CN106201773A - The upgrading of instance system, cloud system and instance system or restoration methods - Google Patents

Abstract

The present invention provides upgrading or the restoration methods of instance system, cloud system and instance system, in instance system, it is provided with two recovers subregion and two regeneration blocks, the reforestation practices of the operating system in support instance system, it is ensured that operating system can successfully be upgraded or recover under multiple abnormal conditions.Additionally, cloud system is provided with ARM server, instance system is provided with agency, constitutes the dual fail-safe of trigger action system upgrade or recovery.

Description

Instance system, cloud system and instance system upgrade or recovery method

technical field

The invention relates to an instance system, a cloud system and an upgrade or restoration method of the instance system.

Background technique

A certain operating system (such as the Android operating system) plus hardware (such as a processor, memory, etc.) required to run the operating system can be regarded as an instance system, and various applications can run in the instance system. A cloud system can be formed by centrally managing several instance systems according to a certain architecture (such as distributed).

For Android systems running on mobile terminals, set-top boxes, game consoles, etc., when the system needs to be upgraded or restored, the user will be prompted through a graphical interface, and the system will store (for example, in a file in a partition) the corresponding information and restart. When the system restarts and checks the corresponding information, it will run in recovery mode and use the corresponding files to upgrade or restore the system. The stand-alone Android system requires the user to ensure uninterrupted power supply when upgrading or restoring. Once there is a power failure in the middle, the system can only be restored manually.

For a cloud system that generally runs in a dedicated computer room, the operating system in the instance system also requires an uninterrupted power supply during upgrade or recovery, which is very difficult in reality. On the one hand, the computer room cannot guarantee uninterrupted power supply throughout the year, and the circuit boards in the instance system may occasionally experience power failures, resulting in random restarts; on the other hand, due to the complexity of managing a large number of instance systems, sometimes the operating system in some instance systems The situation where the operation and maintenance personnel cut off the power for maintenance while it is being upgraded. For a large-scale system such as a cloud system that provides services to the outside world without interruption all year round, manual flashing to restore the operating system is neither economical nor efficient, which seriously affects the user experience. At the same time, the upgrade or recovery of the operating system requires the confirmation of the user, which is obviously unacceptable for cloud systems that are automatically deployed and operated.

In order to solve the above problems, an ARM server (Linux server with ARM architecture) can be set up, its hardware circuit is connected to the circuit board of the example system, and the hard restart of the example system is controlled by changing the level of the connection pipe. However, the ARM server cannot guarantee that it will not fail, and if the ARM server adopts a dual-machine cluster (HA), the cost will be significantly increased.

The content disclosed in the "Background Technology" part of the description is helpful for those skilled in the art to understand the technical solution of the present invention, but it should not be considered that these content must belong to the prior art or common knowledge.

Contents of the invention

In order to overcome the defects reflected in the "Background Technology" section and realize the function of automatically re-executing the upgrade or recovery task after the operating system in the example system is interrupted unexpectedly, the present invention provides the example system, the cloud system and the upgrade or recovery method of the example system.

Example systems, including:

The first recovery partition and the second recovery partition are used to store data required when the operating system in the example system enters the recovery mode;

The first update partition and the second update partition are used to store images of data in other partitions;

The data stored in the first recovery partition and the first update partition remain unchanged, and the data stored in the second recovery partition and the second update partition can be modified;

An agent with the root authority of the operating system in the example system (the agent here is the translation of the English word agent, which refers to a certain hardware or software that has certain functions and can run independently).

Further, the operating system in the example system is the Android system.

Cloud system, including control server, ARM server and example system:

The ARM server manages multiple instance systems, and the control server can not only communicate with multiple ARM servers, but also communicate with agents of multiple instance systems managed by each ARM server.

Further, the control server can choose to control the ARM server or the agent that controls the instance system to upgrade or restore the operating system in the instance system;

When the control server chooses to control the ARM server to upgrade or restore the operating system in the instance system, the ARM server will force the operating system in the instance system to restart through electrical signals;

When the control server chooses to control the agent of the instance system to upgrade or restore the operating system in the instance system, the agent of the instance system sets a file in the corresponding partition to record the tasks to be performed (task refers to system recovery and/or system upgrade).

Further, the example system includes a boot partition, and the programs in the boot partition can be used to set the priority of upgrading or restoring the operating system in the example system according to the ARM server or the agent of the example system.

The upgrade or recovery method of the example system includes the following steps:

Start the operating system in the instance system;

The program in the boot partition of the example system judges whether the operating system needs to enter the recovery mode according to the electrical signal of the ARM server or the file set by the agent in the corresponding partition;

If there is no need to enter the recovery mode, the operating system runs normally, otherwise enter the recovery mode, and determine the task to be performed according to the electrical signal of the ARM server or the file set by the agent in the corresponding partition (task refers to system recovery and/or system upgrade) ;

Read data from the corresponding recovery partition and/or update partition to complete the task.

