instance provided by an embodiment of the present application;

[0032] Figure 4 A functional schematic diagram of an event perception channel provided by an embodiment of the present application;

[0033] Figure 5 A schematic diagram of the structure of an event perception instance provided by an embodiment of the present application;

[0034] Figure 6 A flowchart of a service instance management method provided by an embodiment of the present application;

[0035] Figure 7 A schematic diagram of a service instance management method provided by an embodiment of the present application;

[0036] Figure 8 A schematic diagram of a service instance management method provided by an embodiment of the present application;

[0037] Figure 9 A schematic diagram of the structure of a cloud management platform provided by an embodiment of the present application;

[0038] Figure 10 A schematic diagram of the structure of a computing device provided by an embodiment of the present application;

[0039] Figure 11 A schematic structural diagram of a computing device cluster provided by an embodiment of the present application;

[0040] Figure 12 A schematic structural diagram of a computing device cluster connected through a network provided by an embodiment of the present application. Detailed implementation manners

[0041] The solutions provided by the embodiments of the present application will be described below in conjunction with the accompanying drawings. Among them, in the embodiments of the present application, "a plurality of" means two or more, and "a variety of" means two or more. "First", "second", etc. are only used to distinguish similar objects and do not necessarily need to describe a specific order or the number of objects.

[0042] To facilitate understanding of the solutions provided by the embodiments of the present application, the technical terms that may be involved in the embodiments of the present application will be introduced first.

[0043] Cloud technology refers to a hosting service that unifies a series of resources such as hardware, software, and networks within a wide area network or a local area network to achieve data computing, storage, processing, and sharing.

[0044] Cloud infrastructure: Also known as infrastructure, it is the facility that supports cloud computing services, including at least one data center, and each data center includes multiple physical machines (such as servers). Among them, a physical machine can be used as a resource node or a resource node is deployed on it, and service instances such as containers, database instances, and data warehouse service instances can be run in the resource node to achieve elastic cloud computing services. For example, in the case where the cloud infrastructure includes multiple data centers, the multiple data centers can be distributed in different geographical regions, and the data centers are remotely connected through a backbone network.

[0045] Resource node: A node with resources such as computing and storage, which can be used to deploy service instances. A resource node can be a physical machine, such as a server. A resource node can also be a virtual machine or a bare metal server (BMS) deployed on a physical machine.

[0046] Service instance: A collection of programs or processes used to provide services such as data storage and data computing. Common service instances include: database instances, data warehouse service instances, containers, etc. One service instance can be deployed in at least one resource node.

[0047] Cloud management platform: It is a platform provided by cloud service providers for managing service instances, such as creating, shutting down, and migrating service instances. Among them, cloud service providers can provide cloud services to users in the form of service instances. The cloud management platform can interact with users. Users can register accounts on the cloud management platform and rent cloud services provided by one or more service instances with the accounts, thus becoming tenants of cloud services.

[0048] Quality of Service (QoS): It refers to the quality of the service provided by a service instance. QoS can be represented by metrics such as latency and throughput. Among them, QoS is negatively correlated with latency and positively correlated with throughput. QoS reflects the service-providing ability of a service instance.

[0049] QoS requirement: It refers to the requirement of a cloud provider or tenant for the QoS of a service instance. For example, if the QoS of a service instance includes latency, the requirement for the QoS of this service instance stipulates that the latency cannot be greater than a preset latency threshold. Another example is that if the QoS of a service instance includes throughput, the QoS requirement for this service instance stipulates that the throughput of this service instance cannot be less than a preset throughput threshold. Among them, the QoS requirement of a service instance can be set by a cloud provider or the tenant to which the service instance belongs.

[0050] Scheduled event: Also known as an in-scheduled operation and maintenance event or operation and maintenance operation, it is an event actively issued to improve the reliability and security of resource nodes. Usually, by actively predicting the software and hardware failure risks of a physical machine and generating a scheduled event based on the prediction result. A scheduled event is usually to migrate the resource node in the physical machine to prevent the impact of software and hardware failures of the physical machine on the resource node.

[0051] Unscheduled event: Also known as an out-of-schedule operation and maintenance event or operation and maintenance operation, it is an event triggered by software and hardware failures of a physical machine. Compared with scheduled events, unscheduled events are passively generated to cope with software and hardware failures that have occurred in the physical machine. Unscheduled events are usually also to migrate the resource node in the physical machine to avoid the impact of software and hardware failures of the physical machine on the resource node.

[0052] Live migration: Also known as dynamic migration or real-time migration, it refers to completely saving the running state of the entire resource node in a physical machine and quickly restoring the running state of this resource node to other physical machines. The restored resource node still runs smoothly. Among them, when performing live migration, it is necessary to reduce the frequencies of the central processing unit (CPU), memory, etc. of the resource node to increase the time interval for the CPU and memory to execute instructions or process data. Then, complete the live migration within this time interval.

[0053] Cold migration: Also known as static migration, it refers to shutting down the resource node and migrating the resource node from the current physical machine to another physical machine while the resource node is in a shutdown state. Cold migration requires shutting down the resource node and takes a relatively long time. Compared with hot migration, cold migration has a greater impact on the services of the resource node.

