KR970007401B1 - Disorder management method of distributed processing system using event number - Google Patents

Abstract

The present invention relates to a failure management method for a distributed data processing system using event numbers by failure sources. This method includes a processor failure managing function collecting (31) failures, alarm messages, etc. from the source, and a system failure managing function. According to this invention, when a failure occurs, an event number (32) is allocated from the system failure managing funjction and each transition step of the sources is administrated to manage source failure, recovery request, current states, etc. as the same event number, thus easily maintaining, programming, and enhancing the system reliability.

Description

Fault management method using event number by fault resource in distributed processing system

1 is a structural diagram of an asynchronous transfer mode (ATM) exchange to which the present invention is applied.

2 is a diagram illustrating a path in which faults, faults, and alarm messages are collected according to the present invention.

3 is a process flow diagram in accordance with the present invention.

* Explanation of symbols for main parts of the drawings

A: CIM B: ASM

1: Access Switching Maintenance Processor (ASMP)

2: BCP (Brodcasting Cal1 Processor)

3: ClMP (Central Interconnection Maintenance Processor)

4: GSP (Globa1 Service Processor)

5: MMCP (Man Machine Control Processor)

6: Network Synchronization Control Processor (NSCP)

7: NTP (Number Translation Processor)

8: OMP (Operation and Maintenance Processor)

9: Remote Center Interface Processor (RCIP)

10: SCP (Subscriber Call Rrocessor)

11: SH (Signalling Handler)

12: TCP Call Processor

The present invention relates to a fault management method using an event number for each fault resource in a distributed processing system.

Generally, a distributed processing system such as an electronic exchanger is composed of tens or thousands of processors, and most processors except for a few processors are subscriber related processors.

Most of these subscriber-related processors have the same resources, and failures of these resources also have the same type of message and only the location of occurrence (processor) due to the failure of the same nature.

Therefore, after the failure of a specific processor, the operator must find all the printed recording papers to track the action and the current status of these failure resources, and thus, a large amount of time must be invested in finding the effective maintenance.

The present invention has been made to solve the above problems, by dynamically assigning event protection to the failure of a specific resource in a distributed processing system such as an electronic exchanger, the failure of the resource, self-test result measures and current situation It provides stable service of the system by providing algorithm that can efficiently manage the self-return deduction of fault resources by managing the Phtout Event Number of the output message for the same message. The purpose is to provide a fault management method using the event number for each fault resource.

In order to achieve the above object, the present invention, in the failure management method applied to the distributed processing system, the output message event number for the event received by determining an arbitrary natural number n according to the message output amount and period of the system from 0 to 10 Allocating up to ⁿ -1, and assigning the event number for the event of a specific resource received when the event number for which the event (disability, failure, alarm) information of a specific resource is not assigned is 0 to 10 ⁿ -1. Dynamically assigning one of the 0 to 10 ⁿ -1 and, if the event number is allotted from 0 to 10 ⁿ -1 after the event information of a specific resource is received, a new event number is set from 0 to 10 ^n- Allocating up to 1, notifying the operator of this, deleting the assigned case number when the action on the received case of a specific resource is completed, assigning a new case number, and then loading the system Checking and checking whether it is in the pass state, and outputting an unresolved fault in the case of a call, requesting the operator to take action, and requesting an audit of the remaining case number if not in a call. Characterized in that made.

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

1 is a structural diagram of an Asynchronous Transfer Mode (ATM) exchange to which the present invention is applied, in which A is CIM, B is ASM, 1 is ASMP, 2 is BCP, 3 is CIMP, 4 is GSP, and 5 is MMCP, 6 is NSCP, 7 is NTP, 8 is OMP, 9 is RCIP, 10 is SCP, 11 is SH, and 12 is TCP.

As shown in the figure, the structure of the asynchronous transfer mode (ATM) exchange is as follows.

The Access Switching Maintenance Processor (ASMP) (1) manages the control of maintenance tasks in the ASM (B), such as fault collection and testing of hardware resources in the ASM (B), and the BCP (Brodcasting Call Processor) (2). As a processor that manages multiple access (broadcasting) related service control, it is located in CIM (A) and ASM (B) and controls the overall multiple access function of CIM (A) and called party ASM (B) call processing.

The CIMP (Central Interconnection Maintenance Processor) (3) manages the control of maintenance tasks in the CIM (A), such as fault collection and testing of hardware resources in the CIM (A), and the Global Service Processor (GSP) (4). ) Controls system services provided to the system as a whole.

Man Machine Control Processor (MMCP) 5 allows for dialogue between the operator and the system or between the operating center and the system. For the dialogue between the operator and the system, VDUs (Visial Display Units) including the system console 17 and printers for hard copying output data are used. The local disk (loca1 disk) is used for the purpose of supplementing VDU transparency, command file, and logging file storage, and the Network Synchronization Control Processor (NSCP) (6) To control.

