JP6880961B2 - Information processing device and log recording method - Google Patents

Description

The present invention relates to an information processing device and a log recording method.

The Baseboard Management Controller (BMC) firmware installed in the computer monitors the hardware of the computer and records information about remote controls, hardware events, and so on. Standard specifications are specified in the Intelligent Platform Management Interface (IPMI) for logging.

As the device to be monitored becomes large-scale, the amount of logs to be acquired increases and it takes time for the user to analyze the logs, so that the efficiency of analysis is required.

As a related technique, a technique has been proposed in which only log information necessary for determining the content of a failure or the like is output to a file device when a failure occurs in an application program (see, for example, Patent Document 1).

Further, as a related technique, an information processing device that collects operation history information related to a plurality of tasks without performing exclusive control has been proposed (see, for example, Patent Document 2).
Information processing device that can

Japanese Unexamined Patent Publication No. 2007-58265 Japanese Unexamined Patent Publication No. 2010-66799

The BMC is, for example, a System Event Log (SEL) including a failure type specified in IPMI, a detailed log in which details such as application operation information at the time of a failure and information output from a hardware register are described. And record. The description content of SEL is specified in IPMI.

For example, it is conceivable to start recording the detailed log after giving the identification information to the SEL, and write a message and a recording time in the SEL after the recording of the detailed log is completed. However, the time required to acquire the detailed log differs depending on the event. Therefore, there is a possibility that the order of the SEL identification information and the order of the SEL recording time do not match.
As one aspect, the present invention aims to ensure the integrity of the recorded logs.

In one aspect, the information processing apparatus has an acquisition unit that acquires event data related to the event in response to detection of the event that has occurred, and a first unit that responds to the event after the acquisition of the event data is completed. 1 The relationship between the recording unit that sets the identification information and records the log corresponding to the event data in association with the first identification information, the event that has occurred, and the event indicated by the recorded log. Select whether to have the acquisition unit acquire the event data, have the acquisition unit acquire a part of the event data, or have the acquisition unit acquire all the event data. Includes a selection and.

According to one aspect, the integrity of the recorded logs can be ensured.

It is a figure which shows an example of the whole structure of the system of embodiment. It is a figure which shows the process outline of the BMC firmware. It is a figure which shows an example of SEL. It is a figure which shows an example of the log recording processing in which inconsistency occurs. It is a figure which shows an example of the software structure of BMC. It is a figure which shows an example of the management information of a log acquisition method. It is a figure which shows the error management information. It is a flowchart which shows an example of the process flow of Embodiment. It is a figure which shows an example of the process of step S101 of FIG. It is a figure which shows the 1st example of a log recording process. It is a figure which shows the 2nd example of a log recording process.

<Example of the overall configuration of the system of the embodiment>
Hereinafter, embodiments will be described with reference to the drawings. FIG. 1 shows an example of the overall configuration of the system of the embodiment. The system of the embodiment includes a first server 1 and a second server 2. The second server 2 may be one unit or a plurality of units. The first server and the second server can communicate with each other and send and receive IPMI commands to and from each other.

The first server 1 includes a BMC 11, a Power Supply Unit (PSU) 15, a sensor 16, and a CPU Memory Unit (CMU) 17. The first server 1 is an example of an information processing device.

The BMC 11 includes a storage device 12, a first communication interface 13a, a second communication interface 13b, a third communication interface 13c, and a processor 14. The CMU 17 includes a Central Processing Unit (CPU) 18 and a memory 19.

The BMC 11 is an on-chip management controller, which is mounted on, for example, a motherboard. The BMC 11 records in the storage device 12 various logs related to each hardware in the first server 1, various hardware included in the second server 2, and the like.

The storage device 12 is an example of a computer-readable tangible storage medium. These tangible storage media are not temporary media such as signal carriers. The storage device 12 is, for example, a portable recording medium (for example, an SD memory card) using a semiconductor memory. However, the storage device 12 is not limited to the portable type, and may be built in the BMC 11. A program for executing various processes in the BMC 11 may be stored in the storage device 12.

The first communication interface 13a is used for communication with the second server 2. The second communication interface 13b is used for communication with the PSU 15 and the CMU 17. The third communication interface 13c is an example of General Purpose Input / Output (GPIO) and is used for communication with the sensor 16.