[0054] In cloud technology, cloud services are provided through service instances, and service instances are deployed in resource nodes. Therefore, the QoS of service instances is affected by the state of resource nodes. Among them, the state of resource nodes can specifically be the running state of resource nodes. When the state of a resource node deteriorates, the QoS of the service instance decreases. Operations on resource nodes often affect the state of resource nodes. For example, to cope with physical machine software and hardware failures, physical machine operating system upgrades, sub-healthy states of physical machine hardware resources, etc., hot migration operations or cold migration operations of resource nodes are often triggered. Among them, the hot migration operation of resource nodes causes the CPU of the resource node to downclock, the memory to downclock, etc., making the resource node in a state of low-frequency CPU operation and low-frequency memory operation. This affects the QoS of service instances. If the QoS requirements of service instances are high, the downclocking of the CPU and memory of resource nodes causes the QoS of service instances to be lower than the QoS requirements. In addition, cold migration requires shutting down the resource node, which has a greater impact on the QoS of service instances.

[0055] The cloud management platform for managing service instances and the underlying management platform for managing resource nodes are separated, resulting in the cloud management platform being unable to predict operations on resource nodes (such as planned events or unplanned events). Therefore, often when the operation on the resource node has started to execute and the state of the resource node has affected the QoS of the service instance, the cloud management platform re-schedules the service instance, for example, migrates the service instance to a resource node in another physical machine. This affects the reliability and resilience of cloud services and even affects users' experience of cloud services.

[0056] The embodiment of the present application provides a method for managing service instances based on cloud technology. In this method, before operating on a resource node in a physical machine, the physical machine can send out event information. Among them, the event information includes the expected execution time of the operation, and the operation affects the resource node in the physical machine. The cloud management platform can receive the event information and obtain the expected execution time of the operation from the event information. Then, based on the QoS requirements of the service instance in the resource node, the cloud management platform can migrate the service instance to a resource node in another physical machine before the expected execution time of the operation.

[0057] In this method, the cloud management platform can obtain the expected execution time of an operation that affects a resource node in a physical machine through event information. Before this expected execution time, the cloud management platform migrates the service instance in this resource node to a resource node in another physical machine. In this way, the quality of service of the service instance in the resource node can be prevented from being affected by the operation of the resource node.

[0058] Next, a specific introduction to the service instance management method provided by the embodiments of the present application will be given.

[0059] Figure 1 Shows a system architecture that can be used to implement this method. As Figure 1 shown, this system architecture includes a cloud infrastructure 100 and a cloud management platform 200. The cloud infrastructure 100 includes multiple physical machines, such as physical machine 110, physical machine 120, etc. Exemplarily, the physical machine can specifically be a server.

[0060] In some embodiments, the cloud infrastructure 100 can include multiple data centers. Among them, each data center can include at least one of the multiple physical machines. That is to say, different physical machines among the multiple physical machines can be respectively deployed in different data centers. For example, physical machine 110 and physical machine 120 are respectively deployed in different data centers.

[0061] Each physical machine can deploy one or more resource nodes. For example, physical machine 110 deploys resource node 111, and physical machine 120 deploys node 121. In some embodiments, the resource node deployed in the physical machine is the physical machine itself. In some embodiments, the resource node deployed in the physical machine can be a virtual machine or a bare metal server, etc. The cloud management platform 200 can create one or more service instances in the resource nodes in the cloud infrastructure 100. For example, service instance A1 is created in resource node 111, and service instance B1 is created in resource node 112, etc. Among them, one service instance can be deployed in at least one resource node, and the QoS of the service instance is affected by the state of the resource node where the service instance is located. The cloud management platform 200 can manage the service instances in the resource nodes. For example, migrate the service instance from the current resource node to other resource nodes.

[0062] Refer to Figure 1 , each physical machine includes an exception perception instance and an event generation instance. Among them, the exception perception instance is used to perceive the exception or change of the physical machine and issue an alarm. The event generation instance is used to generate an event based on the alarm and send event information to the cloud management platform 200. The exception perception instance and the event generation instance run directly in the physical machine. For example, they run directly in the operating system of the physical machine in the form of a system service or an application program.

[0063] Exceptions of the physical machine may include hardware failures of the physical machine. Refer to Figure 2 , the exception perception instance may include a hardware failure detection module. The hardware failure detection module is used to detect the status of the hardware of the physical machine. Among them, the hardware may include one or more of a processor, a south bridge chip (platform controller hub, PCH), a data card, a redundant arrays of independent disks (RAID) card, a fan, a memory, a disk, a network card, a motherboard, a power supply, etc. Among them, the processor may be a CPU, a graphics processing unit (GPU), etc. The data card may be a peripheral component interconnect express (PCIE) card, etc.

[0064] In some embodiments, the hardware failure detection module may detect hardware operation metrics, such as the temperature of the CPU, the frequency of the CPU, the accumulated amount of dirty pages in the memory, etc. In some embodiments, the hardware failure detection module may detect kernel events of the physical machine operating system, such as CPU internal error (CPU IERR) events, system-level memory initialization error events, and hard disk drive fault events generated in the kernel log, etc. Among them, the hardware failure detection module may periodically detect hardware operation metrics or kernel events, such as detecting hardware operation metrics or kernel events every 5 seconds.

[0065] As Figure 2 shown, the exception perception instance includes a fault identification module. The fault identification module may obtain the hardware operation metrics and / or kernel events detected by the hardware failure detection module, and based on the hardware operation metrics and / or kernel events, identify whether the hardware has failed.

[0066] As Figure 2 shown, the exception perception instance has a fault identification rule library. The fault identification rule library includes fault identification rules. Among them, the fault identification rules may be preset. Through the fault identification rules, based on the hardware operation metrics and / or kernel events detected by the hardware failure detection module, it can be identified that the hardware has failed. In one example, the fault identification rules are shown in Table 1.

