KR950013840B1 - Automatic servetance unit and operation method of overload of in service control - Google Patents

Abstract

an OA having a response time measuring unit for measuring a treatment time through a time of the input from that of the output of transaction of a first measuring unit for measuring the used amount of a CPU; a measuring unit for measuring the amount of the transaction that is not treated in the real time and is at the standby state; a CCP overload check reference value storing part for storing an overload check reference value under connection with the CCP overload watching and local check processing unit; a double-structured CCP having a CCP overload local controlling unit for controlling the overload corresponding to an overload control level of a corresponding processor under connection with the CCP overload watching and local check processing unit; a SLP overload watching and local check processing unit for collecting data from each measuring unit by being connected thereto; a SLP overload check reference value storing part for storing an overload check reference value under connection with the SLP overload watching and local check processing unit; a double-structured SLP having a SLP overload local controlling unit for controlling the overload corresponding to an overload control level of a corresponding processor under connection with the SLP overload watching and local check processing unit; an overload synthesis check processing unit for the overload degree of the all systems according to the local overload state of the processor; a system overload degree displaying unit for displaying the overload degree to an operator under connection with the overload synthesis check processing unit; an overload synthesis controlling unit for controlling the overload by being connected to the overload synthesis check processing unit; and a system process obstacle watching unit for watching the obstacle within the system.

Description

Distributed overload automatic monitoring device for intelligent network service control and its operation method

1 is a schematic diagram of an intelligent network service control transaction processing system to which the present invention is applied.

2 is a detailed block diagram of a transaction processing system for intelligent network service control according to the present invention.

3 is an overall flowchart of the overload monitoring and local determination processing method according to the present invention.

4 is a flow chart of the overload comprehensive determination processor according to the present invention.

* Explanation of symbols for main parts of the drawings

11, 21: response time meter 12, 22, 51, 61: CPU usage meter

13, 23, 52, 62: Transaction Wait Meter

14, 24: CCP overload local controller

15, 25: CCP Overload Monitoring and Local Decision Processor

16, 26: CCP overload judgment reference value store

53, 63: SLP Overload Monitoring and Local Decision Processor

54, 64: SLP overload local controller

55, 65: SLP Overload Judgment Store

71: each processor failure monitor 72: overload comprehensive judgment processor

73: system overload degree display device 74: overload integrated controller

The present invention relates to a distributed overload automatic monitoring device for intelligent network service control and an operation method thereof.

The intelligent network service control processes the service call (input transaction) originating from the service execution exchange and sends the result (output transaction) back to the service execution exchange. When the excess of the call exceeds the limit of the processing system, Overloading and unexpected abnormal operation can lead to inadequate service control. Therefore, overload management should be carried out to prevent such a phenomenon and protect the system and improve the reliability of the system. However, in the conventional overload state determination processing method for determining such an overload state, a processor for determining the overload state of the system is centrally located in the OAP (Operation and Administration Processor) to determine the overload state uniformly. Because of the intensive processing structure, it is difficult to proactively cope with changes in the system, and it is difficult to determine an appropriate overload judgment reference value for various characteristics of each processor by making the same criteria for all processors.

In order to solve the above problems, the present invention has a local determination processor for determining each processor overload state so as to automatically detect the overload. It is an object of the present invention to provide a distributed overload automatic monitoring apparatus and method for intelligent network service control that can notify and process a processor.

