KR20190108712A - Method for recovering error in non-volatile main memory unit and system therefor - Google Patents

Abstract

Disclosed are an error recovery method in a nonvolatile main memory device environment and a system thereof. According to an embodiment of the present invention, the process error recovery method in a nonvolatile main memory device comprises: a step of storing execution processes of processes running in the nonvolatile main memory device in history storage corresponding to the processes; a step of receiving error information from an operating system if an error occurs in at least one process among the processes; and a step of using the history storage to recover from the error which has occurred in the at least one process corresponding to the received error information. The step of using the history storage to recover from the error analyzes a correlation between the at least one process in which the error has occurred and the remaining processes, and uses the analyzed correlation and the history storage to recover from the error which has occurred in the at least one process and errors of other processes associated with the at least one process.

Description

Method for recovering error in non-volatile main memory unit and system therefor}

The present invention relates to a technique for recovering an error in a nonvolatile main memory, and more particularly, to a method and system for determining an error occurring during an operation of a process in a system having a nonvolatile main memory as a memory and recovering the error. It is about.

In a main memory-secondary storage environment, a process recovery task is usually a temporary file of the currently running data, restarted at abnormal termination, and then read back from the beginning, or a program environment that has been operating normally before. The recovery routine is executed as soon as it is saved in a temporary file and restarted. It is operated by terminating and starting the process based on the volatility of the existing memory, such as overwriting it.

However, in the nonvolatile main memory, since there is no state of executing or terminating a process, the current recovery routine is not applicable to the nonvolatile main memory.

Therefore, there is a need for a technology capable of recovering an error when an error occurs in the nonvolatile main memory.

Embodiments of the present invention provide a method and system for determining an error occurring in a process operation and recovering the error in a system having a nonvolatile main memory as a memory.

The process error recovery method according to an embodiment of the present invention is a process error recovery method in a nonvolatile main memory device, the execution process of each of the processes operating in the nonvolatile main memory device to the history storage corresponding to each of the processes; Storing; Receiving error information from an operating system when an error occurs in at least one of the processes; And recovering an error occurring in the at least one process corresponding to the received error information by using the history store.

The recovering by using the history repository may include analyzing an association between at least one process in which the error occurs and the rest of the process, and using the analyzed association and the history repository, an error occurring in the at least one process and the It is possible to recover from errors in other processes associated with at least one process.

The recovering using the history storage may recover an error occurring in the at least one process by restoring to a value before a time point at which the error occurred in the at least one process using the history storage.

The storing in the history store may store a stack change process of each of the processes in a history store corresponding to each of the processes.

The storing of the history storage may include storing the history storage corresponding to each of the processes in a predetermined space of the nonvolatile main memory device or in the space of each of the processes.

The process error recovery system according to an embodiment of the present invention is a process error recovery system in a nonvolatile main memory, wherein the execution process of each of the processes operating in the nonvolatile main memory is stored in a history store corresponding to each of the processes. A storage unit for storing; A receiver configured to receive error information from an operating system when an error occurs in at least one of the processes; And a recovery unit for recovering an error generated in the at least one process corresponding to the received error information by using the history storage.

The recovery unit analyzes an association relationship between the at least one process in which the error occurs and the rest of the processes, and uses the analyzed association relationship and the history repository to determine an error occurring in the at least one process and another associated with the at least one process. Recover from errors in the process.

The recovery unit may recover an error occurring in the at least one process by restoring to a value before a time point at which the error occurred in the at least one process using the history storage.

The storage unit may store a stack change process of each of the processes in a history store corresponding to each of the processes.

The storage unit may share and store history storage corresponding to each of the processes in a predetermined space of the nonvolatile main memory device or in a space of each of the processes.

According to embodiments of the present invention, when a system software problem occurs in an environment in which a nonvolatile memory is used as a main memory device, the system software error is recovered by using process stack information, thereby recovering an error in the nonvolatile main memory device. Can be recovered.

In a situation where nonvolatile memory serves as both a main memory and a secondary memory, the process of running and exiting the memory is obscured.

In the present invention, by using a history structure that can log process execution in such a situation, the process can be safely restored to a previous state, and the history structure can be used as various caching data.

The present invention can be applied to a system in which the main memory is a nonvolatile storage medium and a system requiring error detection and recovery in an environment in which the main memory is a nonvolatile storage medium.