[0067] Table 1

[0068] Hardware operation metrics and / or kernel events Fault type CPU temperature is greater than the preset temperature threshold Fan failure CPU frequency is greater than the preset frequency threshold CPU failure Internal CPU error CPU failure System-level memory initialization error event Memory configuration error Hard disk drive fault event Remaining hard disk life is too low Accumulated amount of dirty memory pages is greater than the preset threshold Memory is in a sub-healthy state … …

[0069] As Figure 2As shown, the anomaly perception instance further includes an alarm generation module. The alarm generation module is used to generate an alarm based on the hardware fault identified by the fault identification module. Among them, when a hardware fault occurs in the hardware, and the faulty hardware is the hardware required by the physical machine running resource node, the fault identification module can trigger the alarm generation module to generate an alarm. In one example, the alarm generation module can generate an alarm based on the alarm format shown in Table 2.

[0070] Table 2

[0071] Alarm component Alarm identifier Alarm name (fault type) CPU 0x00000001 CPU failure Fan 0x00000002 Fan failure Memory 0x00000003 Memory configuration error Hard disk 0x00000004 Remaining hard disk life is too low … … …

[0072] The alarm generation module can generate an alarm based on the hardware fault, referring to the alarm format shown in Table 2. Among them, the alarm corresponds to the fault type, and different fault types generate different alarms.

[0073] In some embodiments, as Figure 2 shown, the anomaly perception instance includes a software fault detection module. The software fault detection module is used to detect the running state of the software. For example, the response time of the software, the frequency of software downtime or freezing, thread leakage, etc. Among them, the software is the operating system of the physical machine or the management software of the resource node. For example, when the resource node is a virtual machine, the software can be a virtual machine monitor (VMM). The fault identification module can identify whether the software has a fault based on the running state of the software detected by the software fault detection module. For example, when the response time of the software is greater than the preset time threshold, it can be confirmed that the software has a fault. In addition, the fault can also include a sub-healthy state. Among them, if the software has a thread leakage, it means that the software is in a sub-healthy state. When the fault identification module confirms that the software has a fault, it can trigger the alarm generation module to generate an alarm.

[0074] In some embodiments, as Figure 2 shown, the anomaly perception instance includes a change perception module. Among them, the change can be an operation and maintenance operation within the plan. For example, replacing one or more hardwares of the physical machine, upgrading software such as the operating system of the physical machine, installing patches, etc. The operation and maintenance personnel can input the change time and change type through the change perception module. The alarm generation module can generate an alarm based on the change time and change type.

[0075] The event generation instance can generate an event based on the alarm. As Figure 3As shown in the figure, the event generation instance may include an alarm receiving module, an event generation module, and an event information sending module. Among them, the alarm receiving module may receive alarms from the anomaly perception instance and send the received alarms to the event generation module. The event generation module may generate events based on the alarms. An event refers to performing a certain operation on a resource node at the expected execution time. That is to say, an event includes an operation and the expected execution time of the operation. The operation can have multiple types. Common operation types include resource node migration, physical machine restart, etc. Among them, the migration of resource nodes can be divided into hot migration and cold migration. Among them, compared with cold migration, the hot migration of resource nodes takes less time and has less impact on the QoS of service instances in the resource nodes.

[0076] The event generation module may generate different events based on different alarms. Among them, the difference in events may refer to the different types of operations in the events. For example, if the alarm is an alarm corresponding to a CPU failure, the type of operation in the generated event is resource node cold migration. Another example is that if the alarm is an alarm corresponding to a memory configuration error, the type of operation in the generated event is physical machine restart, and so on. In some embodiments, the event generation instance has an event generation rule library. The event generation rule library includes the pre-set corresponding relationship between alarms and events. The event generation module may generate corresponding events based on this corresponding relationship and the alarms.

[0077] The operation in the event is an operation to address the failure or change corresponding to the alarm. That is to say, the operation in the event is an operation that the physical machine will perform to address hardware failures, software failures, or software and hardware changes. For example, if a physical machine hardware fails and this failure affects the operation of the resource node, in order to ensure the operation of the resource node, a migration operation is performed on the resource node. And so on.

[0078] An event can be represented by event information. The event information includes the expected execution time of the operation, the type of operation, etc. Among them, in some embodiments, the event information may also include an event identifier (ID), an event description, and the execution status of the operation, etc. Among them, the execution status of the operation may include to be executed, executing, executed, cancelled, etc.

[0079] In one example, the format of the event information may be as shown in Table 3.

[0080] Table 3

[0081] Field Meaning ID Event ID Type Type of operation in the event Description Event description NotBefore Time when the operation starts to execute NotAfter Time when the operation ends to execute State Execution status of the operation

[0082] Whenever the event generation module generates an event, it may send the event information of the event to the event information sending module. The event information sending module may send out the event information.

[0083] The cloud management platform 200 can receive the event information and schedule service instances in the resource nodes of the physical machines based on the event information. Among them, referring to Figure 1 , event perception instances are deployed in the resource nodes of the physical machines. The event perception instance is a type of service instance, specifically a service instance for perceiving event information. The event perception instance in the physical machine can obtain the event information sent by the event generation instance in the physical machine. The event perception instance can send the obtained event information to the cloud management platform 200.

[0084] In some embodiments, in the field of cloud computing, for security reasons, the virtual network environment where the service instance is located and the physical network environment where the physical machine is located are isolated, and the service instance and the physical machine cannot directly access each other. That is to say, the event perception instance cannot obtain event information by directly accessing the physical machine. Referring to Figure 4 , this embodiment provides an event perception channel. Through the event perception channel, the event perception instance can obtain event information.