The NTP (Number Translation Processor) 7 responds to the number translation request from each SCP (10), TCP (Trunk Cal1 Processor) 12, and BCP (2), and ASM (B) number translation and each ASM ( It performs the translation of the link identification number (ID) accepted in B), and the Operation and Maintenance Processor (OMP) 8 oversees a series of operations and maintenance related functions in the system. Therefore, a magnetic tape (MT) 1 and a disk 16 are managed as auxiliary storage devices.

The MT 15 stores fee records, statistics, maintenance, operation management information, and the like, and the disc 16 stores general programs and data.

The Remote Center Interface Processor (RCIP) 9 controls communication with a remote operation center or a Telecommunication Management Network (TMN), and performs a protocol conversion function as necessary.

Subscriber Cal1 Processor (SCP) 10 is a processor that performs call processing of a general subscriber using the User Network Interface (UN) protocol. The subscriber Cal1 processor (UPC), priority parameter control (UPC), and priority control are performed together with a subscriber matching circuit. Overall traffic control such as congestion control, congestion control, and termination of signaling information cells on the Signaling Handler (SH) (11) sms UNI / NNI (Node Interfac by Netwo) protocol, SCP (10), and TCP (12). The black processes signal information together with the BCP 2.

TCP (12) is a processor that performs call processing with network using NNI protocol. It performs call / connection control on input / output call with relay line matching circuit and performs all functions necessary for matching with network. Preside over

In the past, the operator used to search the entire printer paper or use the history recording function to track the action and current status of these failure resources after the failure of a specific processor.

According to the method of tracking the printer paper, the output message has one output message description (ODM) file, so that the distinction between each OMD is distinguished by a unique integer value (output message reference number). It is a document that describes the output message to the operator in detail by adding detailed explanation and records the action conditions on the system.

In general, a series of flows of failures, actions, and current conditions of a particular resource can take as little as a few seconds to several hours, making it difficult to track the output messages for each step. Because printers are shared devices, there are tremendous output messages depending on the system situation, so it takes time and labor to trace a series of flows of specific resource failures, actions, and current conditions.

There is also a problem caused by the same output message reference numbers for the same kind of work. For example, if you look at the output message according to the start-up of a processor, one message is output according to the start-up of a specific file. The output message reference numbers for the output message are the same and only the processor positions between these messages are different. Processors can be started at the same time in the system, so tracking output messages related to a particular processor is equally time consuming and labor-intensive.

The following command uses the historical recording function: According to this command, when trying to refer back to the output message stored in the disk of MMCP (5), the message by year, month, day, start time, end time, output message reference number and specific input / output (1/0) terminal number See.

1.Format

DIS-MSG-HIS: TYPE = a [, YEAR = b] [, DATE = c] [, STM = d] [, ETM = e] [, PRN = f] [, PORT = g];

2. Description

2.1 Message Description

-Display the message details printed by I / O device and message type.

-The grade is grade 3 and there are SMA, NM, TRKT, SUBT group.

-This message is applied to the main body. The input devices that can be input are PC, CRT, and TTY.

2.2 Format Description

Variable argument default (minimum: maximum) enumlists

a ENUM 1--SYSM, ALM, FLT, STS, MMC

b year 1

c day 1

d time 1

e time 1

f SHORT 1 (0: 9999)

g SHORT 1 (0: 19)

Here, the variable a is the output message type, b is the year, and the default is the year (YYYY).

And, c represents the SMS Date (Date), the default means today (mmdd).

d represents the start time, and in the default case, means 1 hour before the current time (HHMMSS).

e indicates the finished time and, in the default case, the current time (HHMMSS).

f means all messages if no value is given as output message inquiry number.

g is an input / output device number. If a value is not given, g means an input / output device in which a corresponding command is input.

The relevant message is M0300.

The history recording function keeps the status messages, fault messages, alarm messages and MMC messages output during operation for several days. To output a specific message with this history recording function, the operator can output by status command, MMC output message, alarm message, and fault message.

In the past, most message types can be printed by the Printoutt Reference Number. However, due to the failure of the same nature, the message types are the same and only the location of occurrence (processor) is different. Therefore, the output messages related to a specific processor are tracked. Likewise, time consuming and manpower was inevitable.

The following is an output message according to the history recording command input. The printout reference number of this output message is 0300.

