WO2021004256A1

WO2021004256A1 - Node switching method in node failure and related device

Info

Publication number: WO2021004256A1
Application number: PCT/CN2020/097262
Authority: WO
Inventors: 郑营飞
Original assignee: 华为技术有限公司
Priority date: 2019-07-08
Filing date: 2020-06-19
Publication date: 2021-01-14
Also published as: CN112199240B; CN112199240A

Abstract

A node switching method in a node failure and a related device. A master node and a standby node are simultaneously connected to a storage device, but only the master node may access user data in the storage device and provides a service for a user. The standby node may access a status tag of the master node in the storage device. The standby node monitors the status tag of the master node stored in the storage device during operation and determines, according to the status tag, whether the master node fails. If the standby node determines, according to the status tag, that the master node fails, the standby node takes the place of the master node. By means of the method, when multiple nodes sharing one storage device do not sense each other, it is ensured that the standby node can accurately sense the status of the master node and takes the place of the master node when the master node fails, thereby improving the reliability of an application.

Description

Method and related equipment for node switching when node fails

Technical field

The present invention relates to the technical field of cloud computing storage systems, in particular to a method and related equipment for node switching when a node fails.

Background technique

In cloud computing scenarios, applications that provide services to users are generally deployed in containers of virtual machines or physical machines. In order to ensure the reliability of service applications, each service application corresponds to a main container and at least one backup container, and the main container and the backup container are provided with a common storage device. Under normal working conditions, only the main container can read and write data in the storage device to provide services to the outside world, and the standby container cannot read and write data in the storage device. It can only monitor the status of the main container and take over when the main container fails. The work of upgrading to the main container, read and write to the storage device to provide services.

At present, physical machines and storage devices communicate through the small computer system interface (SCSI) protocol. For the main container and backup container deployed in different physical machines, the main container can use the SCSI lock command provided by the SCSI protocol. Lock the storage device. Since a network connection is established between different physical machines, the backup container can monitor the status of the main container through the network connection. After the main container fails, the backup container can monitor the failure of the main container in time, and Immediately upgrade to the main and continue to provide external services. However, in practical applications, the standby container may be created temporarily, and may be created on the same physical machine as the main container, so the standby container cannot establish a network connection with the main container, and thus cannot monitor the status of the main container.

Therefore, how to ensure that the standby container accurately perceives the state of the primary container and switch when the primary container fails is an urgent problem to be solved.

Summary of the invention

The embodiment of the present invention discloses a node method and related equipment when a node fails, which can ensure that the standby node can accurately perceive the state of the primary node in a scenario where multiple nodes sharing a storage device do not perceive each other, and the primary node Take over the master node in the event of a failure, thereby improving application reliability.

In a first aspect, the present application provides an inter-node switching method, including: a standby node detects a status flag of a master node stored in a storage device, and determines whether the master node is faulty according to the status flag, wherein the The master node is a node that accesses data in the storage device and provides services for users; when the backup node determines that the master node is faulty according to the status flag, the backup node takes over the master node.

In the embodiment of this application, the standby node does not need to establish a heartbeat with the master node to directly perceive the status of the master node, and indirectly determines whether the master node is faulty by detecting the status flag of the master node stored in the storage device. In the event of a failure, it takes over the master node and provides external services, thereby improving application reliability.

With reference to the first aspect, in a possible implementation of the first aspect, the status flag is the heartbeat value of the master node; the standby node periodically detects the heartbeat of the master node stored in the storage device Whether the value is updated; if the heartbeat value of the master node is not updated, it is determined that the master node is faulty.

In the embodiment of this application, the master node will periodically update the heartbeat value stored in the storage device, for example, periodically increment by one, so that the standby node can determine whether the master node is faulty by periodically detecting the heartbeat value In this way, it can be ensured that the standby node can still accurately sense the status of the master node without establishing a heartbeat connection with the master node.

With reference to the first aspect, in a possible implementation of the first aspect, the storage device further stores the master node tag, and the backup node updates the master node tag in the storage device to the backup The label of the node.

In the embodiment of this application, the storage device only stores the label of one node (that is, the label of the primary node). When the standby node takes over the primary node, the standby node needs to update the label of the primary node in the storage device to its own label , So that other standby nodes can determine that there is a new primary node currently, avoiding access to storage devices, ensuring data consistency and application reliability.

With reference to the first aspect, in a possible implementation manner of the first aspect, the standby node writes its own mark to the storage device every first preset duration, and after the first preset duration , Reading the node label stored in the storage device every the first preset period of time; within the second preset period of time, when the node label continuously read by the standby node and the node label of the standby node itself If the tags are the same, stop writing their own tags to the storage device; the N is a positive integer greater than or equal to 1.