[0085] The event perception channel can be an interface through which the event perception instance can obtain event information. Specifically, this interface is associated with a storage space of the physical machine. Among them, this storage space is used to store the event information generated by the physical machine. That is to say, the event generation instance can store the event information in this storage space. The physical machine can open or provide this interface to the event perception instance in the resource node of the physical machine. The event perception instance can access the storage space associated with the interface through this interface, so as to read the event information from this storage space. Among them, the event perception instance can periodically access this storage space through this interface to detect whether new event information is generated. For example, the event perception instance can access this storage space through this interface every 1 second. Only the storage space associated with this interface can be accessed through this interface, and other storage spaces cannot be accessed. In this way, the information security of the physical machine is guaranteed.

[0086] In some embodiments, as Figure 5 shown, the event perception instance is preset with a perception rule library. The perception rule library includes the corresponding relationship between the preset event perception instance and the event perception channel. The event perception instance can obtain the event perception channel corresponding to the event perception instance based on this corresponding relationship, so as to obtain event information through this event perception channel.

[0087] In some embodiments, the event perception instance may screen the obtained event information. Specifically, for some event information, the corresponding operation may have been cancelled or completed. In such cases, there is no need to send such event information to the cloud management platform 200 anymore. The event perception instance may screen out the event information corresponding to the operation to be executed from the obtained event information, and send the screened event information to the cloud management platform. As described above, the event information includes the execution status of the operation, and the event perception instance screens the event information based on the execution status of the operation in the event information.

[0088] In some embodiments, as Figure 5 shown, the event perception instance is pre - configured with a QoS requirement library. The QoS requirement library includes the QoS requirements of the service instances in the resource node where the event perception instance is located. Based on the QoS requirement library, the event perception instance may also send the QoS requirements of the service instances to the cloud management platform 200. For example, when the event perception instance sends event information to the cloud management platform 200, it sends the QoS requirements of the service instances to the cloud management platform 200, enabling the cloud management platform to schedule the service instances based on the event information and the QoS requirements of the service instances.

[0089] The above examples introduce the system architecture provided by the embodiments of the present application. Next, in combination with this system architecture, the service instance management method provided by the embodiments of the present application will be introduced.

[0090] This method is executed by the cloud management platform 200. As Figure 6 shown, this method includes the following steps.

[0091] Step 601, the cloud management platform 200 receives event information sent by the physical machine 110 among multiple physical machines. The event information includes the expected execution time of operation C1, where operation C1 affects the state of the resource node 111 of the physical machine 110.

[0092] The cloud management platform 200 may receive event information from the physical machine 110. For example, it receives event information from the event perception instance in the resource node 111 of the physical machine 110. For specific details, please refer to the above introduction and will not be elaborated here.

[0093] This expected execution time is later than the current time and later than the execution time of step 601.

[0094] Operation C1 is an operation that physical machine 110 will execute at the expected execution time, and operation C1 affects the state of resource node 111. Among them, the state of the resource node can be the running state of the resource node. That is to say, when physical machine 111 executes operation C1, the state of resource node 111 will change. Specifically, the state of resource node 111 will deteriorate. As described above, the deterioration of the state of the resource node causes the QoS of the service instance in the resource node to decline. That is to say, the execution of operation C1 will cause the QoS of the service instance in resource node 111 to decline.

[0095] In some embodiments, operation C1 is an operation for dealing with the failure or change of the hardware of physical machine 110, where the hardware is the hardware required for physical machine 110 to run resource node 111. That is, this hardware provides hardware resources for physical machine 1110 to run resource node 111. In some embodiments, operation C1 is an operation for dealing with software failures or changes; among them, the software is the operating system of physical machine 111 or the management software of the resource node. The generation method and function of operation C1 can also refer to the introduction of the embodiments shown in Figure 2 and Figure 3 above, which will not be elaborated here.

[0096] Step 602, before the expected execution time, based on the QoS requirements of the service instances in resource node 111, cloud management platform 200 migrates the service instances in resource node 111 to the resource node 121 of physical machine 120 among multiple physical machines. Among them, physical machine 120 is a healthy physical machine, and cloud management platform 200 has not received event information sent by physical machine 120.

[0097] The service instance has QoS requirements. The QoS requirements of the service instance can be set by the tenant to which the service instance belongs or by the cloud provider.

[0098] In some embodiments, cloud management platform 200 saves the QoS requirements of each service instance. Cloud management platform 200 can execute step 602 based on the QoS requirements of the service instances in resource node 111 it saves.

[0099] In some embodiments, cloud management platform 200 can receive the QoS requirements of the service instances in resource node 111 from resource node 111. For example, as described above, the event awareness instance in resource node 111 can send the QoS requirements of the service instances in resource node 111 to cloud management platform 200.

[0100] In some embodiments, when the resource node is a virtual machine or a bare metal server, the operation types on the resource node may include live migration of the resource node, cold migration of the resource node, and reboot of the physical machine. That is to say, operation C1 belongs to one of live migration of the resource node, cold migration of the resource node, and reboot of the physical machine. Among them, live migration of the resource node, cold migration of the resource node, and reboot of the physical machine will all cause the state of the resource node to deteriorate, that is, live migration of the resource node, cold migration of the resource node, and reboot of the physical machine will all reduce the QoS of the service instances in the resource node. Among them, the degrees of decrease in QoS of the service instances caused by live migration of the resource node, cold migration of the resource node, and reboot of the physical machine increase in sequence.

