KR19990053243A - Circular Redundancy in Open Embedded Systems - Google Patents

Abstract

1. TECHNICAL FIELD OF THE INVENTION

The present invention relates to a cyclic redundancy method in an open embedded system.

2. The technical problem to be solved by the invention

According to the present invention, a redundant and redundant system is configured in a system having the same shape to allow and recover the failures that can be prevented at a specific point provided by the high-availability system and the fault-tolerant system, and the main system manages the entire resource. If the primary system fails, we want to provide a cyclic redundancy method that all the secondary systems manage.

3. Summary of Solution to Invention

The present invention provides a first step in which a main / sub system configures cyclic redundancy to store shape data and memory contents and to monitor a shared disk area; A second step of executing / waiting if a system fails to recover from a single fault before detecting a critical error for transferring control to a circular standby system; And a third step of copying the special area disk contents and switching the central processing unit to the circulatory system when a switch request event occurs.

4. Important uses of the invention

The present invention is used in an open embedded system.

Description

Circular Redundancy in Open Embedded Systems

The present invention provides a flexible and fault-tolerant system supported by a high availability system using a shareable storage device having a single point of failure (SPOF) function when an open embedded system is applied to a communication system such as an exchange. Allows and recovers faults that can be prevented at specific points, which are redundantly configured in a system with the same shape, one system manages all resources, and if the main system fails, all the resources of the main system are managed by the secondary system. It is a cyclic redundancy method.

Control of communication system such as exchange system not only uses processor embedded in the system with high real time and high reliability but also adds hardware such as central processing unit, power module, communication module to ensure system reliability. We have provided a fault tolerant system that consists of hardware that does not require software.

In addition, with the development of processors and other technologies, as processor prices become cheaper and reliability increases, communication channels, disk units, central processing units, and power supplies are provided to solve problems caused by specific hardware failures. A high available system that increases the usability in software by duplexing devices and the like has been applied as an alternative.

In the case of a fault-tolerant system, development time for hardware configuration without the help of software is later than the price drop due to the development of system element technology and the development of technology, so that the current technology cannot be used at all. In the case of high-availability systems, when an additional element is allowed to fail, failure of the element, that is, single fault problem, can be solved through the software recovery step, but there is a case where it is impossible to recover the failure by software such as memory. do.

In particular, the single fault guarantee technology in shared disks has already been proposed through methods such as disk mirroring, and the system bus expansion or system redundancy method through the network has been proposed. Due to the delay of the message, the change of order, and the occurrence of error, it is not only more complicated than the method using the shared disk, but also has the disadvantage of hard to guarantee the reliability, and the method through the system bus expansion requires much development cost and time. There was a problem of low economic efficiency.

Accordingly, the present invention has been made in order to solve the above problems, it is configured using the elements used in the general system software and hardware required for the system, such as operating system, central processing unit, memory, communication method, Redundancy such as memory that does not provide reliability in high-availability system is provided by system redundancy method, and it is possible to remove elements that cannot eliminate single point of failure through shared disk channel for communication system using open embedded system. It is an object of the present invention to provide a cyclic redundancy method for providing a cyclic redundancy method.

1 is an exemplary configuration of a general open embedded system to which the present invention is applied.

2 is a structural diagram of a shared disk area of FIG. 1;

3 is a structural diagram of data for memory permanent preservation.

4 is a flowchart of a cyclic redundancy method according to an embodiment of the present invention.

* Explanation of symbols for the main parts of the drawings

100: main system 110: sub-system

101, 111: central processing unit 102, 112: cache

103, 113: memory 104, 114: bus

105, 115: network 106, 116: disk controller

107, 117: input and output interface 108, 118: other devices

120: network connector 130: storage device

According to an aspect of the present invention, there is provided a circular redundancy method applied to an open embedded system, comprising: a first step in which a main / sub system configures circular redundancy to store shape data and memory contents and to monitor a shared disk area; A second step of executing / waiting if a system fails to recover from a single fault before detecting a critical error for transferring control to a circular standby system; And a third step of copying the contents of the special area disk and converting the central processing unit into the circulatory system when the switching request event occurs.

Hereinafter, an embodiment according to the present invention will be described in detail with reference to FIGS. 1 to 4.

1 is an exemplary configuration of a general open embedded system to which the present invention is applied.

As shown in the figure, the first system 100 and the second system 110 are general systems composed of the same elements, and the central processing units 101 and 111 and caches 102 and 112 for compensating memory speeds. And the memory 103 and 113, the buses 104 and 114 for communication, the network 105 and 115, the disk controllers 106 and 116, the network 105 and 115 and the disk controller 106, Input / output interfaces 107 and 117 for connecting elements such as 116 to the central processing module, network connector 120 for connecting the two systems, storage 130, and other devices 108 and 118. Equipped.

In high-availability systems, they are duplicated to prevent single points of failure such as communication processing devices and storage devices, which are prone to failure, and there is no organic relationship between the systems.

The invention consists of a storage device in which the disk and the bus are redundant, i.e. without a single fault.

FIG. 2 is a structural diagram of the shared disk area of FIG.

As shown in the figure, the system is divided into the original disk partition portion 200 and the partition portion 210 for the cyclic redundancy configuration, the partition portion 210 for the cyclic redundancy configuration can store the shape information. The shape information area 211, the central processing unit state information area 212 which stores the state information of the central processing unit, and the recovery swap area 213 are comprised.

Allocating a specific area for storing system memory on the disk means that the cyclic redundancy system allocates a specific area of the disk so that the operating state of the other party can be stored.

3 is a structural diagram of data for memory permanent preservation, and is a data structure necessary for updating a memory page which is changed in the physical memory 310.