In the solution provided by this application, when the standby node competes for the primary node, it is achieved by writing its own mark to the storage device, and the rule that the mark written later overrides the mark written earlier is used for The mark written later has a greater probability of successful competition. If a backup node N consecutive times, for example, 3 times, the read mark is the same as its own mark, it can be considered that the backup node has successfully competed and become a new one. Master node. Through this kind of competition, the accuracy of selecting the master node can be improved, and multiple nodes can avoid accessing the storage device at the same time, and the reliability of the application can be ensured.

With reference to the first aspect, in a possible implementation manner of the first aspect, after the standby node takes over the master node, the heartbeat value stored in the storage device is cleared, and the heartbeat value is updated periodically.

In the embodiment of this application, after the standby node takes over the primary node, the storage device stores the mark of the standby node, and the stored heartbeat value is the heartbeat value of the original primary node. Therefore, the standby node needs to clear it. Zero indirectly informs other standby nodes that a new master node currently exists, and periodically updates the heartbeat value to make other standby nodes perceive their own status.

With reference to the first aspect, in a possible implementation of the first aspect, the standby node periodically reads the mark and heartbeat value stored in the storage device, and determines whether the mark and the mark of the standby node are Is the same, and whether the heartbeat value is the same as the heartbeat value written by the backup node in the previous cycle; it is determined that the tag stored in the storage device is the same as the tag of the backup node and the heartbeat value is the same as that of the backup node. The heartbeat value written by the node in the previous cycle is the same, and the standby node updates the heartbeat value.

In the embodiment of the present application, after the backup node takes over the master node, it needs to periodically update the heartbeat value so that other backup nodes can perceive its state. Before updating the heartbeat value each time, the standby node needs to determine whether the mark stored in the storage device is the same as its own mark and whether the heartbeat value is the same as the heartbeat value written in the previous cycle. Only the mark stored in the storage device is the same as the one written in the previous cycle. The heartbeat value is updated only when the own label is the same and the heartbeat value is the same as the heartbeat value written in the previous cycle. In this way, it can be ensured that the standby node can detect abnormal conditions in time and ensure the reliability of the application.

With reference to the first aspect, in a possible implementation of the first aspect, the backup node detects whether the storage device has a master node mark; if the storage device does not store the master node mark, the backup node The node takes over the master node.

In the embodiment of this application, the standby node will detect whether the primary node currently exists by detecting the primary node mark stored in the storage device when it is started. If the primary node does not currently exist, the secondary node will directly take over the primary node, and no cycle is required. The ability to detect the status mark of the master node stored in the storage device can improve the efficiency of competition.

In a second aspect, the present application provides a node including: a detection module for detecting the status flag of the master node stored in the storage device, and determining whether the master node is faulty according to the status flag, wherein: The master node is a node that accesses data in the storage device and provides services for users; a takeover module is used to take over the master node when the detection module determines that the master node is faulty according to the status flag .

With reference to the second aspect, in a possible implementation of the second aspect, the status flag is the heartbeat value of the master node; the detection module detects the status flag of the master node stored in the storage device, and When determining whether the master node is faulty according to the status flag, it is specifically used to: periodically detect whether the heartbeat value of the master node stored in the storage device is updated; if the heartbeat value of the master node is not updated, then It is determined that the master node is faulty.

With reference to the second aspect, in a possible implementation of the second aspect, the storage device further stores the master node identifier, and when the takeover module takes over the master node, it is specifically configured to: The label of the master node in the device is updated to the label of the node.

With reference to the second aspect, in a possible implementation of the second aspect, when the takeover module updates the label of the primary node in the storage device to the label of the standby node, it is specifically configured to: Writing the tag of the node device into the storage device for a first preset duration, and after the first preset duration, read the tag stored in the storage device every first preset duration; In the second preset period of time, when the consecutively read mark N times are the same as the written mark, stop writing the mark of the node to the storage device, where N is a positive integer greater than or equal to 1.

With reference to the second aspect, in a possible implementation of the second aspect, after the takeover module takes over the master node, it is also used to clear the heartbeat value stored in the storage device to zero, and periodically update all State the heartbeat value.

With reference to the second aspect, in a possible implementation of the second aspect, when the takeover module periodically updates the heartbeat value, the takeover module is specifically configured to: periodically read the data stored in the storage device And determine whether the label is the same as that of the standby node, and whether the heartbeat value is the same as the heartbeat value written by the standby node in the previous cycle; determine whether the storage device stores The mark of is the same as the mark of the standby node and the heartbeat value is the same as the heartbeat value written by the standby node in the previous cycle, and the heartbeat value is updated.