1 is a flowchart illustrating a process error recovery method in a nonvolatile main memory according to an embodiment of the present invention.
2 illustrates an example diagram for describing a method of operating a history repository.
FIG. 3 illustrates an exemplary diagram for describing a method of operating the history repository of FIG. 2A.
4 is an exemplary diagram for explaining a method of operating the history storage of FIG. 2B.
FIG. 5 shows an operation flowchart of an embodiment of step S130 shown in FIG. 1.
6 illustrates a configuration of a process error recovery system in a nonvolatile main memory according to an embodiment of the present invention.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited or limited by the embodiments. Also, like reference numerals in the drawings denote like elements.

In an environment to which the present invention can be applied, volatile is replaced with nonvolatile memory, and this nonvolatile memory also plays a role of HDDs and SSDs used as auxiliary storage devices. This nonvolatile memory is a memory that can retain previous data even when the power supply is interrupted. The nonvolatile memory refers to all non-volatile storage spaces such as PCM, Xpoint, and STT-MRAM.

In the case of DRAM, which is frequently used as a memory, data stored in the memory is lost when the power supply is interrupted, and data is read again from the auxiliary memory device at the start of the new memory. Some non-volatile memory may provide a function of temporarily initializing the memory to interact with the auxiliary memory. In the limited environment of the present invention, data is not deleted even when the power supply is interrupted. It is an environment that resides and is maintained in memory.

In the environment based on the present invention, the CPU registers move and all applications are executed simply by reading data in the memory. Volatile Memory-A nonvolatile secondary memory memory architecture handles the process by taking the process out of memory and reloading it in the event of an error. When the computer is not turned off, when it is reloaded, the recovery process can be performed by remembering the information of the existing process stored in the temporary file when it is running, or by restarting the system after initializing the memory by shutting down the system. . When the nonvolatile memory replaces the volatile memory, all the operations of loading and linking data from the auxiliary memory to the executable file format are lost, and the error part is also stored in the storage space. Recovery is not possible and if you reload, you will face the same situation as before.

Embodiments of the present invention utilizes process stack information to recover system software errors when a system software or system process problem occurs in an environment in which non-volatile memory (NVM) is used as the main memory. In the case of an error in the volatile main memory, the error can be recovered.

Here, the present invention uses a history store that can log the process execution in such a situation, not only can you safely recover the error process to the previous state, but also can use the history store as various caching data. .

The history store in the present invention may include all data structures and spaces capable of storing various data types such as structures, tables, records, and the like.

The present invention can be limited to the case where a process recovery is required in an environment in which the existing volatile memory is changed to a nonvolatile memory, a function of receiving an error message (or error information) detected from an operating system, and checking the received error message. This can include the ability to run with appropriate routines.

In order to implement this function, the present invention has a storage which can be called history, each process running in the nonvolatile main memory can be connected with the history storage, the history storage is the previous value of the data changed as each process is executed. You will save all of them.

That is, from the moment the process first occupies the CPU, the stack change process is recorded in the history store. Here, the present invention can collect data limited to stack values because the overhead of storing all the data of the process is large.

According to the present invention, when an error occurs in a process based on the collected data history log, the process may be restored to a previous state in which the process normally operates.

Furthermore, in the present invention, since an error of a process occurs and the error may affect a function of another process, the analysis may be performed based on information about analyzing the association between processes or pre-stored association between processes. In addition to the process in which the error occurred, errors occurring in other processes that are related by the generated error can also be recovered.

The process in the present invention may be a process including a thread.

Such a method and system according to the present invention will be described with reference to FIGS.

1 is a flowchart illustrating a process error recovery method in a nonvolatile main memory according to an embodiment of the present invention.

Referring to FIG. 1, in the process error recovery method according to an embodiment of the present invention, the execution process of each of the processes executed in the nonvolatile main memory device is stored in the history storage (S110).

Here, step S110 may store a stack change process of each of the processes in a history store corresponding to each of the processes.

The environment in which the nonvolatile memory of the present invention becomes a main memory device can be applied to various computing systems. For example, it can be applied to numerous environments that are executed through CPU operations such as smartphones, tablets, laptops, and PCs. Of course, the environment proposed in the present invention may include an environment in which the nonvolatile memory has enough performance improvement to have an input / output operations per second (IOPS) approaching the minimum DRAM and the storage space also achieves an SSD specification. In other words, since there are no IO operations occurring in SSDs, the environment running on enhanced nonvolatile memory runs faster and reduces unnecessary overhead of the system.