The virtual memory area 300 is a data structure 301 that exists in the memory and is managed by the operating system. The physical memory 310 or the disk 320 indicates the swap partition 321.

In the case of the swap partition 321, even if the control is transferred to the secondary system, which is a circular system, since it exists on the shared disk 320, no separate recovery is required, but when the physical memory 310 is designated, the swap partition 322 is used. Must be stored.

4 is a flowchart of a cyclic redundancy method according to an embodiment of the present invention.

As shown in the figure, the main / sub system detects the critical error of passing control to the rotating standby system when a preparatory process for configuring the cyclic redundancy occurs and when the system cannot recover from a single point of failure. It can be divided into the execution / waiting phase until the transition and the transition to the circulatory system after the transition request event occurs.

The preparatory process determines which system will be the primary system in operation, or a standby system in the standby state (401), and in the case of the primary system, its shape: the size of memory, the version of the operating system, Shape data such as boot time and type of hardware is stored (402), and in the case of the main system, the entire contents of the physical memory are copied to a backup swap partition which is a specific area of the disk to preserve the data of the current memory (403). .

In order for the system to determine execution / standby, a flag exists in a specific area on the shared disk so that the two systems are accessed independently, so that the value of the special area in the shared disk is correct.

The secondary system stores the shape of the main system, that is, the size of the memory, the version of the operating system, the boot time, the type of hardware possessed, etc. (411), then reads and loads the same system image from the disk (412). The shared disk area (shape information) is continuously monitored (413), and a circular waiting operation is performed (414).

After the primary system saves the changed memory page to disk (403), it continues monitoring the state to store the changed memory page to disk (404) and monitors the state (405).

The primary system is monitoring the state (405), if a failure is detected (406), storing the state in the central processing unit (407), and after saving the changed memory page to disk (408), the need for system replacement. If the system replacement is not necessary (409), and the cause of the failure is repaired (410), the process of monitoring the state is repeated (405), and if a system replacement is required, the system is replaced by the sub-system after replacing the sub-system. It is determined whether there is a decision (415).

If it is determined that there is no system replacement decision, the process repeats the process of performing the waiting for a circular operation 414. If it is determined that the system replacement decision is made, the contents of the special area disk are copied to the memory (416).

There are two main types of events that transition from waiting to transition: asynchronous events such as power failures, manual resets, and synchronous events that cannot recover from a single fault, such as software resets and memory ECC errors. In the former case, the stored data can be used to switch to the sub-system with minimal loss. In the latter case, the sub-system can be switched without loss of information.

If the secondary system determines that there is a system replacement decision while performing a circular wait operation, it will copy the special area disk contents to memory (416), restore the state of the central processing unit (417), and then save the changed memory to disk. (418) The shape data is changed (419) before switching to the main system, and the process of repeating the step 405 of monitoring the state of the main system is repeated.

In the transition process, the contents of registers and caches used by the central processing unit to evacuate to the memory level are stored in the shared disk area (system sensitive information) so that the subsystem can maintain the state when it is switched. It is possible to recover the changed part completely by restoring it. In case of the secondary system, it examines the state of system switching and performs the related information recovery process in reverse order with the primary system, that is, switching from the backup swap partition to physical memory, so that the central processing unit can be restored to the previous state. Will work. The system that has failed displays the cause of the failure and when the recovery is completed, the preparation of the secondary system is performed.

The present invention described above is capable of various substitutions, modifications, and changes without departing from the spirit of the present invention for those skilled in the art to which the present invention pertains, and the above-described embodiments and accompanying It is not limited to the drawing.

As described above, the present invention implements the advantages of the fault tolerant system and the high availability system through a disk sharing system, thereby ensuring high reliability through a cyclic redundancy scheme in an open system composed of existing use elements without additional hardware development. It is possible to reduce the development cost accordingly, and to have flexibility in system application.

Claims (4)

In the cyclic redundancy method applied to the open embedded system, A first step in which the main / sub system configures cyclic redundancy to store shape data and memory contents and to monitor the shared disk area; A second step of executing / waiting if a system fails to recover from a single fault before detecting a critical error for transferring control to a circular standby system; And The third phase of copying the special area disk contents and switching the central processing unit into a circular system when a switch request event occurs. Circular redundancy method comprising a. The method of claim 1, The first step is, A fourth step of determining, when configuring a system, which one is to be the primary system operating or the secondary system in standby state; A fifth step of storing, in the case of the main system, shape data such as a shape of the main system, a size of a memory, a version of an operating system, a boot time, and a type of hardware; A sixth step of copying the entire contents of the physical memory to a backup swap partition which is a specific area of the disk in order to preserve the data of the current memory in the case of the main system; The sub system may include a seventh step of storing a shape of the main system, that is, a size of a memory, a version of an operating system, a boot time, a type of hardware, and the like; An eighth step of reading the same system image from the disk and loading the same as the main system; A ninth step of continuously monitoring the shared disk area (shape information); And Tenth step to perform circular wait operation Circular redundancy method comprising a. The method of claim 2, The second step, An eleventh step, wherein the main system monitors a state; If a failure is detected, storing the state in the CPU; A thirteenth step of storing the changed memory page on a disk; And Fourteenth Step to Identify Necessity of System Replacement Circular redundancy method comprising a. The method of claim 3, wherein The third step, A fifteenth step of copying the contents of the special area disk into the memory when the determination of the system replacement is performed while performing the circular wait operation as a result of the checking of the fourteenth step; Recovering the state of the central processing unit; A seventeenth step of storing the changed memory page on a disk; An eighteenth step of repeating the ninth step after changing the shape data; And If the system replacement decision is not confirmed as a result of the check in the 14th step, the 19th step is repeated from the ninth step after repairing a failure of a single point of failure. Circular redundancy method comprising a.