[0101] The QoS requirements of service instances can have multiple QoS levels, and different QoS levels tolerate different operation types. It can be set that the multiple QoS levels include QoS level D1, QoS level D2, QoS level D3, and QoS level D4 with gradually decreasing QoS. Among them, the QoS requirements of service instances providing game backend computing services are usually QoS level D1, the QoS requirements of service instances providing e-commerce services are usually QoS level D2, the QoS requirements of service instances providing offline AI model training services are usually QoS level D3, and the QoS requirements of service instances providing periodic dial test services are usually QoS level D4.

[0102] As shown in Table 4, QoS level D1 cannot tolerate operations of any operation type, that is, QoS level D1 cannot tolerate live migration of the resource node. The QoS level that cannot tolerate live migration of the resource node also cannot tolerate cold migration of the resource node and reboot of the physical machine. That is to say, live migration of the resource node causes the QoS of the service instances in the resource node to be lower than QoS level D1.

[0103] QoS level D2 can tolerate live migration of the resource node, but cannot tolerate cold migration of the resource node and reboot of the physical machine. That is to say, the QoS of the service instances affected by live migration of the resource node is not lower than QoS level D2, but cold migration of the resource node or reboot of the physical machine causes the QoS of the service instances to be lower than QoS level D2.

[0104] QoS level D3 can tolerate cold migration of the resource node, but cannot tolerate reboot of the physical machine. Among them, the QoS level that tolerates cold migration of the resource node can also tolerate live migration of the resource node. That is to say, the QoS of the service instances affected by cold migration or live migration of the resource node is not lower than QoS level D3, but reboot of the physical machine causes the QoS of the service instances to be lower than QoS level D3.

[0105] The QoS level D4 can tolerate operations of any operation type. That is to say, no matter what operation type is performed on the resource node, the QoS of the service instance in the resource node is higher than the QoS level D4.

[0106] Table 4

[0107] Service instance QoS requirement Tolerable operation type Service instance A1 QoS level D1 Cannot tolerate any operation type Service instance A2 QoS level D2 Hot migration of resource nodes Service instance A3 QoS level D3 Cold migration of resource nodes QoS level D4 Can tolerate any operation type

[0108] In some embodiments, in step 602, the cloud management platform 200 determines whether the QoS requirement of the service instance A1 in the resource node 111 is greater than or equal to a preset QoS threshold. Among them, the preset QoS threshold may be the QoS level D1. If the QoS requirement of the service instance A1 is greater than or equal to the preset QoS threshold, in step 602, the cloud management platform 200 migrates the service instance A1 to the resource node 121. That is to say, in step 602, when the cloud management platform 200 confirms that the QoS requirement of the service instance A1 in the resource node 111 is greater than or equal to the preset QoS threshold, it migrates the service instance A1 to the resource node 121.

[0109] Specifically, the QoS requirement of the service instance A1 being greater than or equal to the preset QoS threshold indicates that the QoS requirement of the service instance A1 cannot tolerate any type of operation. In this case, no matter what type of operation the operation C1 belongs to, the QoS requirement of the service instance A1 cannot tolerate it, that is, the operation C1 defaults to causing the QoS of the service instance A1 to be lower than the QoS requirement of the service instance A1. Therefore, there is no need to further identify what operation type the operation C1 is, and the migration of the service instance is directly performed.

[0110] The migration of the service instance can be performed in the following manner.

[0111] Refer to Figure 7 , first, a service instance A1' is created in the resource node 121. The service instance A1' is a copy of the service instance A1 and can provide the services provided by the service instance A1. When the service instance A1' is successfully created, that is, when the service instance A1' can provide the services provided by the service instance A1, the service instance A1' is started, and the service instance A1 is isolated. New service requests are sent to the service instance A1', and no new service requests are sent to the service instance A1. Among them, the service instance A1 still remains running to process the existing service requests (that is, the service requests sent to the service instance A1 before the service instance A1' is started). Exemplarily, when the service instance A1 finishes processing the existing service requests, the cloud management platform 200 can delete the service instance A1 from the resource node 111.

[0112] Through the above service instance migration method, lossless migration of the service instance is achieved, and the QoS of the services provided by the service instance A1 can be made not lower than the QoS level D1.

[0113] When the expected execution time of operation C1 arrives, the physical machine 110 performs operation C1 on the resource node 111. At this time, the service instance A1 has been migrated out of the resource node 111, and the operation C1 performed on the resource node 111 does not affect the service instance A1. Or rather, the service instance providing the service provided by the service instance A1 has been migrated out of the resource node 111, and the operation 111 performed on the resource node 111 does not affect the QoS of this service.

[0114] In some embodiments, the event information received by the cloud management platform 200 further includes the type of operation C1. In step 602, based on the type of operation C1, it can be determined whether operation C1 will cause the QoS of the service instance in the resource node 111 to be lower than the QoS requirement of this service instance. Among them, it can be determined whether the QoS requirement of this service instance can tolerate the type of operation C1. If it cannot tolerate, it is confirmed that operation C1 will cause the QoS of the service instance in the resource node 111 to be lower than the QoS requirement of this service instance. If it can tolerate, it is confirmed that operation C1 will not cause the QoS of the service instance in the resource node 111 to be lower than the QoS requirement of this service instance.

[0115] For example, as shown in Table 4, it can be set that there is also a service instance A2 deployed in the resource node 111, where the QoS requirement of the service instance A2 is the QoS level D2. As described above, the QoS level D2 can only tolerate the hot migration of the resource node. If operation C1 is the cold migration of the resource node or the restart of the physical machine, it is confirmed that operation C1 will cause the QoS of the service instance A2 to be lower than the QoS requirement of the service instance A2. If operation C1 is the hot migration of the resource node, it is confirmed that operation C1 will not cause the QoS of the service instance A2 to be lower than the QoS requirement of the service instance A2.