With reference to the second aspect, in a possible implementation of the second aspect, before the detection module detects the status flag of the master node stored in the storage device, the detection module is further configured to detect Whether there is a master node mark stored in the storage device; the takeover module is further configured to compete for the master node when the detection module detects that the storage device does not store the master node mark.

In a third aspect, the present application provides a computing device, the computing device includes a processor and a memory, the processor and the memory are connected by an internal bus, the memory stores instructions, and the processor calls all The instructions in the memory are used to execute the method for switching between nodes provided in the foregoing first aspect and in combination with any one of the foregoing first aspects.

In a fourth aspect, the present application provides a computer storage medium that stores a computer program. When the computer program is executed by a processor, it can implement the above first aspect and any combination of the above first aspect. A flow of the method for switching between nodes provided by an implementation method.

In a fifth aspect, the present application provides a computer program product. The computer program includes instructions. When the computer program is executed by a computer, the computer can execute the first aspect and any one of the implementations in the first aspect. The process of the switching method between the provided nodes.

Description of the drawings

In order to explain the technical solutions of the embodiments of the present invention more clearly, the following will briefly introduce the drawings used in the description of the embodiments. Obviously, the drawings in the following description are some embodiments of the present invention. Ordinary technicians can obtain other drawings based on these drawings without creative work.

FIG. 1 is a system architecture diagram of a communication system using SCSI protocol provided by an embodiment of the present application;

Figure 2 is a schematic diagram of an application scenario provided by an embodiment of the present application;

FIG. 3 is a schematic flowchart of an inter-node handover method provided by an embodiment of the present application;

4 is a schematic diagram of a state change of a container during a switching process according to an embodiment of the present application;

FIG. 5 is a schematic diagram of a sequence relationship in a competition process provided by an embodiment of the present application;

FIG. 6 is a schematic diagram of the structure of a node provided by an embodiment of the present application;

Fig. 7 is a schematic structural diagram of a computing device provided by an embodiment of the present application.

Detailed ways

The following describes the technical solutions in the embodiments of the present application clearly and completely in conjunction with the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments.

First of all, some terms and related technologies involved in this application will be explained with reference to the drawings, so as to facilitate the understanding of those skilled in the art.

The small computer system interface (SCSI) is an independent processor standard for system-level interfaces between computers and hardware devices (such as hard disks, optical drives, printers, scanners, etc.). SCSI is a universal interface. Host adapters and SCSI peripheral controllers can be connected to the SCSI bus. Multiple peripherals on a SCSI bus can work at the same time. The SCSI interface can transmit data synchronously or asynchronously.

A container is a virtualization technology in a computer operating system. This technology enables processes to run in a relatively independent and isolated environment (including independent file systems, namespaces, resource views, etc.), thereby simplifying the software deployment process and enhancing the software Portability and security, and improve system resource utilization.

Generally, in a cloud computing scenario, applications that provide services to users are deployed in virtual machines in the form of containers, and virtual machines are generally deployed in physical machines. In order to ensure the reliability of the application and prevent the application from being unavailable due to the failure of a single container, the main container and the backup container can be set. The main container and the standby container access the same storage device. At the same time, the main container can only read and write data in the storage device and provide services to the outside world. The standby container monitors the status of the main container. When the primary container fails, the standby container is upgraded to the primary container to provide external services.

As shown in Figure 1, it is a system architecture diagram in which multiple physical machines are connected to a storage device and communicate through the SCSI protocol. As shown in the figure, the physical machine 121, the physical machine 122, ..., the physical machine 12n are connected to the storage device 110 at the same time. One or more containers are deployed in each physical machine, and each container is deployed with applications that provide services for users. To ensure the reliability of the application and prevent the application from being unavailable due to a single container failure, the main container and the standby container can be set, and the main container and the standby container are located on different physical machines. To ensure that only the main container can access the storage device at the same time, the main container needs to lock the storage device exclusively through the physical machine where the main container is located. At present, the SCSI protocol provides a lock command for locking the storage device, and the physical machine can lock the storage device accessed by the container through the lock command on the container deployed therein. In practical applications, each container is actually allocated to a storage interval on the storage device, so the physical machine actually locks the storage space allocated to the container. After the storage device accessed by the main container is locked, the main container can access the storage device. However, after the main container fails, the lock added by the main container to the storage device cannot be released, that is, the lock remains. The physical machine where the standby container is located is connected to the physical machine where the main container is located through the network, so the standby container and the main container can also establish a connection. During the normal operation of the main container, for the standby container that has established a network connection with the main container , The standby container can periodically send heartbeat information to the main container. When the standby container does not receive a response from the main container within a period of time, it is determined that the main container is faulty, and the standby container can be upgraded to the main container. For the lock remaining due to the failure of the main container, the standby container can use the mandatory cover lock command provided in the SCSI protocol to cover the exclusive lock added by the main container, and add an exclusive lock to the storage device to access the storage device and externally Provide services. However, in some scenarios, the standby container is temporarily created when it needs to be used, and the physical machine where the temporarily created standby container is located may be the same as the physical machine where the main container is located, or it may not be the same as the main container. The new physical machine establishes a network connection, so the newly-built standby container cannot determine the state of the main container by heartbeat. As a result, when the main container fails, the standby container cannot be upgraded to the primary container in time, which affects the reliability of the application.