The processor 14 executes the program deployed in the storage device 12. The PSU 15 supplies power to the first server 1. The PSU 15 communicates with the BMC 11 by an Inter-Integrated Circuit (I2C) communication method.

The sensor 16 performs various detections related to the first server 1. The number of sensors 16 may be one or a plurality. The sensor 16 is, for example, a water leakage detection sensor, a temperature sensor, a sensor for measuring the rotation speed of a cooling fan, or the like.

The CMU 17 is, for example, a board on which a CPU 18 and a memory 19 are mounted. The memory 19 is, for example, a Random Access Memory (RAM). The CPU 18 executes the program expanded in the memory 19.

In addition to the configuration shown in FIG. 1, the first server 1 may include an auxiliary storage device such as a hard disk drive or a semiconductor memory.

The BMC 11 acquires a log recording the power supply status from the PSU 15 by the I2C communication method. The power supply state is, for example, a voltage value. The file format of the log is, for example, a binary format, which can be referenced by using a tool.

Further, the BMC 11 is acquired from the PSU 15 via the second communication interface 13b by the I2C communication method as a log indicating the power supply state (for example, the voltage value).

Further, the BMC 11 acquires various detection data from the sensor 16 via the third communication interface 13c and records a log related to the various detection data. The detection data includes the presence or absence of water leakage, the temperature of the CPU 18 and the like, the rotation speed of the cooling fan, and the like. The file format of the log is, for example, a binary format, which can be referenced by using a tool. The time required to acquire the log from the sensor 16 is, for example, several milliseconds.

Further, the BMC 11 acquires the dump information of the CMU 17 as a log via the second communication interface 13b by the I2C communication method. The dump information is, for example, information output from a register held by the CPU 18 when an error occurs. The time required to acquire the log from the CMU 17 is likely to be longer than the log acquisition time from the PSU 15 and the sensor 16.

Further, the BMC 11 acquires a log from the second server 2 via the first communication interface 13a by using the IPMI command. The time required to acquire the log from the second server 2 is likely to be longer than the time required to acquire the log from the hardware in the first server 1, for example, several minutes.

The processor 14 outputs a log (hereinafter, referred to as a program operation log) indicating the operation of the program to be executed, and stores it in the storage device 12. The program operation log is updated regularly. The processor 14 outputs a program operation log when an event such as a hardware error occurs, for example. If a similar event has occurred before, the BMC 11 may copy the log of the event.

The program operation log is output periodically, for example, and it is highly possible that the time required for recording is shorter than the log acquired from the PSU 15, the sensor 16, the CMU 17, the second server 2, and the like. The file format of the program operation log is binary format or text format. The operation log in binary format can be referenced by using the tool.

The components in the second server 2 may be the same as or different from the first server 1. The first server 1 may be connected to an electronic device other than the server (for example, a storage, a switch, etc.), or may acquire a log related to the connected electronic device.

FIG. 2 is a diagram showing a process outline of the BMC firmware. The processing of BMC firmware is defined as, for example, the IPMI Main process.

The BMC firmware process includes an interrupt reception thread of a CPU (for example, CPU 18), an interrupt reception thread of a PSU (for example, CMU17), and a monitoring thread for values output from various sensors (for example, sensor 16). In each thread, logging is performed in the SEL registration library and the detailed log registration library.

The SEL registration library and the detailed log registration library are stored in, for example, the storage device 12. SEL is an example of a log. The detailed log is an example of event data.

BMC11 creates a thread for each hardware. Further, even if the hardware is the same, if the types of events are different, the BMC 11 may generate a new thread. Further, the BMC 11 analyzes the contents of an event such as a hardware error, and processes the SEL recording and the detailed log recording in each thread.

The BMC 11 performs SEL recording and detailed log recording in response to interrupts from, for example, the CPU 18 and the PSU 15. For example, the BMC 11 may periodically monitor the CPU 18 and the PSU 15 to acquire a log, and perform SEL recording and detailed log recording. In addition, the BMC 11 periodically monitors various sensors, and when an abnormal value is detected, for example, SEL recording and detailed log recording are performed.

