KR950011480B1 - An electronic exchahger - Google Patents

Abstract

receiving process trouble generation information from an operation system and reading execution state information of a corresponding process in a database to request restarting to a starting management function unit; if no response exists or the response fails, requesting reloading to a corresponding execution module; collecting executing process information from the operation system, reading information of the number which is generated in maximum, and comparing the maximum number with the number of the executed process to determinate existence/nonexistence of overgeneration of process; and requesting process generation limit to the operation system and outputting a state message.

Description

Process performance monitoring and fault recovery method of electronic switching system

1 is a control system hardware structure diagram of a TDX-10 system.

2 is a process configuration diagram of execution modules mounted in a processor.

3 is a process failure detection and regeneration process.

4 is a process chart for process overproduction monitoring.

5 is a method for monitoring through the process of confirming the creation status of the process.

The present invention relates to a method for monitoring the execution status of a process, which is the minimum performance unit for each function of the electronic switchboard, and recovering from the failure.

In the distributed hardware structure of the electronic switching system, it can be divided into upper and lower layers according to the functional characteristics, and the processor controlling the hierarchical functions is divided into the main processor (MP) and the peripheral processor (PP). Among them, the functional performance of the main processor (MP) is performed by a mutual exchange of information between a plurality of associated execution modules that are implemented for each processor. Therefore, any process failure (destruction) belonging to one execution module of these related functions immediately leads to a large failure state of the entire function execution.

The present invention has been made to solve the above problems, by managing the state of the process that must be continuously created to perform the main task for each execution module to regenerate when these processors are destroyed due to the failure caused by the failure Process execution status to minimize the impact and ensure service continuity, and at the same time, to prevent the generation of a process for the main function due to the over-generation of the process of a specific function through the process monitoring function. Its purpose is to provide a monitoring and disaster recovery method.

In order to achieve the above object, the present invention, in order to determine whether or not the normal state of the process execution state is divided into the form that the process form is temporarily generated and extinguished (extinotion) when the execution is completed and continuously performed, The failure information of the process is received from the operating system to confirm that the execution form of the process is continuous, and then re-created only. In addition, since the number of processors that can be created in a specific processor is not allowed more than a certain number due to the limitation of memory capacity, the maximum limit that can be created for all processes is set and the process information that is actually created and executed is displayed. The generation should be limited to over-produced processes that are collected.

FIG. 1 shows a hardware structure distributed in functional units of an electronic switching system. The functional type includes a subscriber service module (ASS-S: Access Switching Subsystem-Subscriber; 1-1) and a trunk line sibling module. Trunk; 1-2), Packet Access Module (Access Switching Subsystem-Packet; 1-3), Interconnection Network Subsystem (INS: 1-4) to control number translation and switch connection, and functional hierarchy It consists of a central control module (CCS: Central Control Subsystem; 1-5) that manages overall operation and maintenance functions such as collecting and analyzing failure and status information for each subsystem and outputting an output message. Interprocessor communication (IPC) is used to exchange information for function performance.

In addition, the control system includes an operation maintenance processor (OMP), an input / output processor (MMP), and an access exchange processor (ASP), which are main processors (MPs) implemented in resource information management, operator message processing, database management, and call processing control. , Number translation processor (NTP), network processor (INP), and periphral process (PP), a periphral process that directly controls devices in hardware, ASIP, digital trunk line matching processor (DTIP), and space switch Processor (SSP), Packet Layer Control Processor (PLCP) has a hierarchical structure. In the present invention, only the functions performed in the MP are applied, and the present invention is applied to the execution of all general programs controlling the operating system and the database manager.

FIG. 2 conceptually shows a process mounted in each of the main processors that envision the control system. In one processor, a plurality of execution modules are mounted and each execution module is configured as one or more processes. .

Here, the execution module can be performed in units of software functional units. As shown in the figure, the execution modules, that is, blocks A, B, and C, are mounted in the OMP 2-1, and the execution module A is A1, A2, A3, A4 It consists of four processes, and execution module B consists of three processes, B1, B2, and B3. Execution module C means one process of C1. Similarly, blocks D, E, F, and G are mounted in ASP 2-2, H, I, and J are mounted in NTP 2-3, and each block performs one or more processes to perform a function.