1.Format of M0300

<Display message history when M0300 is normal>

Type Year Sun STM ETM Port

××× ×××× ×××× ×××××× ×××××× ××

Result = normal

<Display message history when M0300 is abnormal>

Type Year Sun STM ETM Port

××× ×××× ×××× ×××××× ×××××× ××

Result = abnormal

Reason = b

2. Description

2.1 Message Description: Displays the message details printed by I / O device and message type.

2.2 Format Description

aaaa: The saved output message

Reason: If abnormal

Not data

Day of the week

Too late start time

The interval is out of range.

The relevant message is C0300.

2 is a path diagram in which faults, faults, and alarm messages are stored according to the present invention, wherein alarm messages generated by each processor in the ASM module are collected by the ASMP 1 and an ASMP (1) ) Is reported to the OMP (8) that manages system failures.

When failure information is received from the alarm source due to the occurrence of a failure, the change status is reflected in the status table, and it is determined whether the failure information is repeatedly output within a predetermined time, and the message output is suppressed in case of duplicate output within the predetermined time. If it is not duplicate, it should be printed.

3 is a process flow diagram according to the present invention.

First, it receives the failure, failure and alarm information of a specific resource (31). At this time, for the fault occurring within a certain time, the output message event number is dynamically generated for each fault resource so that the fault occurrence, self-test result, final value and current situation of the corresponding resource can be known.

When the message is output, a Printout Event Number is assigned to this event (32). The event number is assigned as an n-digit integer and consists of 0 to 10 ⁿ -1 events. If assigned, the operator is notified and a new case number is assigned from 0 to 10 ⁿ -1. If the case number is not assigned, the case number is increased by one (33, 34, 38). However, the integer n is a variable that depends on the amount and duration of the message output of the system.

When an event occurs, an event number is assigned, and the test status and action for this event are also outputted with the same event number, and when the error is recovered, the recovery message is finally extracted from the event list. .

When the case number is allocated and a new case number is assigned, the operator is notified (33, 34), and the current system load is checked (35) to check whether it is busy (36). Print out the unresolved fault, ask the operator for action and exit (37). At this time, considering the situation in which the case number remains even though the software (S / W) defect or the signal is recovered, the audit is requested for the remaining case protection when the call is not in progress and terminated (39). .

When information is received from the alert source, it must be able to confirm this status, run a test routine for fault detail, or transmit it as an audible and audible information to the operator, and transmit the status to the TMN.

0. Types of system messages are classified as follows.

MMC Message: The output of the command the operator entered to check the status of the system.

Status messages: Messages that inform the operator of the current status of a function while performing a function from each software (S / W) block.

Failure message: A message that informs the operator of the occurrence of a failure detected from each failure source or by a software (S / W) block that manages the failure source.

When a fault is detected, it prints out the possible fault location (line & channel, card, device & shell, rack, row).

The location of the row, rack, shelf, and card in each system does not have a symbolic name, but is in the form of a constant as shown below.

Processor (Equipment) Name Row Rack Shelf Card (Device)

××× 1 3 3 2

Detected faults are measured by the frequency of occurrence of each device (threshold value may vary depending on the characteristics of the equipment) and raised to an alarm message when the threshold is exceeded.

Alert message: An alert message is a message that includes audible and audible signals that indicates that the source of the fault must be acted upon immediately. The occurrence of an alarm message is generated when the status of a recurring error message is raised and may occur directly at the source of the error. The failure is divided into minor, major and critical according to the severity. The composition of the output message is as follows.

Exxxxxx Random Messages (××××)

+ +

│ │

Event number Output message reference number

Output message reference number The present invention as described above is composed of a processor failure management function for collecting failure, failure, and alarm messages generated from resources, and a system failure management function that is a central management module. By assigning a number to manage each transition stage of the resource and managing resource failures, self-recovery requests, and current situations with the same event number, system reliability improvement and implementation can be easily programmed by the operator's maintenance and developers. It has an easy effect.

Claims (1)

A failure management method applied to a distributed processing system, the method comprising: determining an arbitrary natural number n according to a message output amount and a period of a system and allocating an output message event number for a received event from 0 to 10 ⁿ -1; When the event (disability, failure, alarm) information is received, if the event number is not assigned from 0 to 10 ⁿ -1, the event number for the event of a specific resource is dynamically assigned to one of the above 0 to 10 ⁿ -1. the method comprising; specific received; the method comprising after the event information reception of the specific resource, the case number is given both in the 0 to 10 ⁿ -1 If the assignment in the new case number 0 to 10 ⁿ -1, and inform the operator Deleting the assigned case number when the action on the event of the resource is completed; allocating a new dictionary number, checking the load on the system to investigate whether the call is in a busy state; and Faults using the event number of each resource in the distributed processing system including the steps of outputting an unresolved fault and asking the operator to take an action, and requesting an audit for the remaining case number when not in a call. How to manage.