FIG. 3 is a diagram showing an example of SEL. SEL is a log showing an outline of an event, and the description content is defined by IPMI. SEL is used, for example, to isolate the occurrence and severity of a disorder. The SEL shown in FIG. 3 includes a SEL_ID, a time (timestamp), a message code (MsgCode), an error level (ErrLvl), a part name, and a message.

The SEL_ID, time, and message code are the contents specified by IPMI. For example, a server that has acquired a log from BMC11 can convert a message code using a tool to acquire an error level, a part name, and a message.

SEL_ID is log identification information given to each thread. The time is the time when the message was recorded. The message code is a code that separates events.

The error level is, for example, information indicating whether the error content is a normal level, a warning level, or a fatal level. The hardware type indicates the type of hardware that caused the event. The message content is a description of the content of the event, such as the error content.

The BMC 11 records one or more detailed logs (not shown) corresponding to the SEL shown in FIG. The detailed log is a log acquired from various hardware (for example, PSU15, sensor 16, CMU17, and second server 2). Alternatively, the BMC 11 may edit the acquired log into a predetermined format to make a detailed log.

Detailed logs are used to investigate the failure factors in detail. The detailed log is a log generated by Original Equipment Manufacturing (OEM), and the recording content and recording method are defined by the user.

In addition, the detailed log is used for analysis of detailed cause investigation of the failure. As the device to be monitored becomes larger, the number of detailed logs increases and it takes time to analyze. The BMC 11 assigns common identification information to the SEL and the detailed log, and manages the relevance by using the identification information to clarify the information necessary for analysis and improve efficiency. For example, when a user refers to a log to analyze an error, the error can be analyzed without referring to all the detailed logs by referring to the SEL and selecting the detailed log to be used for analysis based on the contents of the SEL. Can be analyzed.

For example, when the BMC 11 determines that an event similar to or related to a previously generated event has occurred, the BMC 11 may record only the identification information of the previously generated log as the detailed log, or omits the acquisition of the detailed log. You may. The BMC 11 can reduce the log capacity by recording only the identification information of the previously generated log or omitting the acquisition of the detailed log.

FIG. 4 is a diagram showing an example of log recording processing in which inconsistency occurs. In the example shown in FIG. 4, it is assumed that two events (errors) to be logged occur and processing is performed in thread A and thread B corresponding to the two errors, respectively.

In the SEL recording area shown in FIG. 4, a date (SEL recording date), a time (SEL recording time), and a message code are recorded in the SEL_ID. In FIG. 4, the date (X / X), the time (09: 00), and the message code (○○) are recorded in advance corresponding to the SEL_ID: “1”. Further, in the detailed log, a folder "ID0001" corresponding to SEL_ID: "1" is created in advance.

When the error A for which the log is to be acquired occurs (S11), the BMC 11 secures a free space in the SEL recording area for the error A and sets "2" as the SEL_ID (S12). Since the SEL recording date and time and the log contents are unknown at the time of acquiring the SEL_ID, the BMC 11 temporarily records the message "reserved A" in the SEL recording area instead of the date, time, and message code.

The BMC 11 starts acquiring the detailed log regarding the error A (S13). The BMC 11 creates a folder with a folder name corresponding to the SEL_ID when starting the acquisition of the detailed log. Then, the BMC 11 saves the acquired log in the created folder. In the example of FIG. 4, the BMC 11 has a folder name of "ID0002" corresponding to SEL_ID: "2".

It is assumed that an error B different from the error A occurs after S13 (S14). The BMC 11 secures a free space in the SEL recording area for error B, and sets "3" as the SEL_ID (S15). Since the SEL recording date and time and the log contents are unknown at the time of acquiring the SEL_ID, the BMC 11 temporarily records the message "reserved B" in the SEL recording area instead of the date, time, and message code.

The BMC 11 starts acquiring the detailed log regarding the error B (S16). The BMC 11 creates a folder named "ID0003" corresponding to SEL_ID: "3" when starting the acquisition of the detailed log. Then, the BMC 11 completes the acquisition of the detailed log (S17).

The BMC 11 performs SEL recording in the SEL recording area after saving the detailed log in thread B (S18). The BMC 11 records the date, time, and message code in the area where the SEL_ID is "3". The recorded date and time are the date and time when the SEL was recorded.