In order to achieve the above object, the present invention provides an intelligent network service control transaction processing system comprising: response time measuring means for measuring whether a processing time is measured from an input time of a transaction and a CPU time measuring means for measuring the usage of each processor; Transaction latency measurement means for measuring the amount of transactions waiting in real time on the flow of transactions in each processor; CCP overload monitoring and local judgment connected to each measurement means to collect data from each measurement instrument; A CCP overload determination reference value storage for storing an overload determination reference value connected to a processing means, the CCP overload monitoring and local determination processing means, and connected to the CCP overload monitoring and local determination processing means according to the overload control level of the processor. Overload control CCP overload local control means, which have the same structure as the first and second CPUs, which have a redundant structure, CPU usage measuring means for measuring processor usage, and cannot process in real time in the flow of transactions in each processor. Transaction latency measurement means for measuring the amount of transactions waiting, SLP overload monitoring and local determination processing means connected to each measurement means for collecting data from each measurement means, and the SLP overload monitoring and local determination processing means An SLP overload determination reference value storage for storing an overload determination reference value, and an SLP overload local control means connected to the SLP overload monitoring and local determination processing means to perform a corresponding overload control according to the overload control level of the processor. Redundant as the same structure Overload comprehensive determination processing means connected to the first and second SLPs and the overload monitoring and local determination processing means to determine the degree of overload of the entire system according to the local overload state trend of each processor in case of overload. A system overload degree display device connected to the overload comprehensive judgment processing unit for displaying by the operator so as to be recognized by the overload comprehensive control unit, an overload comprehensive control unit connected to the overload comprehensive judgment processing unit to control and reduce the overload; And an OAP connected to the processing means and having a system processor failure monitoring means for monitoring a failure of each processor in the system.

In addition, the method of operating the distributed overload automatic monitoring device for intelligent network service control includes the first step of determining whether each transaction processor (CCP, SLP) is operated when the overload monitoring starts, and after performing the first step, CCP1 / 2 or If the SLP1 / 2 fails simultaneously, after overload monitoring and local determination processor stop and the system stops and terminates, after performing the first step, each processor operates the CCP to drive the data in the CCP. After the third step of performing the third step, and after performing the third step to collect the measured value of the CCP and the input reference value of the overload judgment reference repository and the logic to table the collected data, the two tables are mapped and compared The local overload level is determined and the overload data and the local overload level are transmitted to the overload comprehensive judgment processor. After performing step 4 and requesting to start local overload control, it is determined whether the previous overload condition has been performed when the non-overload was performed. If not, the fifth step of stopping and returning each CCP / SLP measuring instrument in the periodic monitoring time band; after performing the first step, in case of SLP, the measuring instrument in the SLP is driven to generate data, and the measured value and overload of the SLP is After performing the sixth step of collecting the input reference value of the judgment reference storage and comparing the logical data with the sixth step, after performing the sixth step, the two tables are mapped and compared to determine the local overload level in case of overload. The local overload level is transmitted to the overload integrated judgment processor and simultaneously to the local overload controller. After performing the seventh step and the seventh step of requesting to start the local overload control, it is determined whether the overload state was previously performed when the non-overload was performed. And an eighth step of stopping and returning each CCP / SLP meter in a periodic monitoring time band.

Hereinafter, with reference to the accompanying drawings will be described an embodiment of the present invention;

1 is a schematic diagram of an intelligent network service control transaction processing system to which the present invention is applied, in which, 10 and 20 are Common Channel Signaling Processor (CCP), 30 and 40 are Local Area Network (LAN) devices, and 50 and 60 are SLP ( Service Logic Processor) and 70 represent an Operation and Adiministration Processor (OAP), respectively.

The transaction processing system for intelligent network service control according to the present invention is a system that controls various communication network services by providing an appropriate response when a query is input from a service exchanger, as shown in the drawing. The transaction is inputted to the transaction through the I / O link device connected to the transaction. The transaction is in the form of a signal protocol message, so that the CCP (10, 20), the front end processor, performs the interconnection control and transaction handler processing in the protocol layer processing. SLP 50, 60, which is a later processor, processes the dialogue handler on the protocol through LAN devices 30, 40, and services per service existing in the memory database of SLP 50, 60 for service-specific responses. The result is then returned in reverse order to generate a response by the logic data. The. Here, the system has duplicated each processing unit in pursuit of high performance and high reliability, and under normal circumstances, the processing of transactions is balanced between mutual loads in the front, back, front and back processors (10, 20, 50, 60). If an element is abnormal, all load sharing is performed by the normal processor. Here, the transaction input amount to this system can be used regardless of the processing capacity of the system due to the nature of the user's unlimited network.