Further, after the task is completed, restart the operating system.

Further, after the task is completed, delete the file set by the agent in the corresponding partition (if the system recovery and/or system upgrade task is performed according to the file set by the agent in the corresponding partition).

Since the upgrading or restoring method of the example system is implemented on the basis of the aforementioned example system and cloud system, in order to avoid unnecessary repetition, some descriptions in the upgrading or restoring method of the example system are omitted. Those skilled in the art can have a clear and complete understanding of the upgrade or recovery method of the example system through comparison.

In the technical solution of the present invention, words such as "comprising" and "for" should be understood in an open-ended manner. What those skilled in the art can know by reading this specification in combination with the prior art or common knowledge will not be repeated in this specification.

The example system, the cloud system, and the upgrade or recovery method of the example system provided by the present invention set two recovery partitions and two update partitions in the example system, support the recovery mode of the operating system in the example system, and ensure In this case, the operating system can be successfully upgraded or restored. In addition, the ARM server is set up in the cloud system, and the agent is set up in the instance system, which constitutes a double insurance that triggers the upgrade or recovery of the operating system. If the external ARM server fails, the instance system can upgrade or restore the operating system through its own agent. Conversely, if the agent cannot operate normally due to the problem of the instance system itself, the ARM server can force the operating system to be upgraded through electrical signals. or restore. When the operating system encounters unexpected situations such as power failure during the upgrade or recovery process, it can automatically re-execute the upgrade or recovery task after the power supply is restored or other accidents are excluded to ensure the success of the upgrade or recovery.

Description of drawings

Fig. 1 is a schematic diagram of a simplified structure of a cloud system in a specific embodiment. The lines in the figure represent connections or communication relationships.

FIG. 2 is a flowchart of an example system upgrade or restore method in a specific embodiment.

detailed description

Embodiments of the present invention will be further described in detail below. However, it should be noted that the scope of the present invention is not limited to the specific technical solutions described. Any technical solution obtained by performing the same or equivalent replacement of the technical elements in the described specific technical solution or a technical solution that can be obtained by those skilled in the art without creative work on the basis of the described specific technical solution shall be regarded as For falling into the protection scope of the present invention.

A certain operating system plus hardware (such as a processor, memory, etc.) required to run the operating system can constitute an instance system. A cloud system can be formed by centrally managing several instance systems according to a certain architecture. The operating system in the example system may be an Android system. In order to run the Android system normally, partitions as shown in the table below are set on the memory of the instance system.

uboot kernel ramdisk cache data system resources recovery1 recovery2 update1 update2

The first recovery partition (recovery1) and the second recovery partition (recovery2) are used to store data (including programs, configuration files, etc.) required when the Android system enters the recovery mode.

The first update partition (update1) and the second update partition (update2) are used to store images of data in other partitions (such as kernel, ramdisk, system, resource partitions, etc.).

During the running of the Android system, the data in the recovery2 and update2 partitions can be modified to store newer dynamic update data; but the data in the recovery1 and update1 partitions remains unchanged, storing the initial state data of the Android system (factory settings) . If the Android system uses the data in the recovery2 and/or update2 partitions for system upgrade, and an error occurs that seriously affects the system operation, you can use the unchanged data in the recovery1 and/or update1 partitions to return to the initial state of the Android system to avoid Android The failure of the system to operate normally affects the use of users.

In the example system, an agent (agent) with the root authority of the Android system also needs to be set. The simpler implementation of the agent is an application software with the root authority of the Android system. The application software is designated as a system service by the init.rc script when the Android system starts, automatically starts, and specifies that the application software can be rerun immediately if the process corresponding to the application software is killed.

The cloud system includes a control server, an ARM server, and an instance system. A simplified structure of the cloud system is shown in Figure 1. In Figure 1, it is assumed that the cloud system includes a control server, two ARM servers and six instance systems, and each ARM server manages three instance systems. The control server has an X86 architecture and may be called an X86controller. An ARM server is a Linux server with an ARM architecture and can be called an arm controller. In a cloud system, one arm controller can manage multiple instance systems. The hardware circuit of the arm controller is connected to the circuit board where the processor running the Android system in the example system is located, and the Android system can be forced to restart through electrical signals, which is specifically realized by the combination of the GPIO level of the connecting pins.

The X86 controller can communicate with multiple arm controllers, and also communicate with the agent of each instance system managed by each arm controller, and control the upgrade or recovery of the Android system by controlling the arm controller or agent. The communication described above may be in a wired or wireless manner. When the agent of the example system communicates with the X86controller, it can receive instructions from the X86controller and download the corresponding files. For example, download update.img to this example system and write it to the update2 partition, or download recovery.img to this example system and write it to the recovery2 partition.