In order to solve the above problems, this application provides a method. Even when the network connection is not established between the main and standby containers, the standby container can detect the failure of the main container in time, thereby upgrading to the main container and continuing to provide services. .

Figure 2 shows a possible application scenario of an embodiment of the present application. As shown in FIG. 2, in this application scenario, the physical machine 2100 and the physical machine 2200 are connected to the storage device 2300. A virtual machine 2110 and a virtual machine 2120 are deployed in the physical machine 2100, a container 2111 is running in the virtual machine 2110, and a container 2121 is running in the virtual machine 2120; a virtual machine 2210 is deployed in the physical machine 2200, and a container 2211 is running in the virtual machine 2210 . Container 2111, container 2121, and container 2211 form a container cluster, where container 2111 is the main container, and container 2121 and container 2211 are standby containers. The same application is deployed in the main container and the backup container. During normal operation, the main container shown accesses the storage device 2300 to provide external services. In other embodiments, the container 2111, the container 2121, and the container 2211 may also be directly deployed on the physical machine.

In the embodiment of the present invention, the storage device 2300 may be a physical storage device, such as a storage array, or a hard disk, or a section of storage space on the physical storage device, allocated to the container 2111, the container 2121, and the container 2211 storage container Data generated by the deployed application. The storage device 2300 includes a mark storage area 2310, a heartbeat information storage area 2320, and a data storage area 2330. The mark storage area 2310 is used to store the mark of the main container, the heartbeat information storage area 2320 is used to store the heartbeat value of the main container, and the data storage area 2330 is used to store data generated during the operation of the main container. Among them, the main container (ie container 2111) can access all areas of the storage device 2300, while the backup container (ie container 2121 and container 2211) cannot access the data storage area 2330, but can access the mark storage area 2310 and the heartbeat information storage area 2320 to monitor the status of the main container.

When the container 2111 serves as the main container, the heartbeat value in the heartbeat information storage area 2320 is periodically updated, for example, it is periodically incremented by one. The container 2121 and the container 2211 respectively periodically monitor the mark storage area 2310 and the heartbeat information storage area 2320 to monitor the status of the container 2111. If it is found that the heartbeat value of the heartbeat information storage area 2320 exceeds a preset time (for example, two monitoring cycles) and is not updated, It can be determined that the container 2111 has failed and cannot provide services normally. At this time, the container 2121 and the container 2211 respectively write their own labels to the label storage area, and compete to select a new main container. The process of the container 2121 and the container 2211 competing for the main container will be described in detail below. Assuming that the container 2121 competes successfully and becomes the new main container, the label of the container 2121 will be stored in the label storage area 2310, and the label of the container 2111 will no longer be stored. The container 2121 will clear the heartbeat value in the heartbeat information storage area and proceed. Periodically update, and access the data in the data storage area 2330 to provide external services. If the container 2211 fails to compete, it will continue to monitor the mark storage area 2310 and the heartbeat information storage area 2320 to monitor the status of the container 2121.

By setting the mark storage area and the heartbeat information storage area in the storage area, even if there is no network connection between the standby container and the main container, the standby container can also write in the heartbeat information storage area 2320 by detecting the main container The heartbeat information determines the status of the main container, and can quickly select a new main container and provide services to the outside when the main container fails, ensuring the reliability of the service provided by the container.

In combination with the application scenario shown in Figure 2, the container switching method provided by the embodiment of the present application will be described below in conjunction with Figures 3 and 4. Figure 3 is a flowchart of the container switching method, and Figure 4 is the state change of the container during the container switching process Figure. This application takes any container as an example for detailed description. As shown in Figure 3, the method includes but is not limited to the following steps:

S301: When the container is started, it is detected whether the mark storage area 2310 of the storage device 2300 is written with the mark of the main container. If the mark storage area has written the mark of the main container, step S302 is executed. If the mark storage area is not written Enter the mark of the main container, step S303 is executed.

After the container is started, it first needs to detect whether the mark storage area 2310 of the storage device 2300 is written with the mark of the main container. The mark storage area 2310 in the storage device 2300 is used to store the mark of the main container. The size of the mark storage area 2310 can be set according to actual needs, for example, it can be set to 512 bytes, which is not limited in this application. The label of the container can uniquely identify a container.