After S18, it is assumed that the acquisition of the detailed log is completed for error A (S19). The BMC 11 performs SEL recording in the SEL recording area after saving the detailed log in thread A (S20). The BMC 11 records the date, time, and message code in the area where the SEL_ID is "A".

In the above processing, the date, time, and message code corresponding to SEL_ID: "1" and "3" are recorded at the time of processing S18, but the date, time, and message corresponding to SEL_ID: "2" are recorded. No code is recorded. That is, a space is generated in the middle of the SEL recording area secured in the order of SEL_ID.

Further, as a result of the processing of S11 to S20, in the SEL recording area, the date and time corresponding to SEL_ID: "3" (x / x 10:40) from the date and time corresponding to SEL_ID: "2" (x / x 10:50). Is the previous date and time. That is, the order of SEL_ID and the order of date and time are reversed.

A state in which a space is created in the middle of the SEL recording area or a state in which the order of the SEL_ID and the order of the date and time are reversed is a state not expected by the program, and therefore, an abnormality may occur in various processes. For example, the IPMI command for reading the SEL sent from an external information processing device or the like may result in an error, or the acquisition of log data may partially fail. In addition, the log may not be restored when the server is restarted. Further, if there is a space in the middle of the SEL area, new logging may not be possible.

The format of the detailed log varies, and the collection time also varies. For example, as described above, it takes more time to collect detailed logs from other devices than to collect detailed logs (sensor information, etc.) from the hardware in the own device. Further, in a large-scale system, a plurality of failure events may occur at almost the same time, and a large number of logs may be recorded in a short period of time, which causes the above-mentioned problems.

Recording the SEL before recording the detailed log solves the above problem, but when determining the SEL to be recorded after analyzing the collected detailed log, it is preferable to record the SEL after recording the detailed log.

Moreover, since the SEL recording method is supported by the IPMI standard, the recording process is provided as one of the standard packages. Therefore, for example, it is difficult to devise such as changing the SEL_ID in accordance with the completion of the detailed log.

In addition, a method of preventing the above problem by exclusion between threads can be considered. For example, it is conceivable that the processing of thread B is not performed during the processing of thread A. However, if exclusion between threads is performed, the log recording time becomes long when a plurality of events occur in a short period of time.

<Example of BMC>
In order to avoid the above-mentioned problems, the BMC 11 in the embodiment performs a process for ensuring the consistency of the log recorded with respect to the event that has occurred. In the following description, the event that occurs is a hardware error, but the event that occurs is not limited to the error. For example, the event that occurs may be information that the program outputs periodically.

FIG. 5 is a diagram showing an example of the software configuration of BMC. The BMC 11 includes a communication unit 21, a storage unit 22, a selection unit 23, a recording unit 24, and an acquisition unit 25.

The communication unit 21 transmits / receives data to / from various hardware. The communication unit 21 corresponds to the first communication interface 13a, the second communication interface 13b, and the third communication interface 13c shown in FIG.

The storage unit 22 includes a logically divided SEL recording area and a detailed log recording area. Further, the SEL is stored in the SEL recording area, and the detailed log is stored in the detailed log recording area. In addition, the storage unit 22 stores management information of the log acquisition method and error management information, which will be described later. The storage unit 22 is realized by the storage device 12 shown in FIG.

The selection unit 23 does not allow the acquisition unit 25 to acquire the detailed log, or the acquisition unit 25 acquires a part of the detailed log, based on the relationship between the newly generated event and the event indicated by the recorded SEL. Select whether to have the acquisition unit 25 acquire all the detailed logs. Details of the processing of the selection unit 23 will be described later.

When an event to be acquired by the log occurs, the recording unit 24 sets the identification information corresponding to the thread number as the second identification information corresponding to the detailed log and assigns the identification information to the detailed log. Further, the recording unit 24 creates a folder in the detailed log recording area to which a folder name corresponding to the second identification information is assigned. Then, the recording unit 24 saves the acquired detailed log in the created folder.

When the recording of the detailed log is completed, the recording unit 24 sets the SEL_ID corresponding to the event that has occurred, and records the SEL in association with the SEL_ID and the recording time.

