JPH07281929A - Trace information collecting device - Google Patents

Abstract

PURPOSE:To prevent erasion of the useful information stored in a trace area. CONSTITUTION:A failure detector means 2 decides whether the harfware or software abnormality is occurred right after the trace information are collected. If occurred, the trace information collected at the time of occurrence of the abnormality are not stored in a trace area 4 of a storage means 1 under the control of a control means 3. Meanwhile the collected trace information are immediately compared with the trace information collected in the precedent time. When the perfect coincidence is secured between both trace information, the former information are not stored in the area 4. Under such conditions, a fact is show that the same trace information are overlapping with each other in the extension bit of the area 4. Then the overlapping frequency is stored. When the overlapping frequency reaches a prescribed level, the abnormality is processed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a trace information collecting device for collecting information used for analyzing a processing process in a computer system.

[0002]

2. Description of the Related Art Generally, trace information is collected in order to monitor the operation of a computer system in time series. Such trace information is collected each time an event of interest occurs that requires analysis in analyzing the operation of the device. For example, the communication control device collects trace information each time an event such as the start of transmission or the end of reception occurs and stores it in a specific trace area in the memory. Then, this trace information is used when, for example, analyzing the state of the communication control device when a failure occurs.

[0003]

However, the above-mentioned conventional techniques have the following problems. That is, a failure may occur in the computer system such as the communication control device during the collection of the trace information described above. In general, a failed device often repeats the same operation or process. For example, in the communication control device, if an abnormality occurs in the line portion, the same reception state may continue indefinitely. In that case, if the trace information at the time of reception is collected, the same trace information is stored in the trace area one after another, and the trace area becomes full. When the trace area is cyclically used, the trace information of the current reception abnormality is overwritten on the trace information collected before the failure occurs and stored in the trace area. As a result, there is a problem that useful trace information is erased.

Also, when a program is executed in a loop state due to a software failure (bug in the program), the same trace information is repeatedly collected and useful trace information is overwritten. There was a problem that it was erased by.

[0005]

The trace information collecting apparatus of the present invention is characterized by the following points in order to solve the above-mentioned problems. (1) A storage means is provided for storing the trace information collected when a predetermined event occurs at a predetermined time. A failure detection means for detecting a failure of the device is provided. A control unit is provided that prohibits the storage of the trace information collected when the fault detection unit detects a fault in the device.

(2) A storage means for storing the trace information collected when a predetermined event occurs at a predetermined time is provided. A detection unit that compares the trace information collected this time with the trace information collected immediately before and detects whether or not the same trace information is repeatedly collected is provided. When the detection unit repeatedly collects the same trace information, the control unit is provided to prohibit storing the duplicated trace information in the storage unit.

(3) A storage means is provided for storing the trace information collected when a predetermined event occurs at a predetermined time. When the same trace information is repeatedly collected, a counter for counting the number of repetitions is provided. The storage unit is provided with a control unit that prohibits the duplicated trace information from being stored and that abnormally terminates the device when the number of times the counter has counted exceeds a predetermined number.

[0008]

[Operation] Immediately after collecting the trace information, it is determined whether or not an abnormality such as hardware has occurred, and when the abnormality occurs, the trace information collected at that time is stored in the trace area of the storage means. do not do. Immediately after collecting the trace information, the trace information is compared with the previously collected trace information, and when the two completely match, the trace information collected this time is not stored in the trace area of the storage means. In that case, the extension bits of the trace area indicate that the same trace information is duplicated, and the number of duplications is stored. Then, when the number of times of duplication reaches the predetermined number of times, the abnormality processing is performed.

[0009]

Embodiments of the present invention will now be described in detail with reference to the drawings. FIG. 1 is a block diagram of an embodiment of the trace information collecting apparatus of the present invention. In the figure, a communication control device 10 controls data communication of the information processing device 11, and another information processing device 13 via a communication line 12.
Is connected to the communication control device 14. Communication control device 1
0 indicates a storage unit 1, a failure detection unit 2, and a control unit 3.
And are provided. The storage means 1 comprises a RAM (random access memory) or the like, and stores programs and data executed by the control means 3. Addresses n to m of the storage means 1 are used as a trace area 4. The trace area 4 stores trace information collected every time a predetermined event of the communication control device 10 occurs. The trace information is collected in the communication control device 10 every time an event such as transmission or reception occurs. The contents of this trace information and the collection procedure will be described later.