Correspondingly, the X86 controller can choose to control the agent or arm controller of the instance system, use the data in the update1 partition to restore the Android system, or use the data in the update2 partition to upgrade the Android system. When the X86controller selects the agent that controls the instance system to restore or upgrade the Android system, the agent will set a file in the data or cache partition to record the tasks to be performed after entering the recovery mode, such as data recovery using the update1 partition For the Android system, use the data in the update2 partition to upgrade the Android system; after restarting, use the program in the recovery1 partition or the program in the recovery2 partition. After the above files are completed, the agent will restart the Android system.

The uboot partition is the boot partition of the Android system in the example system. When the Android system starts, the program in the uboot partition is executed first. The program in the uboot partition can set the priority of upgrading or restoring the Android system according to the armcontroller or the agent of the instance system through logical operations. For example, the Android system upgrade or recovery based on the arm controller can be set to be prioritized over the upgrade or recovery based on the agent.

Based on the example system and cloud system described above, if the Android system in the example system encounters a power outage or other unexpected circumstances while running in recovery mode, resulting in incomplete partition data such as system and kernel, and the Android system cannot run normally, you can use the arm controller Set an appropriate electrical signal trigger, and enter the recovery mode again until it succeeds. You can also enter the recovery mode by itself through the task file set by the agent in the corresponding partition, and re-execute the unfinished system upgrade or recovery task. The above-mentioned process can be described with the upgrade or recovery method of the example system shown in Figure 2 (i.e. the upgrade or recovery method of the operating system in the example system), the method includes the following steps:

S201: start the Android system;

S202: the program in the uboot partition judges whether the Android system needs to enter the recovery mode according to the electrical signal of the arm controller or the file set by the agent in the corresponding partition;

If there is no need to enter recovery mode, the operating system is running normally, otherwise

S203: enter the recovery mode, and determine the tasks to be executed according to the electrical signal of the arm controller or the file set by the agent in the corresponding partition;

S204: Read data from the corresponding recovery partition (including recovery1 and recovery2) and/or update partition (including update1 and update2), and complete the task;

S205: Restart the Android system.

In addition, if the recovery or upgrade task of the Android system is based on the files set by the agent in the corresponding partition, the corresponding files are deleted after the task is completed, so as to avoid calling these files again when the Android system starts normally next time.

Those skilled in the art can completely construct other solutions on the basis of the specific technical solutions described above. For example, the operating system in the example system may adopt other operating systems, and the partitions on the memory of the corresponding example system may change. In the upgrade or recovery method of the example system, if some operating systems do not need to be restarted to complete the recovery or upgrade task, step S205 is deleted. I won't list them all here.

Claims (8)

1. instance system, it is characterised in that described instance system includes:

First recovers subregion and second recovers subregion, for storing needs when the operating system in instance system enters reforestation practices Data；

First regeneration block and the second regeneration block, for storing the mirror image of other partition datas；

First data recovering subregion and the first regeneration block storage keep constant, and the second recovery subregion and the second regeneration block are deposited The data of storage can be modified；

There is the agency of operating system root authority in instance system.

Instance system the most according to claim 1, it is characterised in that the described operating system in instance system is Android System.

3. cloud system, including controlling server, ARM server and the instance system described in claim 1 or 2, it is characterised in that The described multiple instance system of ARM server admin, described control server can either be with multiple stage ARM server communication, also The agent communication of the multiple instance systems that can be managed with every ARM server.

Cloud system the most according to claim 3, it is characterised in that described control server can select to control ARM clothes The agency of business device or control instance system carries out upgrading or the recovery of operating system in instance system；

When controlling server and selecting control ARM server carry out the upgrading of operating system in instance system or recover, ARM services Device forces the operating system in instance system to restart by the signal of telecommunication；

When controlling server and selecting the agency controlling instance system carry out the upgrading of operating system in instance system or recover, real The agency of example system arranges file in corresponding subregion, and record needs performing of task.

Cloud system the most according to claim 4, it is characterised in that described instance system includes boot partition, starts and divides Program in district can be used in arranging the agency according to ARM server or instance system and carries out operating system in instance system Upgrading or the priority recovered.

6. the upgrading of instance system or restoration methods, be characterised by, described method comprises the steps:

Start the operating system in instance system；

The literary composition that program in instance system boot partition is arranged in corresponding subregion according to the signal of telecommunication or the agency of ARM server Part, it is judged that operating system is the need of entering reforestation practices；

If need not enter reforestation practices, then operating system is properly functioning, otherwise enters reforestation practices, according to ARM server The signal of telecommunication or the file that arranges in corresponding subregion of agency, determine the task of needing to perform；

Read data from corresponding recovery subregion and/or regeneration block, complete task.

Method the most according to claim 6, it is characterised in that described complete task after, restarting operating systems.

Method the most according to claim 6, it is characterised in that described complete task after, delete agency at corresponding subregion The file of middle setting.