KR20010057827A - Method of Managing System Errors in the Network System - Google Patents

Abstract

The present invention relates to a system fault management method in a network system, and particularly stores a trap and a log file generated in an NVRAM in the event of a system failure in NVRAM. The present invention relates to a system error management method in a network system to maintain a description.

The method of the present invention includes the steps of creating a session by checking the session creation request information according to the session request in RMON and starting a timer; Collecting and updating related information while periodically checking and checking a timer identifier and whether a predetermined time has elapsed; Generating an alarm / event by checking whether the update value is equal to or greater than a preset threshold value; And generating a trap or log file, checking attributes of the trap identifier and the event identifier, and determining whether to store the trap or log file according to the identified attributes.

Description

How to manage system errors in network systems {Method of Managing System Errors in the Network System}

The present invention relates to a method for managing a system error in a network system. In particular, a trap (alert message) and a log file generated in the event of a system failure are stored in NVRAM to maintain a description. A system error management method in a network system.

In general, the RMON Management Information Base (RFC) of Request for Comment (RFC) 1757 has a Packet / Byte, Broadcast / Multicast number generated within one segment. Statistics Group for the data, etc., History Group that provides information on various traffic and errors within the time interval set by the Manager, and specific values periodically By checking and reaching the threshold (threshold), the manager manages the alarm group that the agent maintains the record, and the number of traffic and errors generated by each device connected to the segment by host. Host Top N Group, which finds the hosts that generated the most traffic in a certain period of time among the hosts generated from the host group, Collects information about traffic and errors between two hosts based on the data-link layer, or MAC (Media Access Control) address, to find out who has the most traffic and errors. The Matrix Group to see, the Manager monitors the trend of a specific packet, and the Filter Group to notify if the filtered packet occurs more than a certain threshold value, and the Manager collects the packets generated in the segment. It includes a packet capture group that allows analysis and an event group that keeps a record of certain events as they occur and sends traps to the manager.

Here, when an abnormal event occurs in the system, the RMON group generates a trap to the manager so that the manager can take action. In particular, the filter group, the host top-end group, the event group, The alarm group generates a trap for the manager.

In this method, it is possible to distinguish whether to generate only a trap, to store only a log file in the system, and to simultaneously generate a trap and a log file.

In addition, in order to support the RMON MIB, the system registers a timer that periodically collects MIB information, and creates a session to generate an entry for each group recommended by the RFC. (Creation)

Therefore, the generated session collects information for each given time-out value, and when a certain threshold is reached, a trap is generated in the NMS using SNMP or a description of the situation occurred in the log file of the system. Describe it.

If the NMS wants to view the log file, it can be viewed using the SNMP, but if the system is rebooted due to a critical fault, the information on the log file will be cleared. And the reason for the trap is unknown.

Then, in the network system of the prior art, looking at the configuration for system error management with alarms and events for the error packets (Error Packets), as shown in Figure 1, the SNMP 11, the NMS 12 and And an RMON processing module 13, an error processing module 14, and a device 15.

Looking at the system error management method of the configuration as described above are as follows.

First, the SNMP 11 generates a session for collecting error packets of a statistic group according to the recommendation of the RFC. In order to generate the session through the RMON processing module 13, the corresponding session number and port number ( Port number) and session owner are created, and a timer is generated to collect and update information by periodically checking the timeout at the end of the creation request.

Once a session of statistics has been created, create an alarm and event session in the same way as described above. In the case of the alarm, the difference from the statistics requires additional information on when the alarm is generated.

That is, if more than 10,000 error packets are generated for one port of the system, when the alarm can be generated, the corresponding value should be recorded in the session entry when an alarm session is created. The interval value for whether the timer collects information of the error packet, the sample type value that distinguishes whether the error packet value is absolute or delta, and the error packet value Values indicating whether to generate an alarm when reaching '10, 000 'while increasing from minimum value to maximum value between minimum value' 0 'to maximum value' 0xFFFFFFFF 'supported by this system, or both In other words, the values of Falling Threshold, Rising Threshold, Polling and Rising Threshold should be recorded in the session entry.