If this is the first boot in this environment, the first init process used will be the target of history logging. There are two ways to operate a stack in a non-volatile memory environment: 1) how all processes share one space in non-volatile memory, and 2) separate stack space for each process. . That is, the history store in the present invention may be stored by sharing a predetermined space of the nonvolatile main memory device as shown in FIG. 2A, or may be stored in each space of the processes as shown in FIG. 2B.

For example, when a shared stack is used for the history storage, as shown in FIG. 3, the data of each stack is stored with a data type for identifying a process. When POPs are removed from the stack and deleted from memory, there is no way to retrieve the data again, so you should always be careful with delete operations on systems that only store data in memory. Therefore, when a POP occurs, the state of the stack at that time is newly stored in the history store. This shared stack area can be used as a storage method if the storage method is known to be sequential, such as a history list.

At this time, in case of data written in the history, it is necessary to have information about the process to which the data belongs so that it is possible to know which stack is added to the process. It is preferable.

On the other hand, when the history store is stored in the space of each process, as shown in FIG. 4, the history of each process is stored one by one.

Here, the storage method is the same as when using a common stack, but several histories may be operated at the same time. Of course, the stack change value stored in the history storage may be stored using a storage method known as sequential, and all storage methods known as sequential may be used.

Referring back to FIG. 1, when an error occurs in at least one of the processes with the stack change process being stored in the history store corresponding to each of the processes by the execution process of each process by step S110. An error generated by the operating system is detected, and an error message or error information on the detected error is received from the operating system (S120).

Here, the operating system may know what error occurred in the running process, and the description thereof will be omitted since it is obvious to those skilled in the art.

When error information generated in at least one processor is received from the operating system in operation S120, an error generated in at least one process corresponding to the received error information is recovered using data stored in the history storage (S130).

Here, step S130 may recover an error occurring in at least one process by restoring to a value before the time at which the error occurred in at least one process using data stored in the history store.

Furthermore, in step S130, an error of a process occurs, and the error may be affected by a function of another process, so that the correlation between processes may be analyzed or based on information previously stored in the association between processes. In addition to the process in which the error occurred, errors occurring in other processes that are related by the generated error can also be recovered.

For example, step S130 analyzes an association relationship between at least one process in which an error occurs and the remaining processes as shown in FIG. 5 (S510).

Here, the step S510 may store in advance information on the relationship between the functions of each of the processes, the value that is changed by the process execution process and the resulting value for the function of another process is changed by the value so changed The information can be known by analysis. Of course, the association analysis for each of the processes may apply not only the above description but also all the methods for analyzing the functional relationships.

If another process related to the error-producing process is analyzed in step S510, the error recovery process recovers not only an error of the error-producing process but also an error of another process associated therewith, thereby causing a process in the nonvolatile memory. It is possible to reliably recover the errors that occur.

In addition, although the method according to the present invention is not shown in FIG. 1, if an error occurred in the process is not solved by the recovery to the previous value stored in the history store, the recovery using the operating system is executed or the user is customized. By performing an error recovery operation, the error may be recovered.

As described above, the method according to the present invention recovers a system software error by using process stack information when a system software problem occurs in an environment in which a nonvolatile memory is used as a main memory. You can recover from the error.

In addition, the method according to the present invention uses a history structure that can log the process execution process, it is possible not only to safely recover the process to the previous state, but also to generate a process error using the association information between processes. In this case, it is possible to stably recover not only an error of the corresponding process but also an error of another process caused by this.

6 illustrates a configuration of a process error recovery system in a nonvolatile main memory according to an embodiment of the present invention, and illustrates a configuration of a system that performs the method of FIGS. 1 to 5.

Referring to FIG. 6, the system 600 according to the present invention includes a storage unit 610, a receiver 620, and a recovery unit 630.

The storage unit 610 stores the execution process of each process executed in the nonvolatile main memory in the history storage.

Here, the storage unit 610 may store a stack change process of each of the processes in a history store corresponding to each of the processes, and each of the processes by using a method in which all processes share and use a space of the nonvolatile memory. A history corresponding to each of the processes may be stored in a stack space of each of the processes.

When an error occurs in at least one of the processes and the error is detected by the operating system, the receiver 620 receives an error message or error information on the detected error from the operating system.

When the error information generated in at least one processor is received from the operating system, the recovery unit 630 recovers an error generated in at least one process corresponding to the received error information by using data stored in the history store.

Here, the recovery unit 630 may recover an error occurring in at least one process by restoring to a value before a time point of occurrence of an error in at least one process using data stored in the history store.