After recording the SEL, the recording unit 24 updates (renames) the folder name and the second identification information of the detailed log area from the thread number to the name based on the SEL_ID based on the SEL_ID. SEL_ID is an example of the first identification information.

Based on the selection result of the selection unit 23, the acquisition unit 25 acquires the detailed log related to the event that has occurred from various hardware for which the detailed log acquisition target is detected, in response to the detection of the event that has occurred. The various hardware for which the detailed log is acquired are, for example, the PSU 15, the sensor 16, the CMU 17, the second server 2, and the like shown in FIG. The acquisition unit 25 acquires a detailed log when the temperature acquired from the sensor 16 exceeds a predetermined value or when a predetermined request is received from the second server 2 or the like.

FIG. 6 is a diagram showing an example of management information of the log acquisition method. The selection unit 23 determines the type of error after the occurrence of the error, and determines whether or not a related error has occurred within a predetermined time in the SEL stored in the storage device 12 in the past. When a related error has occurred, the selection unit 23 checks the detailed log related to the error and estimates whether the newly recorded detailed log is the same as the detailed log stored in the storage device 12. If it is presumed that they are the same, the selection unit 23 causes the acquisition unit 25 to acquire only a part of the detailed log or does not acquire the detailed log. It is assumed that the detailed log of the acquisition target is set in advance for each type of error.

For example, when the selection unit 23 detects the same error as the error indicated by the SEL stored in the storage device 12, it selects not to acquire the detailed log. The selection unit 23 determines whether or not the error is the same, for example, based on the message code (MsgCode) shown in FIG. 3 and the content of the newly generated error.

When the selection unit 23 detects an error related to the error indicated by the SEL stored in the storage device 12, if it is detected in the same thread, one of the detailed logs of the acquisition target set for the error. Select to have the acquisition unit 25 acquire the detailed log of the unit. For example, when a related error is detected in the same thread, the program operation log is likely to have the same contents. Therefore, the selection unit 23 selects not to acquire the program operation log as a detailed log. For example, the selection unit 23 acquires the register information of the CPU 18 to the acquisition unit 25 because the detailed logs such as the register information of the CPU 18 may be different.

For example, when the selection unit 23 selects to have the acquisition unit 25 acquire a part of the detailed logs, if the error is an error related to the power supply, the selection unit 25 causes the acquisition unit 25 to acquire only the detailed logs from the PSU 15.

Further, when the selection unit 23 detects an error related to the error indicated by the SEL stored in the storage device 12, if it is detected by another thread, the details of all the acquisition targets set for the error Choose to get the log.

When related errors occur frequently, that is, when the time when a newly generated error occurs and the time when an error occurs recorded in the recorded SEL are within a predetermined range, the error that occurs later is preceded. The error that occurred is likely to be more important. This is because it is highly possible that it was caused by the cause of the later error caused by the earlier error. Therefore, the selection unit 23 selects not to acquire the detailed log if the related error occurs frequently and is detected in the same thread. Further, the selection unit 23 selects to acquire a part of the detailed log set for the error if the related error occurs frequently and is detected by a different thread.

As described above, the selection unit 23 can reduce the log capacity to be stored and shorten the log acquisition time by selecting to omit the acquisition of the detailed log estimated to have the same contents.

FIG. 7 is a diagram showing error management information. The management information shown in FIG. 7 is stored in the storage unit 22. The “message” shown in FIG. 7 is an error message based on information acquired by the BMC 11 from various hardware when an error occurs. The "error type" is the type of error corresponding to the "message". Further, the detailed logs 1 to 4 are detailed logs acquired by the acquisition unit 25 from the PSU 15, the sensor 16, the CMU 17, and the second server 2 shown in FIG. 1, for example.

The "error type" shown in FIG. 7 is used when the selection unit 23 determines whether the newly generated error is related to or is the same as the error that occurred in the past.

In FIG. 7, “◯” indicates that the detailed log is the acquisition target. “Δ” indicates that the detailed log is the acquisition target but can be omitted. “All detailed logs” in FIG. 6 are detailed logs marked with “◯” and “Δ” in FIG. 7. Further, the “partial detailed log” in FIG. 6 is a detailed log marked with “◯” in FIG. 7.

