| 摘要 |
PURPOSE:To prevent the dissipation of parts required for a failure analysis out of bits of trace information overwritten in a short cycle, and to hold it in a file for longer time, by collecting a failure event, and the bit of trace information in the neighborhood of the event when a failure is generated, in a log collection file. CONSTITUTION:In a log collection when a log even is generated, a log collecting means 2 writes the log event at a log buffer 32 by a log buffer write process 31, and also, a peripheral trace collecting means 3 decides whether the peripheral trace is needed for the log event by a log even decision process 41, and when it is needed, the required number of items of the most new bit of trace information are collected from a trace buffer 22 by a peripheral trace data collection process 42, and they are written on the log buffer 32. And similarly in a trace collection, the data is written on a log collection file 34 at a time when the log buffer 32 is filled, by a log collection file write process 33. In this way, it is possible to collect the trace information of a system by which the failure event is generated, on the log collection file 34.
|