In addition, the recovery unit 630 may cause an error in a process, and the generated error may affect a function of another process. Based on this, not only the process in which the error occurred, but also errors occurring in other processes that are related by the generated error, can be recovered together.

Although the description is omitted in the system of FIG. 6, it will be apparent to those skilled in the art that the system according to the present invention may include all the contents described in the method of FIGS. 1 to 5.

The system or apparatus described above may be implemented with hardware components, software components, and / or combinations of hardware components and software components. For example, the systems, devices, and components described in the embodiments may include, for example, processors, controllers, arithmetic logic units (ALUs), digital signal processors, microcomputers, field programmable arrays (FPAs). ), A programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions, may be implemented using one or more general purpose or special purpose computers. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to the execution of the software. For convenience of explanation, one processing device may be described as being used, but one of ordinary skill in the art will appreciate that the processing device includes a plurality of processing elements and / or a plurality of types of processing elements. It can be seen that it may include. For example, the processing device may include a plurality of processors or one processor and one controller. In addition, other processing configurations are possible, such as parallel processors.

The software may include a computer program, code, instructions, or a combination of one or more of the above, and configure the processing device to operate as desired, or process it independently or collectively. You can command the device. Software and / or data may be any type of machine, component, physical device, virtual equipment, computer storage medium or device in order to be interpreted by or to provide instructions or data to the processing device. Or may be permanently or temporarily embodied in a signal wave to be transmitted. The software may be distributed over networked computer systems so that they may be stored or executed in a distributed manner. Software and data may be stored on one or more computer readable recording media.

The method according to the embodiments may be embodied in the form of program instructions that may be executed by various computer means and recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the media may be those specially designed and constructed for the purposes of the embodiments, or they may be of the kind well-known and available to those having skill in the computer software arts. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic disks, such as floppy disks. Magneto-optical media, and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

Although the embodiments have been described by the limited embodiments and the drawings as described above, various modifications and variations are possible to those skilled in the art from the above description. For example, the described techniques may be performed in a different order than the described method, and / or components of the described systems, structures, devices, circuits, etc. may be combined or combined in a different form than the described method, or other components Or even if replaced or substituted by equivalents, an appropriate result can be achieved.

Therefore, other implementations, other embodiments, and equivalents to the claims are within the scope of the claims that follow.

Claims (10)

A method for recovering a process error in a nonvolatile main memory,
Storing a process of executing each of the processes running in the nonvolatile main memory in a history store corresponding to each of the processes;
Receiving error information from an operating system when an error occurs in at least one of the processes; And
Recovering an error occurring in the at least one process corresponding to the received error information using the history store;
Process error recovery method comprising a.
The method of claim 1,
Restoring using the history store
Analyze an association between the at least one process in which the error occurred and the rest of the processes, and use the analyzed association and the history repository to generate errors in the at least one process and errors in other processes associated with the at least one process. Process error recovery method characterized in that for recovering.
The method of claim 1,
Restoring using the history store
Recovering an error occurring in the at least one process by restoring to a value before an error occurs in the at least one process using the history storage.
The method of claim 1,
The storing in the history repository is
Storing a stack change process of each of the processes in a history store corresponding to each of the processes.
The method of claim 1,
The storing in the history repository is
And storing the history storage corresponding to each of the processes in a predetermined space of the nonvolatile main memory device or in the space of each of the processes.
In a process error recovery system in a nonvolatile main memory,
A storage unit to store an execution process of each of the processes operating in the nonvolatile main memory in a history storage corresponding to each of the processes;
A receiver configured to receive error information from an operating system when an error occurs in at least one of the processes; And
A recovery unit for recovering an error occurring in the at least one process corresponding to the received error information by using the history storage
Process error recovery system comprising a.
The method of claim 6,
The recovery unit
Analyze an association between the at least one process in which the error occurred and the rest of the processes, and use the analyzed association and the history repository to generate errors in the at least one process and errors in other processes associated with the at least one process. Process error recovery system, characterized in that for recovering.
The method of claim 6,
The recovery unit
And recovering the error occurring in the at least one process by restoring to the value before the time at which the error occurred in the at least one process using the history storage.
The method of claim 6,
The storage unit
Storing a stack change process of each of the processes in a history store corresponding to each of the processes.
The method of claim 6,
The storage unit
And storing the history storage corresponding to each of the processes in a predetermined space of the nonvolatile main memory device or in the space of each of the processes.

Images