When the value of '10, 000 'is reached for the error packet by the recorded values, the system generates an event to generate the trap or log file to the error manager of the system.

Accordingly, the event session generated for the generated event records a description in a log table of the system, and transmits a trap to the NMS 12 through the SNMP 11. In addition, the manager solves the error by diagnosing the terminal and the system connected to the system.

By the way, in the implementation of the RMON, when generating a trap or log file to the NMS (12) can be generated only as recommended in the RFC, the system reset due to a serious error for the generated (System Reset) ) Stores past messages and log files in random access memory (RAM), making it difficult to maintain them continuously.

In addition, it is inconvenient to continuously monitor the system in order to determine the cause of the error that is not recommended. Errors other than the relevant RFC recommendation include device critical errors, memory allocation errors, and abnormal processing tresses in each processing module if each processing module fails to allocate resources for operation. (Trace) may correspond to this.

In addition, since RMON only sees the time when it occurs due to an operation tick, there is a disadvantage that the date and time of error occurrence cannot be seen.

As such, in the case of a system reset due to a critical error, since traps and log files are stored in RAM, it is difficult to maintain them continuously, and it is inconvenient to continuously monitor the system in order to determine the cause of an error which is not recommended. There was this.

In order to solve the problems described above, the present invention, by using NVRAM to maintain the log file in the system error and alarm / event of RMON after system reset to maximize the tolerance (error) of the error (network system error) It is suitable for management, and traps and log files generated in the event of a system failure are stored in NVRAM to maintain a description so that the manager can accurately determine the system's error and overcome it. The purpose of this is to provide accurate information by dividing the date and time using (Real Time Clock).

1 is a block diagram showing a configuration for system fault management in a conventional network system.

2 is a block diagram illustrating a configuration for system error management in a network system according to an embodiment of the present invention.

3 is a diagram illustrating a system error management method in a network system according to an embodiment of the present invention.

Explanation of symbols on the main parts of the drawings

21: Simple Network Management Protocol

22: NMS (Network Management System)

23: RMON (Remote Monitoring) Processing Module

24: Fault Process Module

25: Devices

26: non-volatile RAM (NVRAM)

A system error management method in a network system according to an embodiment of the present invention for achieving the above object is a process of creating a session by checking session creation request information according to a session request in RMON and starting a timer at the same time. and; Collecting and updating related information while periodically checking and checking a timer identifier and whether a predetermined time has elapsed; Generating an alarm / event by checking whether the update value is equal to or greater than a preset threshold value; And generating a trap or log file, checking attributes of the trap identifier and the event identifier, and determining whether to store the trap or log file according to the identified attributes. Here, the trap includes a trap identifier and an event identifier, and the log file includes a table index, a descriptive word, a date, a time, and a module identifier, and the date and time are stored by the RTC.

In addition, the process of determining whether to store the trap or log file is characterized in that it comprises the step of storing the log file in the main memory to determine whether the identified attribute should be stored only in RAM. Alternatively, the process of determining whether to store the trap or log file includes determining whether the checked attribute should be stored only in NVRAM and storing the log file in the NVRAM. Alternatively, the step of determining whether to store the trap or the log file includes: generating a corresponding trap by checking whether the identified attribute should generate only a trap; And graphically displaying the trap on a screen when the trap is requested to be returned. Alternatively, the step of determining whether to store the trap or the log file comprises: identifying whether the identified attribute should be stored in NVRAM and generating a trap, storing the log file in the NVRAM, and simultaneously generating the trap; And graphically displaying the trap on a screen when the trap is requested to be returned.

Meanwhile, the method for managing a system error in a network system according to an embodiment of the present invention may further include generating a log file / trap having a module identifier, a trap identifier, and a trap level in case of a system error irrelevant to the RMON. It features.