[0116] It can also be set that there is also a service instance A3 deployed in the resource node 111, where the QoS requirement of the service instance A3 is the QoS level D3. As described above, the QoS level D3 can tolerate the cold migration and hot migration of the resource node. If operation C1 is the restart of the physical machine, it is confirmed that operation C1 will cause the QoS of the service instance A3 to be lower than the QoS requirement of the service instance A3. If operation C1 is the hot migration or cold migration of the resource node, it is confirmed that operation C1 will not cause the QoS of the service instance A3 to be lower than the QoS requirement of the service instance A3.

[0117] In step 602, when it is confirmed that operation C1 will cause the QoS of the service instance in the resource node 111 to be lower than the QoS requirement of this service instance, the cloud management platform 200 migrates the service instance in the resource node 111 to the resource node 121.

[0118] For example, asFigure 8 As shown, operation C1 will cause the QoS of service instance A2 in resource node 111 to be lower than the QoS requirement of service instance A2, and cloud management platform 200 will migrate service instance A2 from resource node 111 to resource node 121. Among them, service instance A2' is created in resource node 121. Service instance A2' is a copy of service instance A2 and can replace service instance A2 in resource node 111 to provide corresponding services. For specific reference, please refer to the above introduction and will not be elaborated here.

[0119] In step 602, when it is confirmed that C1 will not cause the QoS of the service instance in resource node 111 to be lower than the QoS requirement of this service instance, cloud management platform 200 may not migrate this service instance. When performing operation C1 on resource node 111, this service instance is scheduled in the traditional manner.

[0120] For example, as Figure 8 shown, operation C1 will not cause the QoS of service instance A3 in resource node 111 to be lower than the QoS requirement of service instance A3. In this case, cloud management platform may not migrate service instance A3 out of resource node 111.

[0121] In some embodiments, when cloud management platform 200 receives event information, cloud management platform 200 prohibits deploying new service instances in the resource node corresponding to the event information (i.e., resource node 111). In this way, it is avoided that new service instances are affected by operation C1 and the scheduling pressure on cloud management platform 200 is reduced.

[0122] In summary, when a physical machine is about to perform an operation that affects the resource node of this physical machine, it may send event information to the cloud management platform. This event information includes the expected execution time of this operation. When the cloud management platform receives this event information, it may obtain this expected execution time and, before this expected execution event, migrate the service instances in this resource node to resource nodes in other physical machines. In this way, it can be avoided that the QoS of the service instances in the resource node is affected by this operation.

[0123] Referring to Figure 9 , an embodiment of the present application provides a cloud management platform 900. The cloud management platform 900 is connected to the cloud infrastructure. The cloud management platform is used to manage multiple service instances deployed in the cloud infrastructure. The cloud infrastructure includes multiple physical machines. Each physical machine in the multiple physical machines is used to deploy one or more resource nodes. Each service instance in the multiple service instances is deployed in at least one resource node in the cloud infrastructure. The QoS of the service instance is affected by the state of the resource node where the service instance is located. As Figure 9 shown, the cloud management platform 900 includes:

[0124] A receiving module 910 is configured to receive event information sent by a first physical machine among the multiple physical machines. The event information includes an expected execution time of an operation, where the operation affects the state of a first resource node of the first physical machine.

[0125] A migration module 920 is configured to, before the expected execution time, based on QoS requirements of service instances in the first resource node, migrate the service instances in the first resource node to a second resource node of a second physical machine among the multiple physical machines.

[0126] In some embodiments, the event information further includes a type of the operation. The migration module 920 is configured to: based on the type of the operation, confirm that the operation will cause the QoS of the service instances in the first resource node to be lower than the QoS requirements; and migrate the service instances in the first resource node to the second resource node.

[0127] In some embodiments, the migration module 920 is configured to: confirm that the QoS requirements are greater than or equal to a preset QoS threshold; and migrate the service instances in the first resource node to the second resource node.

[0128] In some embodiments, the cloud management platform 900 further includes: a prohibition module configured to prohibit deploying new service instances in the first resource node.

[0129] In some embodiments, the operation is used to cope with a hardware failure or change, where the hardware is used by the first physical machine to run the first resource node; or the operation is used to cope with a software failure or change; where the software includes at least one of an operating system of the first physical machine and management software of the first resource node.

[0130] In some embodiments, the service instances in the first resource node include any one or more of containers, database instances, and data warehouse instances.

[0131] In some embodiments, the first resource node is a virtual machine or a bare metal server, or the first resource node is the first physical machine itself.

[0132] Wherein, both the receiving module 910 and the migration module 920 can be implemented by software or can be implemented by hardware. Exemplarily, next, taking the receiving module 910 as an example, the implementation manner of the receiving module 910 is introduced. Similarly, the implementation manner of the migration module 920 can refer to the implementation manner of the receiving module 910.

[0133] As an example of a software functional unit, the receiving module 910 may include code running on a computing instance. The computing instance may include at least one of a physical host (computing device), a virtual machine, and a container. Further, the above computing instance may be one or more. For example, the receiving module 910 may include code running on multiple hosts / virtual machines / containers. It should be noted that the multiple hosts / virtual machines / containers for running the code may be distributed in the same region or in different regions. Further, the multiple hosts / virtual machines / containers for running the code may be distributed in the same Availability Zone (AZ) or in different AZs, and each AZ includes one data center or multiple geographically proximate data centers. Usually, one region may include multiple AZs.