In the example shown in FIG. 7, the detailed log of the acquisition target is recorded for each message, but for example, the detailed log of the acquisition target may be recorded for each MsgCode shown in FIG.

<Example of processing flow of the embodiment>
FIG. 8 is a flowchart showing an example of the processing flow of the embodiment. The process shown in FIG. 4 starts when the BMC 11 detects a predetermined event such as a hardware error.

Based on the relationship between the event that occurred and the event indicated by the recorded SEL, the selection unit 23 does not allow the acquisition unit 25 to acquire the detailed log, or causes the acquisition unit 25 to acquire a part of the detailed log, or all the details. Select whether to acquire the log (step S101). The selection unit 23 makes a selection based on, for example, the management information shown in FIGS. 6 and 7.

When it is selected in step S101 to acquire a part or all of the detailed log (YES in step S101), the recording unit 24 acquires the thread number (step S103).

The recording unit 24 determines whether or not a folder corresponding to the event that has occurred is created in the detailed log recording area of the storage unit 22 (step S104). If NO in step S104, the recording unit 24 sets the identification information corresponding to the thread number in the detailed log, and sets the folder with the folder name corresponding to the identification information in the detailed log recording area of the storage unit 22. Create (step S105).

If YES in step S104, or after the processing in step S105, the acquisition unit 25 acquires detailed logs from various hardware (step S106). The recording unit 24 records the acquired detailed log (step S107). The recording unit 24 stores, for example, the acquired detailed log in the created folder. The recording unit 24 may convert the acquired detailed log into a predetermined format and then save it in the folder.

If it is selected in step S101 to acquire a part of the detailed log, in step S105, the acquisition unit 25 acquires a part of the detailed log of the acquisition target. The acquisition unit 25 acquires only the detailed log marked with “◯” in the management information of FIG. 7.

The recording unit 24 determines whether or not the recording of the detailed log has been completed (step S108). If NO in step S108, the recording unit 24 returns to the process of step S104. If YES in step S108, the recording unit 24 sets the SEL_ID corresponding to the error that has occurred, and records the SEL in association with the SEL_ID and the recording time (step S109). The recording unit 24 records, for example, a SEL including a SEL_ID, a date, a time, and a message code in the SEL recording area.

The recording unit 24 acquires the recorded SEL_ID from the SEL recording area (step S110). The recording unit 24 updates (renams) the second identification information associated with the detailed log and the folder name of the detailed log recording area based on the acquired SEL_ID (step S110). That is, the recording unit 24 changes the second identification information and the folder name associated with the detailed log from the thread number to the name corresponding to SEL_ID.

The recording unit 24 can streamline the log analysis performed later by changing the identification information and the folder name associated with the detailed log from the thread number to the name corresponding to the SEL_ID.

FIG. 9 is a diagram showing an example of the process of step S101 of FIG. The selection unit 23 determines the error type based on the information transmitted from various hardware (step S201). The selection unit 23 may determine the error type based on the management information shown in FIG. 7.

The selection unit 23 determines the relationship between the newly generated error and the error that has occurred in the past based on the SEL stored in the storage device 12 and the detailed log (step S202). The selection unit 23 determines whether or not there is a relationship based on, for example, the newly generated error content and the recorded SEL message code. In addition, the selection unit 23 may select a detailed log relating to an error that may be related, and may determine the relationship between the newly generated error and the error that has occurred in the past by referring to the detailed log.

The selection unit 23 is based on the management information (for example, FIG. 6) stored in the storage device 12 when the newly generated error is the same error as the error that occurred in the past or an error related to the error (YES in step S203). , Select the log acquisition method (step S204). The selection unit 23 selects whether to have the acquisition unit 25 acquire the detailed log, the acquisition unit 25 to acquire a part of the detailed log, or the acquisition unit 25 to acquire all the detailed logs.

If NO in step S203, the selection unit 23 selects to acquire all detailed logs (step S205).

As described above, since the BMC 11 of the embodiment sets the SEL_ID and records the SEL including the SEL_ID after the acquisition of the detailed log is completed, it is possible to prevent the order of the SEL_ID and the SEL recording time from being reversed. it can. Further, the BMC 11 of the embodiment can prevent the occurrence of a state in which a space is formed in the middle of the SEL recording area. That is, the BMC 11 of the embodiment can ensure the consistency of the recorded log. Therefore, the BMC 11 of the embodiment can restore the log without leaving an extra log even if the device is restarted during the log recording, for example.