If the mark storage area 2310 stores the mark of the main container, it means that the main container already exists in the container cluster where the container is located, and the container is in the standby state shown in FIG. 4 as a backup container.

If the mark storage area does not store the mark of the main container, it means that the main container does not exist in the container cluster where the container is located, and the container can compete for the main container. At this time, the container is in the election state shown in FIG. 4.

S302: The container periodically detects whether the heartbeat value of the heartbeat information storage area 2320 has changed within a preset period of time, if it changes, then continue to perform step S302; if there is no change, the container is in the state shown in FIG. 4 Election status, and step S303 is executed.

As shown in Figure 2, the heartbeat information storage area 2320 is used to store the heartbeat value of the main container. When the main container is running normally, the heartbeat value of the heartbeat information area 2320 will be updated periodically. If other backup containers detect the heartbeat If the value is always updated, it means that the main container is always working. If other backup containers detect that the heartbeat value has not been updated within the preset time period, for example, if the heartbeat value detected in two consecutive detection cycles is the same, it means that the main The container is malfunctioning.

It should be understood that when the backup container detects whether the heartbeat value of the main container is updated, it will record the heartbeat value read in the previous cycle, and then compare the heartbeat value read in the current cycle with the heartbeat value recorded in the previous cycle. If the heartbeat value read is the same as the recorded one, it means that the heartbeat value of the main container has not been updated; if the heartbeat value currently read is different from the recorded heartbeat value, it means that the heartbeat value of the main container has been updated.

Exemplarily, the heartbeat value of the primary container recorded by the backup container is 8, that is, the primary container updated the heartbeat value to 8 in the previous cycle, and the current heartbeat value read by the backup container is 9, and the backup container will compare the current read The standby container can determine that the heartbeat value has been updated by the main container, and the standby container continues to periodically detect the heartbeat value of the main container and update the recorded heartbeat value to 9.

When the main container updates the heartbeat value, it can first read the main container mark stored in the mark storage area 2310 and determine whether the read main container mark is the same as its own mark, and then read the heartbeat information storage area 2320. And determine whether the read heartbeat value is the same as the heartbeat value written in the previous cycle, if the read mark is the same as its own mark and the read heartbeat value is the same as the heartbeat value written in the previous cycle Similarly, the main container updates the heartbeat value, which can be a value, and the main container updates the heartbeat value by incrementing the value. For example, the heartbeat value stored in the heartbeat information storage area at the end of the previous cycle is 15. Then, the heartbeat value is updated to 16 in this cycle. At this time, the main container will still be in the main container state shown in Figure 4; if the read mark is different from its own mark or the read heartbeat value is equal to The heartbeat value written in one cycle is not the same, it means that the current system has an unpredictable failure, and the main container needs to perform state switching, as shown in Figure 4. At this time, the main container needs to exit and restart, and the entire container cluster needs to re-elect one The new main container, for example, the network of the main container is unstable, causing the main container to fail when writing its own mark to the mark storage area 2310 or updating the heartbeat value, but it cannot be sensed by itself, so the read main container's mark and It is different from itself; or the main container has been disconnected for a long time, and subsequently recovered (but the main container itself does not perceive it). At this time, another standby container has written a new mark in the mark storage area 2310, thus Causes the mark read by the main container to be different from its own.

It can be understood that the main container periodically reads the mark stored in the mark storage area 2310 and compares it with its own mark to determine whether to update the heartbeat value or exit and restart. This can be in extreme cases (for example, the unstable network of the main container causes the network The connection is intermittent, the storage device 2300 has an unpredictable failure that causes the content in the identification storage area to change, etc.), restart in time to avoid continuing to read and write the data in the storage device 2300 to ensure data consistency and application reliability .

S303: The container periodically writes its own mark into the mark storage area 2310, and reads the mark stored in the mark storage area 2310 in the same cycle.

When the container is started in step S301, it is determined that the mark storage area 2310 does not store the identifier of the main container, that is, the main container does not exist, or in step S302, it is determined that the heartbeat information in the heartbeat information storage area 2320 has not changed, That is, when the main container fails, the container is in the election state shown in FIG. 4 and can compete for the main container.