The fault detecting means 2 comprises various detectors for detecting a fault of hardware or software, and detects various faults of the communication control device 10. For example, as will be described later, the line synchronization loss, the abnormality of the communication LSI, and the like are detected. The control unit 3 includes a CPU and the like, executes communication control, and stores trace information in the trace area 4 each time a predetermined event such as transmission or reception occurs.
Then, the trace information collected when the failure of the communication control device 10 is detected by the failure detecting means 2 is not stored in the trace area 4.

FIG. 2 shows a processing procedure of the control means 3 shown in FIG. In FIG. 2, in a normal process of the control means 3, the route R1 is taken. In the route R1, it is determined whether the transmission is performed in step S1, the reception is performed in step S2, or whether the timeout occurs in step S3. As a result, when events such as transmission and reception occur,
In steps S4, S5 and S6, processing corresponding to each event is performed. Then, in steps S7, S8, and S9, the trace information corresponding to each event is collected. For example, the communication line 12
When the data is received from the route R2, the control means 3 determines the reception in step S2 in the middle of the route R1 and then the route R2.
Branch to. Then, in step S5, the reception process is started. When the reception process is completed, the trace information is collected in step S8. In the trace information, as shown in FIG.
It includes the line number of the plurality of communication lines, the reception result, the type of received data, and the like. After collecting the trace information, the normal route R1 is returned to.

Details of the trace information will be described with reference to FIG. The trace information for one time consists of 16 bits, for example. This is divided into a 4-bit A section and a 12-bit B section. The line number (for example, "0" ...
“16”) is stored. Part B stores detailed contents of events such as reception start and abnormal reception end. Such trace information is sequentially stored in the trace area 4 in the storage means 1 under the control means 3. That is, for example, as shown in the drawing, the n-th address to the m-th address are the trace areas. The storage address of the latest trace information is stored in the pointer (not shown), and the content of the pointer is updated every time the trace information is newly stored. After the storage up to address m is completed, the pointer is n
Returned to the address. As a result, the trace area is cyclically used.

FIG. 4 shows a trace information collecting procedure according to the present invention. As shown in the figure, the trace information is collected in step S1, and immediately after that, it is determined in step S2 whether or not an alarm has been generated by the fault detecting means 2. When the alarm is generated, the trace information is not stored in the trace area 4 and the collection of the trace information is ended as it is. On the other hand, if no alarm is generated, the trace information is stored in the trace area 4 in step S3,
The pointer is updated in step S4.

FIG. 5 shows details of the alarm. Examples of the hardware failure alarm include synchronization loss detection (line abnormality) of a dedicated line and communication LSI I / O access timeout. Further, as a software failure alarm, there is a watchdog timeout for detecting that a loop state of the program has occurred. If any one of these times out, the alarm signal 51
Is turned on, and the abnormal state display is set in the register 52. Here, if the control means 3 reads the register 52, it can be determined whether or not an alarm has occurred. Register 52
The type of timeout is set in the content of, and the cause of failure can be analyzed by referring to this.

Next, a second embodiment of the present invention will be described. The hardware of the second embodiment is almost the same as that of FIG. 1, but the trace information used is different. Further, in the second embodiment, the control means 3 of FIG. 1 also serves as a detection means for comparing the trace information collected this time with the trace information collected immediately before and detecting whether or not they are the same. FIG. 6 shows details of the trace information in the second embodiment of the present invention. In FIG. 6, the trace information is composed of, for example, 16 bits, the 0th bit is an extension bit, and the 1st bit and thereafter are the A part and the B part. The extension bits are used to distinguish the trace information from the normal case to the special case where the same information is repeated. Part A is, for example, 3 bits, B
The part has, for example, 12 bits and has the same meaning as in the trace information of FIG.

When the trace information is stored, it is normally shown in FIG.
It is used in the format shown in (a). On the other hand, when the same trace information is repeatedly stored, it is used in the format shown in FIG. 6 (b). That is, if the collected trace information has the same content as the previous trace information, the extension bit is set to "1" and a count indicating how many times the trace information having the same content continues in the next 16-bit trace area. Set the value.