<First example of log recording processing>
FIG. 10 shows a first example of the log recording process. In the example shown in FIG. 10, it is assumed that two events (errors) to be logged occur and the threads A and B corresponding to the two errors are processed.

In the SEL recording area shown in FIG. 10, a date (SEL recording date), a time (SEL recording time), and a message code are recorded in association with the SEL_ID. In FIG. 10, a date (X / X), a time (09: 00), and a message (○○) are recorded in advance for SEL_ID: “1”. Further, in the detailed log recording area, a folder "ID0001" corresponding to SEL_ID: "1" is created in advance.

When the error A for which the log is to be acquired occurs (S21), the recording unit 24 creates a folder (TMP2222) corresponding to the thread number "2222" in the detailed log recording area. Then, the acquisition unit 25 starts acquiring the detailed log regarding the error A (S22). The recording unit 24 saves the detailed log acquired for the error A in the “TMP2222”.

When an error B for log acquisition, which is different from the error A, occurs after the processing of S22 (S23), the recording unit 24 creates a folder (TMP3333) corresponding to the thread number (3333) in the detailed log recording area. To do. The recording unit 24 creates a folder and starts acquiring a detailed log regarding error B (S24). The recording unit 24 saves the detailed log acquired for the error B in the "TMP3333". Then, regarding error B, it is assumed that the acquisition of all the detailed logs to be acquired is completed (S25).

The recording unit 24 secures a free space in the SEL recording area for error B and performs SEL recording (S26). Since "1" is used as the SEL_ID, the recording unit 24 sets the SEL_ID to "2". The recording unit 24 records the date and time at the time of SEL recording and the message code in the SEL recording area in association with SEL_ID: "2".

The recording unit 24 acquires the recorded SEL_ID "2" from the SEL recording area (S27). The recording unit 24 renames the name of the folder "TMP3333" created for error B to the name "ID0002" corresponding to SEL_ID "2".

Then, regarding error A, it is assumed that the acquisition of all the detailed logs to be acquired is completed (S30). The recording unit 24 secures a free space in the SEL recording area and performs SEL recording (S31). Since "1" and "2" are used as the SEL_ID, the recording unit 24 sets the SEL_ID to "3". The recording unit 24 records the date and time at the time of SEL recording and the message code in the SEL recording area in association with SEL_ID: “3”.

The recording unit 24 acquires the recorded SEL_ID: "3" from the SEL recording area (S32). The recording unit 24 renames the name of the folder "TMP2222" created for error A to the name "ID0003" corresponding to SEL_ID "3".

As described above, in the first example, unlike the example shown in FIG. 4, a state in which a space is created in the middle of the SEL recording area and a state in which the order of the SEL_ID and the order of the date and time are reversed do not occur. That is, the information processing apparatus of the embodiment can ensure the consistency of the recorded log.

<Second example of log recording processing>
FIG. 11 shows a second example of the log recording process. In the example shown in FIG. 11, it is assumed that two events (errors) to be logged occur in one thread. For example, when creating one thread for one interrupt from hardware, it is assumed that two errors occur for one interrupt.

When the error A for which the log is to be acquired occurs (S41), the recording unit 24 creates a folder "TMP3333" corresponding to the thread number "3333" in the detailed log recording area. Then, the acquisition unit 25 starts acquiring the detailed log 1 regarding the error A (S42). The recording unit 24 saves the detailed log 1 acquired for the error A in the “TMP3333”.

Further, when the error B of the log acquisition target occurs (S43), the recording unit 24 creates a folder “TMP3333_1” corresponding to the thread number “3333” in the detailed log recording area. Then, the acquisition unit 25 starts acquiring the detailed log 2 regarding the error B (S44). The recording unit 24 saves the detailed log 2 acquired for error B in "TMP3333_1".

When the error B is related to the error A, the recording unit 24 may save the detailed log regarding the error B in the folder "TMP3333" without creating a new folder "TMP3333_1".