Since there may be multiple standby containers in a container cluster, such as the container 2121 and the container 2211 in FIG. 2, when the primary container does not exist in the container cluster or the primary container fails, the multiple standby containers will simultaneously compete for the primary container. As shown in Figure 5, during the competition, the container 2121 first writes its own label 1 to the label storage area 2310 at time t1, and the container 2211 writes its own label 2 to the label storage area 2310 at time t2, and t1 is less than t2. Therefore, the mark 2 written by the container 2211 will overwrite the mark 1 written by the container 2121, that is, the mark 2 is stored in the mark storage area 2310. After the container 2121 writes the mark 1, after a period of time (for example, the duration of a sleep cycle), the mark stored in the mark storage area 2310 is read at time t3. At this time, the mark read by the container 2121 is mark 2. After an interval of a sleep cycle, write its own mark to the mark storage area 2310 again at t5, that is, mark 1. After the container 2211 writes mark 2, an interval of sleep cycle, read the stored mark storage area 2310 at t4 At this time, the label read by the container 2211 is label 2, and after another sleep period, at time t6, write its own label, that is, label 2 to the label storage area 2310 again.

S304: The container detects whether the label stored in the label storage area 2310 read continuously for N times is the same as the label written by the container within a preset period of time, if they are the same, go to step S305; if they are not the same , Go to step S302.

Specifically, it can be seen from the description of FIG. 4 that each backup container needs to write its own mark to the mark storage area 2310 first, then read it, and then write it again, and the cycle repeats periodically. For the competition method shown in FIG. 5, for the container that writes its own label first, for example, container 2121, the label it reads each time is the label of the container written last. The mark of is not the same as the written mark, but for the last container to write its own mark, the mark read each time is the same as the written mark, such as container 2211, if within a preset time period If the label read by the container for N consecutive times is the same as the label written, it can be determined that the container has been successfully upgraded to the new main container, and other containers will abandon this competition and no longer write themselves to the mark storage area 2310 Mark, re-check the heartbeat value of the heartbeat information storage area 2320, and wait for the next competition. N is a positive integer greater than or equal to 1, for example, 3 or 4, which is not limited in this application.

It can be seen that in the process of selecting a new main container through competition, each container periodically executes writing marks and reading marks and judging whether they are consistent, so as to finally determine the only mark that is read every time and write mark. Import the container with the same label and upgrade the container to the main container, which can improve the accuracy of the main container selection and ensure that the elected main container is the only one, so as to avoid the situation that multiple containers access the storage device at the same time, and ensure the application Reliability.

S305: The container is upgraded to the main container, and the data in the storage device is accessed to provide external services.

Specifically, after the container is determined to be upgraded to the primary container, it will access the data in the storage device to provide services to the outside world, clear the heartbeat value of the heartbeat information storage area 2320, and then periodically update the heartbeat value At this time, the container is in the main container state shown in FIG. 4.

It should be understood that steps S301 to S305 involved in the foregoing method embodiments are only schematic descriptions and summaries, and should not constitute specific limitations. The involved steps can be added, reduced, or combined as needed.

The embodiments described in Figures 3 and 4 are scenarios where there are multiple standby containers in a container cluster, and when the primary container fails or there is no primary container, multiple standby containers compete to become the primary container, but for the container cluster only In the scenario of a standby container, in step S302, if it is detected that the heartbeat of the primary container is not updated, that is, when the primary container fails, the standby container can be directly upgraded to the primary container, that is, the mark of the standby container is directly written The storage area is marked without performing step S304, and the step of determining the main container through multiple writing and reading.

In addition, although the foregoing embodiment is described with a container as an example, the method provided by the present invention is also suitable for switching between a physical machine and a virtual machine. With regard to the switching method of the physical machine and the virtual machine, except for the different objects of the switching, the other switching methods are the same as those of the container, which will not be repeated here.

In the embodiment of the present invention, by setting the status flag of the master node, such as the heartbeat information of the master node, in the storage device, and the master node periodically updates the heartbeat information, the standby node will also periodically detect the heartbeat information. However, if the heartbeat information is not updated, it can compete for the master node. In this way, even if the master node and the backup node have not established a network connection through their respective physical machines, the backup node can detect the failure of the master node in time and compete for the master node. The node continues to provide services.

In addition, since in the embodiment of the present invention, there is no need to establish a network connection between the master node and the backup node, if the master node and the backup node are both physical machines, the backup physical machine does not need to establish a network connection with the master physical machine. It can be determined whether the main physical machine is faulty. For the primary node and the standby node are virtual machines and containers, the standby node can be deployed on any physical machine, for example, deployed on the same physical machine as the primary node, or deployed on a physical machine that has no network connection with the primary node , Thereby reducing the constraints of virtual machine and container deployment.

The foregoing describes the methods of the embodiments of the present application in detail. In order to facilitate better implementation of the above-mentioned solutions of the embodiments of the present application, correspondingly, related equipment for cooperating with the implementation of the foregoing solutions is provided below.

Refer to Fig. 6, which is a schematic structural diagram of a node provided by an embodiment of the application. As shown in FIG. 6, the node 600 includes a detection module 610 and a takeover module 620. among them,

The detection module 610 is used to detect the status flag of the master node stored in the storage device, and determine whether the master node is faulty according to the status flag, wherein the master node accesses data in the storage device And provide service nodes for users.