[0134] Similarly, the multiple hosts / virtual machines / containers for running the code may be distributed in the same Virtual Private Cloud (VPC) or in multiple VPCs. Usually, one VPC is set within one region. For cross-region communication between two VPCs within the same region and between VPCs in different regions, a communication gateway needs to be set in each VPC, and the interconnection between VPCs is achieved through the communication gateway.

[0135] As an example of a hardware functional unit, the receiving module 910 may include at least one computing device, such as a server, etc. Alternatively, the receiving module 910 may also be a device implemented using an application-specific integrated circuit (ASIC) or a programmable logic device (PLD). Among them, the above PLD may be implemented by a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0136] The multiple computing devices included in the receiving module 910 may be distributed in the same region or in different regions. The multiple computing devices included in the receiving module 910 may be distributed in the same AZ or in different AZs. Similarly, the multiple computing devices included in the receiving module 910 may be distributed in the same VPC or in multiple VPCs. Among them, the multiple computing devices may be any combination of computing devices such as servers, ASICs, PLDs, CPLDs, FPGAs, and GALs.

[0137] It should be noted that in other embodiments, the receiving module 910 can be used to execute Figure 6 any step in the method shown, and the migration module 920 can be used to execute Figure 6 any step in the method shown. The steps to be implemented by the receiving module 910 and the migration module 920 can be specified as needed, and all functions of the cloud management platform 900 are implemented by respectively implementing Figure 6 different steps in the method shown.

[0138] This application also provides a computing device 1000. As Figure 10 shown, the computing device 1000 includes: a bus 1002, a processor 1004, a memory 1006, and a communication interface 1008. The processor 1004, the memory 1006, and the communication interface 1008 communicate with each other through the bus 1002. The computing device 1000 can be a server or a terminal device. It should be understood that this application does not limit the number of processors and memories in the computing device 1000.

[0139] The bus 1002 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 10 only one line is shown herein, but it does not mean that there is only one bus or one type of bus. The bus 1002 can include a path for transmitting information between various components (such as the memory 1006, the processor 1004, and the communication interface 1008) of the computing device 1000.

[0140] The processor 1004 can include any one or more of a central processing unit (CPU), a graphics processing unit (GPU), a micro processor (MP), or a digital signal processor (DSP), etc.

[0141] The memory 1006 may include a volatile memory, such as a random access memory (RAM). The memory 1006 may also include a non-volatile memory, such as a read-only memory (ROM), a flash memory, a hard disk drive (HDD), or a solid state drive (SSD).

[0142] The executable program code is stored in the memory 1006, and the processor 1004 executes the executable program code to implement the functions of the aforementioned receiving module 910 and migration module 920 respectively, so as to implement Figure 6 the method shown. That is to say, the instructions for executing Figure 6 the method shown are stored on the memory 1006.

[0143] The communication interface 1008 uses a transceiver module such as, but not limited to, a network interface card or a transceiver to implement the communication between the computing device 1000 and other devices or communication networks.

[0144] The embodiment of the present application also provides a computing device cluster. The computing device cluster includes at least one computing device. The computing device may be a server, such as a central server, an edge server, or a local server in a local data center. In some embodiments, the computing device may also be a terminal device such as a desktop computer, a laptop computer, or a smart phone.

[0145] As Figure 11 shown, the computing device cluster includes at least one computing device 1000. The same instructions for executing Figure 6 the method shown may be stored in the memory 1006 of one or more of the computing devices 1000 in the computing device cluster.

[0146] In some possible implementation manners, the memory 1006 of one or more of the computing devices 1000 in the computing device cluster may also store partial instructions for executing Figure 6 the method shown respectively. In other words, the combination of one or more computing devices 1000 may jointly execute the instructions for executing Figure 6 the method shown.

[0147] It should be noted that the memories 1006 in different computing devices 1000 in the computing device cluster may store different instructions, respectively for executing some functions of the cloud management platform 900. That is to say, the instructions stored in the memories 1006 in different computing devices 1000 can implement the functions of one or more modules in the receiving module 910 and the migration module 920.

[0148] In some possible implementation manners, one or more computing devices in the computing device cluster may be connected through a network. Among them, the network may be a wide area network or a local area network, etc. Figure 12 A possible implementation manner is shown. As Figure 12 shown, two computing devices 1000A and 1000B are connected through a network. Specifically, they are connected to the network through the communication interfaces in each computing device. In this type of possible implementation manner, the memory 1006 in the computing device 1000A stores instructions for executing the function of the receiving module 910. At the same time, the memory 1006 in the computing device 1000B stores instructions for executing the function of the migration module 920.

[0149] It should be understood that Figure 12 the functions of the computing device 1000A shown in

[0150] can also be completed by multiple computing devices 1000. Similarly, the functions of the computing device 1000B can also be completed by multiple computing devices 1000.

[0150] The embodiment of the present application also provides another computing device cluster. The connection relationship between the computing devices in this computing device cluster can be similarly referred to Figure 11 and Figure 12 the connection manner of the computing device cluster. The difference is that the memories 1006 in one or more computing devices 1000 in this computing device cluster may store the same instructions for executing Figure 6 the method shown.

[0151] In some possible implementation manners, the memories 1006 in one or more computing devices 1000 in this computing device cluster may also respectively store partial instructions for executing Figure 6 the method shown. In other words, a combination of one or more computing devices 1000 can jointly execute the instructions for executing Figure 6 the method shown.

[0152] The embodiment of the present application also provides a computer program product containing instructions. The computer program product may be software or a program product containing instructions that can run on a computing device or be stored in any available medium. When the computer program product runs on at least one computing device, it causes at least one computing device to execute Figure 6 the method shown.