FIG. 6C shows an example of actual collection of trace information. In FIG. 6C, address n and n + 1
Different trace information is collected and stored in the address every time. Further, at the address n + 2, only the first trace information when the same trace information is repeatedly collected is stored. Then, the number of times k is stored in the address n + 3. That is, after the trace information having the content stored in the address n + 2 is continuously collected and stored k times, the normal trace information is collected again from the address n + 4 by resetting or abnormal processing.

FIG. 7 is a flow chart showing the operation of the trace information collection system of FIG. 6, that is, the processing procedure of the second embodiment of the present invention. After collecting the trace information in step S11, first, in step S12, it is compared with the content of the previous trace information. If the content is different, the trace information is stored in step S13. Then, in step S14, the address is incremented by "1" (+1) to indicate the next storage position, and the process ends. If the contents are the same in step S14, it is next determined whether or not the extension bit of the previous trace information is "1". If it is "1", the extension bit has already been set to "1", so the count value is incremented by "1" in step S16.

If the extension bit is not "1" in step S15, the extension bit is set to "1" in step S17, and the count value = 1 is set to the next address in step S18. Then, since the storage position of the next trace information is the current address + 1 address, the address is incremented by "1". Then, in step S20, it is determined whether the count value is overflow. For example, if the trace area has a 16-bit width, it is determined whether or not it is "FFFF". And
If it is an overflow, there is a possibility of a failure, so in step S21, the abnormality processing determined for each device is performed. For example, processing such as resetting the communication LSI or notifying an operator of an information processing apparatus, which is a higher-level apparatus of the system, of an abnormality is performed.

As described above, when the trace information having the same content is repeatedly collected, it is compressed and stored, thereby preventing the trace information having the same content from being overwritten in the entire area of the trace area. it can. As a result, it is possible to prevent the effective trace information from being destroyed, and it is possible to efficiently isolate the failure and confirm the operation in the analysis of the operation. Further, since the failure can be found from the contents of the trace information, the failure can be found and notified in a short time.

Needless to say, the present invention is not limited to the above-mentioned embodiments, and various modifications can be made without departing from the features of the invention. For example, when the same trace information is repeatedly collected, the number of repetitions is calculated as the count value. However, when it is not necessary to know the number of times, such a count value may not be provided. Even in that case, by looking at the extension bits of the trace area 4, it is possible to know whether or not a failure has occurred.

[0022]

As described above, according to the trace information collecting apparatus of the present invention, the collection of trace information is stopped when a failure occurs in the apparatus or the trace information having the same content is repeatedly collected a predetermined number of times. Since the collection of trace information is stopped when it is detected that it is possible to prevent the trace area from becoming full with the same information, it was obtained in the state immediately before the failure. It is possible to prevent the trace information of the current reception abnormality from being overwritten and erased.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of a trace information collecting device of the present invention.

FIG. 2 is a flowchart illustrating a processing procedure of a control unit.

FIG. 3 is an explanatory diagram of details of contents of trace information.

FIG. 4 is a flowchart illustrating a trace information collecting procedure according to the present invention.

FIG. 5 is an explanatory diagram of details of an alarm.

FIG. 6 is an explanatory diagram of details of trace information according to the second embodiment of this invention.

FIG. 7 is a flowchart illustrating a processing procedure of a second embodiment of the present invention.

[Explanation of symbols]

 1 Storage Means 2 Failure Detection Means 3 Control Means

Claims (3)

[Claims] 1. A storage unit for storing trace information collected when a predetermined event occurs at a predetermined time, a failure detection unit for detecting a failure of the apparatus, and a failure detection unit for collecting the failure information of the apparatus. And a control means for prohibiting the storage of the trace information in the storage means. 2. The storage means for storing the trace information collected when a predetermined event occurs at a predetermined time is compared with the trace information collected this time and the trace information collected immediately before, and the same trace information is repeated. And a control means for prohibiting storing duplicated trace information in the storage means when the same trace information is repeatedly collected by the detection means. Trace information collection device. 3. A storage means for storing trace information collected when a predetermined event occurs at a predetermined time, a counter for counting the number of repetitions when the same trace information is repeatedly collected, and a duplication in the storage means. And a trace information collecting device for prohibiting the storage of the trace information and for abnormally terminating the device when the number of times the counter has counted exceeds a predetermined number.