Specifically, the detection module 610 is configured to execute the aforementioned steps S301, S302, and S304, and optionally execute optional methods in the aforementioned steps.

The takeover module 620 is configured to take over the master node when the detection module 610 determines that the master node is faulty according to the status flag.

Specifically, the takeover module 620 is configured to perform the foregoing steps S303 and S305, and optionally perform optional methods in the foregoing steps.

In a possible implementation, the status flag is the heartbeat value of the master node; the detection module 610 detects the status flag of the master node stored in the storage device, and determines the status flag according to the status flag. When the master node is faulty, it is specifically used to: periodically detect whether the heartbeat value of the master node stored in the storage device is updated; if the heartbeat value of the master node is not updated, determine that the master node is faulty.

In a possible implementation, the storage device also stores the master node tag, and when the takeover module 620 takes over the master node, it is specifically configured to update the master node tag in the storage device to The label of the node device.

In a possible implementation, when the takeover module 620 updates the label of the primary node in the storage device to the label of the standby node, it is specifically configured to: The tag of the node device is written into the storage device, and after the first preset duration, the tag stored in the storage device is read every first preset duration; within the second preset duration, When the consecutively read N tags are the same as the written tags, stop writing the tag of the node device to the storage device, where N is a positive integer greater than or equal to 1.

In a possible implementation, after the takeover module 620 takes over the master node, it is further configured to clear the heartbeat value stored in the storage device to zero, and periodically update the heartbeat value.

In a possible implementation, when the takeover module 620 periodically updates the heartbeat value, the takeover module 620 is specifically configured to periodically read the mark and the heartbeat value stored in the storage device, and determine Whether the mark is the same as the mark of the standby node, and whether the heartbeat value is the same as the heartbeat value written by the standby node in the previous cycle; determine whether the mark stored in the storage device is the same as that of the standby node If the tags are the same and the heartbeat value is the same as the heartbeat value written by the standby node in the previous cycle, the heartbeat value is updated.

In a possible implementation, before the detection module 610 detects the status mark of the master node stored in the storage device, the detection module 610 is also used to detect whether the storage device stores the master node. Node tag; the takeover module 620 is also used to take over the master node when the detection module detects that there is no master node tag stored in the storage device.

It should be understood that the structure of the node described above is merely an example, and should not constitute a specific limitation, and various modules of the node can be added, reduced, or combined as needed. In addition, the operations and/or functions of each module in the node are used to implement the corresponding process of the method described in FIG. 3 above. For brevity, details are not repeated here.

Refer to FIG. 7, which is a schematic structural diagram of a computing device provided by an embodiment of the application. As shown in FIG. 7, the computing device 700 includes a processor 710, a communication interface 720, and a memory 730. The processor 710, the communication interface 720, and the memory 730 are connected to each other through an internal bus 740. It should be understood that the computing device may be a database server.

The computing device 700 may be the physical machine 2110 or 2120 in FIG. 2, in which a container or a virtual machine is built. The functions performed by the container in FIG. 2 are actually performed by the processor 710 of the physical machine.

The processor 710 may be composed of one or more general-purpose processors, such as a central processing unit (CPU), or a combination of a CPU and a hardware chip. The aforementioned hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The aforementioned PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a general array logic (generic array logic, GAL), or any combination thereof.

The bus 740 may be a peripheral component interconnect standard (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus 740 can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is used in FIG. 7, but it does not mean that there is only one bus or one type of bus.

The memory 730 may include a volatile memory (volatile memory), such as a random access memory (random access memory, RAM); the memory 730 may also include a non-volatile memory (non-volatile memory), such as a read-only memory (read-only memory). Only memory (ROM), flash memory (flash memory), hard disk drive (HDD) or solid-state drive (SSD); memory 730 may also include a combination of the above types. The program code may be used to implement the functional modules shown in the node device 600, or to implement the method steps in the method embodiment shown in FIG. 6 with the standby node as the execution subject.

The embodiments of the present application also provide a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, it can implement part or all of the steps of any one of the above method embodiments, and realize the above The function of any one of the functional modules described in Figure 6.

The embodiments of the present application also provide a computer program product, which when it runs on a computer or a processor, enables the computer or the processor to execute one or more steps in any of the foregoing methods. If each component module of the aforementioned equipment is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in the computer readable storage medium.

In the above-mentioned embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in an embodiment, reference may be made to related descriptions of other embodiments.

It should also be understood that in the various embodiments of the present application, the size of the sequence number of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not be implemented in this application. The implementation process of the example constitutes any limitation.