In order to achieve a single function, information exchange between processes is performed using an interprocessor communication (IPC), and an external processor signal between the B3 process of B block mounted in OMP (2-1) and the D2 process of D block mounted in ASP is used. (2-4) and the internal signals (2-1, 2-2, 2-3) between the B2 process of the B block in the OMP 2-1 and the C1 process of the C block in the processor.

The information shown in [Table 1] below defines the elements that make up a process in the form of a database, and D_PDF_IND (3-1) of the process definition number attribute for identifying each processor internally, D_PRIORITY (3-2), an attribute that indicates priority of importance, and D_EXE_TYPE (3-3), an attribute that distinguishes between maintaining the tree type of the process after creation and destroying it when it is finished. It consists of D_MAX_CNO (3-4), an attribute representing the maximum number of processes created, and D_EM_NAME (3-5), an attribute representing the name of the process. The above information is used to monitor and recover from process failures and overproduction.

TABLE 1

3 is an overall processing flow diagram in accordance with the present invention.

Performance information and failure occurrence information of general processing functions (4-1, 4-3) are received by the process monitoring and recovery function (4-5) through the operating system (4-2), in the performance monitoring function database (Acquisition of the definition number (3-1) of the process through the management function (4-4) (3-3) and the maximum number of generation (3-4), execution of the process described in [Table-1] For the process that needs to be carried out continuously by extracting the information such as the module name (3-5), the regeneration of over-produced process and the action of suppression or retry of the over-produced process are taken through the start management (4-6) function. In addition, such state occurrence information is composed of a fault and a state message and is output through the operator registration 4-7.

4 is a process flow chart for determining and regenerating a process that requires regeneration in function performance when a process failure occurs.

First, the definition inentifer of the failed processor is received from the operating system (5-1) and the characteristics of the process are analyzed using the stored information in the database (5-2). If it is a process that needs to be performed continuously, the processor starts-up management (5-4) and requests regeneration (5-5) and then waits for a certain time using a timer (5-5). If not, terminate execution.

When a normal processing completion message is connected from the start management function (5-12), a process restart completion message is output (5-13) and terminated.

If there is no response from the start management function for a certain time (5-7), the number of start request is searched (5-8), if it is three times or less, the number of requests is increased and re-request is performed (5-11). After setting the count counter to an initial state (5-9), a restart failure message is output (5-10), and a reloading request is requested (5-14) to the execution module.

In order to prevent the generation of a process that performs an important function due to the overproduction of a specific function, FIG. 5 collects the entire information of the periodically generated process and monitors the process of confirming the creation status of each individual process. It shows how.

The monitoring function performed periodically is driven at regular intervals using the timer provided by the operating system (6-1). The overall process creation information consists of definition number and number of generations and is obtained using the operating system (6- 2).

Using the process definition number (6-3), the maximum number of information that can be created by the process is read from the database (6-4) and compared with the number actually generated (6-5). Do it (6-9), otherwise ask the operating system for a production limit for the process (6-6), print an overproduction status message (6-7), and then confirm that it is complete (6-8). Repeat the same process for the next process.

Therefore, the present invention, which is performed as described above, minimizes the service failure time by regenerating the corresponding process by internally determining the system at the time of failure of the process, and has a priority in consideration of the importance of each function under the limited control resources. By preventing the performance failure of important functions due to the excessive generation of low-performance processes, it is possible to guarantee stable service throughout the system, thereby improving the quality of service (QOS).

Claims (2)

In the method of monitoring the execution status of a process applied to an electronic switching system and recovering from a failure, the process failure information is received from the operating system and the execution status information of the corresponding process is read from a database to determine whether the process is continuously performed. After the first steps (5-1 to 5-4) and the first steps (5-1 to 5-4) to request a restart by the start management function, if there is no response from the start management function or schedule Retry the number of times and if failure continues, the second step (5-5 to 5-14) requesting the reloading of the execution module, and periodically the information of the currently created and running process from the operating system Reads the maximum number of information generated in the database using the process definition number and compares it with the actual number of running processes. After performing the third step (6-1 to 6-5) and the third step (6-1 to 6-5), only the fructose-generated process requests the process creation restriction to the operating system and sends a status message. And a fourth step (6-6 to 6-8) for outputting and terminating the execution state of the processor. 2. The execution state monitoring and fault recovery of a processor according to claim 1, wherein the retrying to the start management function in the second steps 5-5 to 5-14 is performed three times. Way.