Then, regarding error B, it is assumed that the acquisition of all the detailed logs 2 to be acquired is completed (S45). The recording unit 24 secures a free space in the SEL recording area and performs SEL recording regarding the error B (S46). Since "1" is used as the SEL_ID, the recording unit 24 sets the SEL_ID to "2". The recording unit 24 records the date and time at the time of SEL recording and the message code in the SEL recording area in association with SEL_ID: "2".

The recording unit 24 acquires the recorded SEL_ID "2" from the SEL recording area (S47). The recording unit 24 renames the name of the folder "TMP3333_1" created for the error A to the name "ID0002" corresponding to the SEL_ID "2" (S48).

Then, regarding error A, it is assumed that the acquisition of all the detailed logs 1 to be acquired is completed (S49). The recording unit 24 secures a free space in the SEL recording area and performs SEL recording regarding the error A (S50). Since "1" and "2" are used as the SEL_ID, the recording unit 24 sets the SEL_ID to "3". The recording unit 24 records the date and time at the time of SEL recording and the message code in the SEL recording area in association with SEL_ID: “3”.

The recording unit 24 acquires the recorded SEL_ID "3" from the SEL recording area (S51). The recording unit 24 renames the name of the folder "TMP3333" created for the error A to the name "ID0003" corresponding to the SEL_ID "3" (S52).

As described above, in the second example, unlike the example shown in FIG. 4, a state in which a space is created in the middle of the SEL recording area and a state in which the order of the SEL_ID and the order of the date and time are reversed do not occur. That is, the information processing apparatus of the embodiment can ensure the consistency of the recorded log even when a plurality of errors occur in one thread.

<Others>
The present embodiment is not limited to the embodiments described above, and various configurations or embodiments can be taken within a range that does not deviate from the gist of the present embodiment.

1 1st server 2 2nd server 11 BMC
12 Storage device 13a First communication interface 13b Second communication interface 13c Third communication interface 14 Processor 15 PSU
16 sensor 17 CMU
18 CPU
19 Memory 21 Communication unit 22 Storage unit 23 Selection unit 24 Recording unit 25 Acquisition unit

Claims (7)

An acquisition unit that acquires event data related to the event in response to the detection of the event that has occurred.
After the acquisition of the event data is completed, a recording unit that sets the first identification information corresponding to the event and records the log corresponding to the event data in association with the first identification information.
Based on the relationship between the event that occurred and the event shown in the recorded log, the acquisition unit is not allowed to acquire the event data, or the acquisition unit is made to acquire a part of the event data. Or a selection unit that selects whether to have the acquisition unit acquire all the event data,
An information processing device characterized by being equipped with. The recording unit
Set the second identification information corresponding to the event data,
The information processing apparatus according to claim 1, wherein after recording the log, the second identification information is updated based on the first identification information. The recording unit
The information processing apparatus according to claim 1 or 2, wherein the time when the log is recorded is recorded in association with the log. The selection unit, the generated event, if the log recorded is the same as the events indicative, 1 to claim, characterized in that choose not to acquire the event data to the acquisition unit The information processing apparatus according to any one of 3. When the event that has occurred is related to the event indicated by the log, the selection unit either causes the acquisition unit to acquire a part of the event data, or causes the acquisition unit to acquire all the events. The information processing apparatus according to any one of claims 1 to 4, wherein data is to be acquired. The selection unit, the generated event is associated with an event indicated by the log recorded, and the occurrence time of the event that has occurred is, occurrence time and the predetermined range of events indicated by the log recorded In the case of, any one of claims 1 to 5 , wherein the acquisition unit is not allowed to acquire the event data , or the acquisition unit is selected to acquire a part of the event data. The information processing device described in the section. According to the detection of the event that has occurred, the event data related to the event is acquired, and the event data is acquired.
After the acquisition of the event data is completed, the first identification information corresponding to the event is set, and the log corresponding to the event data is recorded in association with the first identification information.
The information processing device executes the processing ,
In the acquisition of the event data, the event data is not acquired or a part of the event data is acquired based on the relationship between the event that has occurred and the event indicated by the recorded log. A logging method, characterized in that the information processing apparatus further executes a process of selecting whether to acquire all the event data.

Images