A person of ordinary skill in the art may be aware that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software depends on the specific application and design constraint conditions of the technical solution. Professionals and technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered beyond the scope of this application.

Those skilled in the art can clearly understand that, for the convenience and conciseness of description, the specific working process of the above-described system, device, and unit can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.

In the several embodiments provided in this application, it should be understood that the disclosed system, device, and method may be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division, and there may be other divisions in actual implementation, for example, multiple units or components can be combined or It can be integrated into another system, or some features can be ignored or not implemented. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may be in electrical, mechanical or other forms.

In addition, the functional units in each embodiment of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit.

If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on this understanding, the technical solution of this application essentially or the part that contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including Several instructions are used to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other media that can store program code .

The steps in the method of the embodiment of the present application can be adjusted, merged, and deleted in order according to actual needs.

The modules in the devices in the embodiments of the present application may be combined, divided, and deleted according to actual needs.

As mentioned above, the above embodiments are only used to illustrate the technical solutions of the present application, not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, a person of ordinary skill in the art should understand that: The technical solutions recorded in the embodiments are modified, or some of the technical features are equivalently replaced; these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

A method for switching between nodes, characterized in that the method includes:

The standby node detects the status flag of the master node stored in the storage device, and determines whether the master node is faulty according to the status flag, wherein the master node is a node that accesses data in the storage device and provides services for users ；

When the standby node determines that the primary node is faulty according to the status flag, the standby node takes over the primary node.
The method according to claim 1, wherein the status mark is a heartbeat value of the master node;

The detection of the status flag of the master node stored in the storage device by the standby node, and determining whether the master node is faulty according to the status flag includes:

The standby node periodically detects whether the heartbeat value of the master node stored in the storage device is updated;

If the heartbeat value of the master node is not updated, it is determined that the master node is faulty.
The method according to claim 1 or 2, wherein the storage device further stores the master node mark, and the backup node taking over the master node comprises:

The standby node updates the label of the master node in the storage device to the label of the standby node.
The method according to claim 3, wherein the updating of the label of the primary node in the storage device by the standby node to the label of the standby node comprises:

The standby node writes its mark to the storage device every first preset time period, and after the first preset time period, reads it from the storage device every first preset time period Stored node tag;

In the second preset time period, when the node label continuously read N times by the standby node is the same as the label of the standby node, it stops writing its own label to the storage device; the N is greater than or equal to A positive integer of 1.
The method of claim 2, wherein the method further comprises:

After the standby node takes over the master node, it clears the heartbeat value stored in the storage device to zero, and periodically updates the heartbeat value.
The method according to any one of claims 1 to 5, wherein before the standby node detects the status flag of the master node stored in the storage device, the method further comprises:

The standby node detects whether a master node mark is stored in the storage device;

If the primary node mark is not stored in the storage device, the backup node competes for the primary node.
A node, characterized in that it includes:

The detection module is used to detect the status flag of the master node stored in the storage device, and determine whether the master node is faulty according to the status flag, wherein the master node is a user for accessing data in the storage device Nodes that provide services;

The takeover module is configured to take over the master node when the detection module determines that the master node is faulty according to the status flag.
The node according to claim 7, wherein the status flag is the heartbeat value of the master node; the detection module detects the status flag of the master node stored in the storage device, and based on the status When the mark determines whether the master node is faulty, it is specifically used for:

Periodically detecting whether the heartbeat value of the master node stored in the storage device is updated;

If the heartbeat value of the master node is not updated, it is determined that the master node is faulty.
The node according to claim 7 or 8, wherein the storage device further stores the master node mark, and when the takeover module takes over the master node, it is specifically configured to use the storage device The label of the master node is updated to the label of the node.
The node according to claim 8, wherein the takeover module is specifically configured to: when updating the label of the primary node in the storage device to the label of the standby node:

Write the tag of the node device to the storage device every first preset time length, and read the storage device from the storage device every first preset time after the first preset time length Mark of

In the second preset period of time, when the consecutively read N tags are the same as the written tags, stop writing the tag of the node to the storage device, where N is a positive integer greater than or equal to 1.
The node device according to claim 8, wherein:

After the takeover module takes over the master node, it is also used to clear the heartbeat value stored in the storage device to zero, and periodically update the heartbeat value.
The method according to any one of claims 7-11, wherein, before the detection module detects the status mark of the master node stored in the storage device, the detection module is further configured to detect the Whether the master node mark is stored in the storage device;

The takeover module is further configured to compete for the master node when the detection module detects that the storage device does not store the master node mark.
A computing device, characterized by comprising a processor and a memory, the processor and the memory are connected through an internal bus, the memory is stored with instructions, and the processor calls the instructions in the memory to execute such as The method of any one of claims 1-6.