[0153] The embodiments of the present application further provide a computer-readable storage medium. The computer-readable storage medium may be any available medium that can be stored by a computing device or a host migration device such as a data center including one or more available media. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive), etc. The computer-readable storage medium includes instructions that direct the computing device to execute Figure 6 the method shown

[0154] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features. However, such modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present application.

Claims

1. A service instance management method based on cloud technology, characterized in that: The method is applied to a cloud management platform connected to a cloud infrastructure, the cloud management platform is used to manage multiple service instances deployed in the cloud infrastructure, the cloud infrastructure includes multiple physical machines, each of the multiple physical machines is used to deploy one or more resource nodes, each of the multiple service instances is deployed in at least one resource node in the cloud infrastructure, and the service quality QoS of the service instance is affected by the state of the resource node where the service instance is located; the method comprises: The cloud management platform receives event information sent by a first physical machine among the multiple physical machines, the event information including an expected execution time of an operation, wherein the operation affects a state of a first resource node of the first physical machine; Before the expected execution time, the cloud management platform migrates the service instance in the first resource node to a second resource node of a second physical machine among the multiple physical machines based on the QoS requirement of the service instance in the first resource node.

2. The method according to claim 1, characterized in that The event information also includes the type of the operation; the cloud management platform migrates the service instance in the first resource node to a second resource node of a second physical machine among the multiple physical machines based on the QoS requirement of the service instance in the first resource node, including: The cloud management platform confirms, based on the type of the operation, that the operation will cause the QoS of the service instance in the first resource node to be lower than the QoS requirement; The cloud management platform migrates the service instance in the first resource node to the second resource node.

3. The method according to claim 1, characterized in that The cloud management platform migrates the service instance in the first resource node to a second resource node of a second physical machine among the multiple physical machines based on the QoS requirement of the service instance in the first resource node, including: The cloud management platform confirms that the QoS requirement is greater than or equal to a preset QoS threshold; The cloud management platform migrates the service instance in the first resource node to the second resource node.

4. The method according to any one of claims 1 to 3, characterized in that The method also includes: the cloud management platform prohibiting deployment of new service instances in the first resource node.

5. The method according to any one of claims 1 to 4, characterized in that The operation is used to deal with a failure or change of hardware, wherein the hardware is used for the first physical machine to run the first resource node; or, The operation is used to deal with a failure or change of software; wherein the software includes at least one of an operating system of the first physical machine and management software of the first resource node.

6. The method according to any one of claims 1 to 5, characterized in that The service instance in the first resource node includes any one or more of a container, a database instance, and a data warehouse instance.

7. The method according to any one of claims 1 to 6, characterized in that The first resource node is a virtual machine or a bare metal server, or the first resource node is the first physical machine itself.

8. A cloud management platform, characterized in that: The cloud management platform is connected to a cloud infrastructure, and the cloud management platform is used to manage multiple service instances deployed in the cloud infrastructure, wherein the cloud infrastructure includes multiple physical machines, each of the multiple physical machines is used to deploy one or more resource nodes, each of the multiple service instances is deployed in at least one resource node in the cloud infrastructure, and the service quality QoS of the service instance is affected by the state of the resource node where the service instance is located; the cloud management platform includes: a receiving module, configured to receive event information sent by a first physical machine among the multiple physical machines, wherein the event information includes an expected execution time of an operation, wherein the operation affects a state of a first resource node of the first physical machine; A migration module is used to migrate the service instance in the first resource node to a second resource node of a second physical machine among the multiple physical machines based on the QoS requirement of the service instance in the first resource node before the expected execution time.

9. The cloud management platform according to claim 8, characterized in that: The event information also includes the type of the operation; the migration module is used to: Based on the type of the operation, confirming that the operation will cause the QoS of the service instance in the first resource node to be lower than the QoS requirement; Migrate the service instance in the first resource node to the second resource node.

10. The cloud management platform according to claim 8, characterized in that: The migration module is used to: Confirming that the QoS requirement is greater than or equal to a preset QoS threshold; Migrate the service instance in the first resource node to the second resource node.

11. The cloud management platform according to any one of claims 8 to 10, characterized in that: The cloud management platform also includes: A prohibition module is used to prohibit the deployment of new service instances in the first resource node.

12. The cloud management platform according to any one of claims 8 to 11, characterized in that: The operation is used to deal with a failure or change of hardware, wherein the hardware is used for the first physical machine to run the first resource node; or, The operation is used to deal with a failure or change of software; wherein the software includes at least one of an operating system of the first physical machine and management software of the first resource node.

13. The cloud management platform according to any one of claims 8 to 12, characterized in that: The service instance in the first resource node includes any one or more of a container, a database instance, and a data warehouse instance.

14. The cloud management platform according to any one of claims 8 to 13, characterized in that: The first resource node is a virtual machine or a bare metal server, or the first resource node is the first physical machine itself.

15. A computing device cluster, characterized in that: comprising at least one computing device, each computing device comprising a processor and a memory; The processor of the at least one computing device is configured to execute instructions stored in the memory of the at least one computing device, so that the computing device cluster executes the method according to any one of claims 1 to 7.

16. A computer-readable storage medium, characterized in that: The method comprises computer program instructions. When the computer program instructions are executed by a computing device cluster, the computing device cluster performs the method according to any one of claims 1 to 7.

17. A computer program product comprising instructions, characterized in that When the instructions are executed by a computing device cluster, the computing device cluster is caused to perform the method according to any one of claims 1 to 7.

Images