For this reason, if a large amount of transaction input occurs, the system may be unable to handle its original functions due to the overload of the system, and the system may cause obstacles. In particular, since this system has a distributed processing structure, the characteristics of the overload are different according to the front and rear processors 10, 20, 50, and 60, which are distributed devices, and the criteria for determining the overload state may be set differently. Based on the overload condition, the entire system should be judged comprehensively to monitor overload at the entire system level. To this end, the overload comprehensive determination device is configured in the OAP 70 of the operation management of the system, and a device for performing automatic monitoring of the overload of the system is assigned to each distributed device to monitor the degree of overload of the system.

2 is a detailed block diagram of an intelligent network service control transaction processing system according to the present invention, in which 11, 21 are response time meters, 12, 22, 51, and 61 are CPU usage meters, and 13, 23, 52, and 62 are Transaction Latency Meter, 14, 24 are CCP Overload Local Controllers, 15, 25 are CCP Overload Monitoring and Local Decision Processors, 16 and 26 are CCP Overload Determination Reference Stores, 53, 63 are SLP Overload Monitoring and Local Decision Processors, 54, 64 denotes an SLP overload local controller, 55 and 65 represent an SLP overload judgment reference value store, 71 represents a system fault handler, 72 represents an overload comprehensive judgment processor, 73 represents a system overload degree indicating device, and 74 represents an overload integrated controller.

As shown in the figure, the transaction processing system for intelligent network service control measures the usage time of each processor and the response time measuring device 11 which measures whether the processing time is from the input time to the output time of the transaction in the CCP 10. CPU usage meter 12, a transaction latency meter 13 for measuring the amount of transactions waiting to be processed in real time on the flow of transactions in each processor, each connected to each of the meter (11 to 13) CCP overload determination reference value storage 16 for storing an overload determination reference value connected to a CCP overload monitoring and local determination processor 15 for collecting data from a meter, the CCP overload monitoring and local determination processor 15, and the CCP Connected to the overload monitoring and local determination processor 15, the corresponding widow according to the overload control level of the processor The CCP overload local controller 14 is configured to take control, and the CCP 20 has a duplex structure configured in the same manner as the CCP 10, and the SLP 50 is a CPU usage meter for measuring the usage of each processor ( 54) is connected to each of the transaction latency measuring device 52 and each of the measuring devices 51 and 52 to measure the amount of transactions waiting in real time on the flow of transactions in each processor. SLP overload determination reference value storage 55 for storing the overload determination reference value connected to the collected SLP overload monitoring and local determination processor 53, the SLP overload monitoring and local determination processor 53, and the SLP overload monitoring and local determination It is provided with an SLP overload local controller 54 that is connected to the processor 53 and measures to perform the corresponding overload control according to the overload control level of the processor. In addition, the SLP 60 has a redundant structure configured in the same manner as the SLP 50, and according to the local overload state transition of each processor when the overload monitoring and local determination processor 15, 25, 53, and 63 are overloaded. An overload comprehensive determination processor 72 for determining the degree of overload of the entire system, a system overload degree display device 73 for displaying to be recognized by an operator so as to be connected to the overload comprehensive determination processor 72, and the overload comprehensive determination processor And a system processor failure monitor 71 connected to the overload comprehensive controller 74 for reducing the overload, and connected to the overload comprehensive determination processor 72 to monitor the failure of each processor in the system. .

3 is a flowchart illustrating a method for monitoring an overload and a method for determining a local overload level for each processor as a general flow chart for each overload monitoring and local determination processing method according to the present invention.

First, when overload monitoring starts, it is determined whether each transaction processor (CCP, SLP) is operating (301). If CCP1 / 2 or SLP1 / 2 is simultaneously failed due to the characteristics of the system, transaction processing is not performed. The processor 15, 25, 53, 63 stops and the system stops and ends (304). In this case, the system is restarted and returned to its original state. As a result of the determination 301, when each processor operates as a CCP-SLP, it executes transaction processing to generate data by driving the measuring devices 11 to 13 in the CCP (302). In the case of SLP, the measuring devices 51 and 52 in the SLP are driven to generate data (303).