The present invention utilizes NVRAM to provide fault processing mechanisms, to provide device and processing module error traces on the NMS, and to provide fault and event trap levels that take into account various system resources. Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

As a configuration for system error management in a network system according to an embodiment of the present invention, as shown in FIG. 2, an SNMP 21, an NMS 22, an RMON processing module 23, and an error processing module ( 24), a plurality of devices 25, and NVRAM 26, to generate and maintain the RMON and error traps or log files of the system.

The SNMP 21 maintains a MIB table that can be processed according to a request authorized from the NMS 22, and also has a trap identifier capable of processing each trap, having information for generating each group of RMONs, There is a module capable of generating UDP packets to the NMS 22 for NMS identifiers, generated traps.

The NMS 22 requests the SNMP 21 for statistical sessions, alarm / event sessions, and log file information.

The RMON processing module 23 is a module for periodically interfacing with the devices 25 for the RMON MIB recommended in the RFC and collecting information and maintaining the collected information. The RMON processing module 23 is a statistical group, a host group, an event / alarm. It maintains RMON entry tables such as groups and interfaces with timers to periodically update information. Here, the statistics group collects statistical information of each device 25 and maintains a session generation number, statistics MIB, creation index, owner, timer value, and the like. It maintains table entries such as session generation number, generation index, owner, trap identifier, timer value, etc. that generate traps to NMS 22, and the alarm group generates traps through events when certain thresholds are reached for any MIB. Let's do it. Also, an alarm state indicating whether the corresponding session generation number, generation index, owner, sample type to distinguish whether it is an absolute value or a change value, whether a specific value of the MIB occurs in Rising, Polling, or Rising and Polling. Alarm Status Values, and interval time indicating how many MIBs are collected.

The error processing module 24 is a module for processing traps or log files of the RMON, a trap identifier for identifying traps, a log table for storing log files, a module identifier for identifying each module of the system, It manages an interface with a real time processor that obtains a real date time (RDT). Here, the log table includes an index for identifying the table, a module identifier, a system date / time, and a description of the corresponding trap.

The NVRAM 26 is an area for storing real-time errors in the system. The NVRAM 26 may store a module identifier, an index, a descriptive value, a date, a time, and the like.

A system error management method in a network system according to an embodiment of the present invention is as follows.

First, the mechanism for generating traps and log files in the system and RMON will be described with reference to the flowchart of FIG. 3. The NMS 22 authorizes the SNMP 21 session request for a statistical group to be managed. In addition, in order to perform network management through traps and log files, an alarm / event session must be created. At this time, the NMS 22 grants an alarm / event session request to the corresponding SNMP 21 (step S1).

Accordingly, the SNMP 21 receives a session request of a statistics group or an alarm / event from the NMS 22 and then applies a request for creating a statistics session or an alarm / event session to the RMON processing module 23 (step). S2).

Then, the RMON processing module 23 receives a request for generating a statistical session or an alarm / event session from the SNMP 21 and checks whether the information of the corresponding generation request is correct (step S3). If the corresponding generation request information is not correct, the SNMP 21 is notified of the failed session creation (step S4), and the statistics session or alarm / event session is performed when the generation request information is correct in the third step (S3). In addition, each timer is designated and operated at the same time (step S5).

That is, in the case of generating a table of the corresponding statistics session, it is checked whether there is information such as a session number, a port number, a session owner, and the like in the generation request information and generates the corresponding statistics table by using the table of the corresponding alarm / event session In the case of, check whether there is information such as an interval time, a threshold value, an event identifier, a tap identifier, a variable of a sample type (Variables; Source, MIB), and a session number in the generation request information, and use the corresponding alarm / event table. Will be created.

Then, the timer of the statistic or alarm / event checks whether a predetermined time has elapsed (step S6), and also checks whether there is a statistic timer identifier or an alarm / event timer identifier (step S7).

When the session is normally generated as described above, the statistics MIB is periodically collected and the alarm / event group is also periodically monitored for the values according to the set values, and the information is collected and updated (step S8).