Then, the measured values of the CCPs are collected and the collected data is tabulated (305).

Subsequently, the input reference values of the overload judgment reference stores 16 and 26 are also tabulated in the same manner (306), and the two tables are mapped and compared (307).

As a result of the comparison (307), in case of overload, the local overload level is determined (308), and the overload data and the local overload level are transmitted to the overload comprehensive determination processor (309), and at the same time, the local overload controller is started to control the local overload. In case of overloading, 310 is notified that the overload is resolved since the overload is resolved. When the next monitoring time band is reached by stopping and returning each of the CCP1 and 2 measuring devices in the band, the CCP overload monitoring is repeatedly performed according to the above procedure to proceed continuously.

In addition, the measured value of the SLP is collected and the collected data is tabulated (312).

Then, the input reference values of the overload judgment reference store are also tabulated in the same manner (313) and the two tables are mapped and compared (314).

As a result of the comparison (314), when the overload is determined (315), the local overload level is determined (315), and the overload data and the local overload level are transmitted to the overload comprehensive determination processor (319). In case of non-overloading, it is checked whether it was in overload before (317), and if it is in case of overload before, it is notified that the overload is resolved. When the next monitoring time band is reached by stopping the measuring instrument and returning 318, the SLP overload monitoring is repeatedly performed according to the above procedure to continuously execute the overload monitoring by the CCP / SLP.

4 is a flowchart of the overload comprehensive determination processing method according to the present invention.

Each failure status of the transaction processing system for service control is first reviewed (401), and the overload monitoring and local decision processor results of each processor are received only when the CCP / SLP pair is used (402 to 405), and the CCP / SLP pair is returned. If a simultaneous failure occurs, the entire system is stopped (406). In addition, all the centrally-collected overload monitoring and local determination processor (15, 25, 53, 63) form a table that can be compared and mapped the results (407), from the table to predict the overload trend of the system to determine the emergency processing processor (408), and determine the overload overall level to control (409). The final determined result is displayed on the system overload display device 73 and transmits a drive signal to the overload integrated controller 74 to transmit the overload level and the main overload position information simultaneously to determine the overload of the entire system (410).

Therefore, the present invention can be applied to a transaction processing of a large amount and irregular input of a transaction input device and to an application system in which a processing structure is composed of a loosely coupled structure in pairs of front and rear separation, and particularly, for service control in a communication network. have. By applying this, even if the operator does not recognize the complex internal state of the system, the system can be automatically notified whether the system is overloaded or not, and it is composed of the overload controller to prevent the abnormality of the system due to the overload in advance. In addition, this method can be applied to the engineering of the facility by applying the data to the analysis to reinforce the fault handling capability of the treatment system.

Claims (3)

Response time measuring means (11) for measuring whether the processing time from the input time of the transaction to measure the processing time, CPU usage measuring means (12) for measuring the usage of each processor, in real time on the flow of transactions in each processor Transaction latency measuring means 13 for measuring the amount of transactions waiting to be processed and CCP overload monitoring and local judgment processing means connected to the respective measuring means 11 to 13 to collect data from each measuring instrument. (15), the CCP overload determination reference value storage 16 for storing the overload determination reference value connected to the CCP overload monitoring and local determination processing means 15, and the CPP overload monitoring and local determination processing means 15 And a CCP overload local control means 14 for taking control of the overload control according to the overload control level of the processor. CPPs 20 and 30, which have a redundant structure as the same structure, CPU usage measuring means 54 for measuring processor usage, and transactions waiting in real time on the flow of transactions in each processor, Transaction latency measurement means 52 for measurement, SLP overload monitoring and local determination processing means 53 connected to the respective measurement means 51 and 52 to collect data from each measurement means, and the SLP overload monitoring and SLP overload determination reference value storage 55 for storing the overload determination reference value, connected to the local determination processing means 53, and connected to the SLP overload monitoring and local determination processing means 53 according to the overload control level of the corresponding processor. SLP (50, 60) having the same structure and redundant SLP overload local control means (54) to take overload control, Overload comprehensive determination processing means 72 for determining the degree of overload of the entire system according to the local overload state trend of each processor in case of overload connected to the load monitoring and local determination processing means 15 and 53 as a LAN device. A system overload degree display device 73 for displaying the operator so that it can be recognized by the decision processing means 72, and an overload synthesis for reducing and controlling the overload connected to the overload comprehensive determination processing means 72 Intelligent network characterized in that it comprises an OA (70) having a system processor failure monitoring means (71) connected to the control means (74), said overload comprehensive determination processing means (72) for monitoring the failure of each processor in the system. Distributed overload automatic monitoring device for service control. Response Time Meters (11, 21), CPU Usage Meters (12, 22, 51, 61), Transaction Latency Meters (13, 23, 52, 62), CCP Overload Local Controllers (14, 24), CPP Overload Monitoring and Local judgment processor 15, 25, CPP overload judgment reference value storage 16, 26, SLP overload monitoring and local judgment processor 53, 63, SLP overload local controller 54, 64, SLP overload judgment reference value store 55 , 65), operation of distributed overload automatic monitoring device for intelligent network service control having a system each processor failure monitor 71, a comprehensive overload determination processor 72, a system overload degree indicating device 73, and an overload integrated controller 74; In the method; When the overload monitoring starts, after performing the first step 301 for determining whether each transaction processor (CCP, SLP) is operating or after performing the first step 301, if the CCP1 / 2 or SLP1 / 2 is simultaneously interrupted, the overload monitoring And stopping each local decision processor 15, 25, 53, and 63 and causing the system to stop and terminate, after performing the second step 304 and the first step 301, when each processor operates as a CCP. Third steps (305, 306) of generating the data by driving the measuring devices (11 to 13, 21 to 23), collecting the measured values of the CCP and the input reference values of the overload judgment reference storage and tabulating the collected data with comparative logic. After performing the third step (305, 306), the two tables are mapped and compared to determine the local overload level at the time of overload, and transfers the overload data and the local overload level to the overload comprehensive decision processor, and at the same time, the local overload controller Delivering to After performing the fourth step (307 to 309, 319) and the fourth step (307 to 309, 319) for requesting to start local overload control, it is determined whether the overload state was previously overloaded (310). In this case, since the overload is eliminated, a fifth step (310, 311, 319) and the first step (301) of determining and returning each CCP / SLP meter in the periodic monitoring time band are notified to the comprehensive judgment device. After performing SLP, in the case of SLP, it operates the measuring devices 51, 52, 25, and 26 in SLP to generate data, collects the measured values of SLP and the input reference value of overload judgment reference storage, and compares the collected data with a table. After performing the sixth steps 312 and 313 and the sixth steps 312 and 313, the two tables are mapped and compared to determine the local overload level during overload, and to overload the overload data and the local overload level. After the seventh step (314 to 316, 319) and the seventh step (314 to 316, 319) requesting to transmit to the local processor and the local overload controller to start the local overload control, The eighth step (317 to 319) to stop and return each CCP / SLP measuring instrument in the periodic monitoring time zone is notified to the comprehensive judge if the previous overload is resolved. Method for operating a distributed overload automatic monitoring device for intelligent network sub time control, characterized in that consisting of. The method according to claim 2, further comprising: in the overload comprehensive judgment processing method; Step 401 of reviewing each failure state of the transaction processing system for service control, after performing step 401, receiving the overload monitoring and local decision processor results of each processor only in the case of CCP / SLP pairs, and receiving CCP / When the simultaneous failure occurs without the SLP pair, the entire system stops (402 to 406), and after performing the steps (402 to 406), a table for comparing and mapping the centrally collected overload monitoring and local decision processor results is provided. Determining the emergency processing processor by predicting the overload trend of the system in the table (step 407 and 408), and after performing the steps 407 and 408, the overload comprehensive level to be controlled is determined and a drive signal is sent to the overload comprehensive controller. Determining the overload of the entire system by simultaneously transmitting the overload level and the main overload location information (409, 410). How the operation of automatic monitoring system for controlling intelligent network services in a distributed overload.