At this time, it is checked whether the collected values deviate from a preset threshold value, that is, the current MIB value is larger than the alarm value in the case of a rising threshold value (step S9), and collected in the ninth step S9. If the values deviate from the preset threshold, the RMON processing module 23 requests the error processing module 24 to perform alarm / event generation (step S10).

Then, the error processing module 24 receives an alarm / event generation request from the RMON processing module 23 to generate an alarm / event (step S11), and generates a trap or log file (step S12). . Here, the trap has a trap identifier and an event identifier, and the information of the table of the log includes a table index, a descriptive value, a date, a time, a module identifier, and the date and time are not managed correctly. Use an RTC that accurately represents the date and time, not the tick. At this time, it is determined whether to store the trap or log file, and the trap or log file is stored in the NVRAM 26 (step S13).

In other words, it checks whether the attribute of the trap identifier and the event identifier is a log file which should be stored only in RAM and stores the log file in main memory. Alternatively, it is determined whether the attributes of the trap identifier and the event identifier are log files that should be stored only in the NVRAM 26, and the log files are stored in the NVRAM 26. Alternatively, it is determined whether the attribute of the trap identifier and the event identifier should generate only a trap, and generates a corresponding trap to the SNMP 21, and the SNMP 21 corresponds to a request of the NMS 22. The trap is applied to the NMS 22 to allow the NMS 22 to be graphically displayed on the screen. Alternatively, it is checked whether the attribute of the trap identifier and the event identifier is a log file to be stored in the NVRAM 26 and a trap should be generated to store the log file in the NVRAM 26 and at the same time the SNMP A corresponding trap is generated to the 21, and the SNMP 21 applies the trap to the NMS 22 at the request of the NMS 22 so that the NMS 22 is displayed graphically on a screen. To give.

Through the above operation, a trap or log file is generated. The log file is stored in the NVRAM 26 and the system is reset by expressing the correct date and time using the RTC and determining whether to store the corresponding information. If this happens, you can keep the log file intact and find out the cause of the critical error.

Meanwhile, in another embodiment, even when a system error irrelevant to the RMON, that is, a reboot or a significant error occurs, the device 25 may transmit a module identifier or a trap to the error processing module 24 through the RMON processing module 23 in the device 25. After generating a log file / trap having an identifier and a trap level, the log file is stored in the NVRAM 26 in the same operation as described above and returned at any time when the NMS 22 requests it.

As described above, trap and log files are stored in NVRAM according to the present invention, so that even if a system reset occurs due to a critical error, the cause of the error can be known. And time can be distinguished accurately.

Claims (7)

Checking the session creation request information according to the session request in RMON, creating a session, and starting a timer; Collecting and updating related information while periodically checking and checking a timer identifier and whether a predetermined time has elapsed; Generating an alarm / event by checking whether the update value is equal to or greater than a preset threshold value; Create a trap or log file, and check the properties of the trap identifier and the event identifier, and determines whether to save the trap or log file according to the identified properties, the system error management in the network system Way. The method of claim 1, The trap includes a trap identifier and an event identifier, the log file includes a table index, descriptive, date, time and module identifier, the date and time are stored by the RTC in the network system How to manage system errors. The method of claim 1, The determining of whether to store the trap or log file comprises determining whether the checked attribute should be stored only in RAM and storing the log file in main memory. The method of claim 1, The determining of whether to store the trap or log file comprises determining whether the checked attribute should be stored only in NVRAM and storing the log file in the NVRAM. The method of claim 1, Determining whether to store the trap or log file comprises the steps of: identifying whether the identified attribute should generate only a trap and generating a corresponding trap; And graphically displaying the trap on a screen when the trap is requested to be returned. The method of claim 1, The determining of whether to store the trap or the log file includes: determining whether the identified attribute should be stored in NVRAM and generating a trap, storing the log file in the NVRAM, and simultaneously generating the trap; And graphically displaying the trap on a screen when the trap is requested to be returned. The method of claim 1, And generating a log file / trap having a module identifier, a trap identifier, and a trap level in the event of a system error irrelevant to the RMON.