QNX6 Performance Monitor tools

  1. what are the best performance monitoring tools to use for measuring
    -CPU utilization by each process
    -RAM utilzation by each process

2.how to configure tracelogger/IDE kernel trace logging to log for 24 hours and collect a snapshot of system state at every 1 minute.

Depends on what you mean by snapshot. You could start tracelogger in daemon mode (-d1 -n0) and it will run forever. Then have another application enable all classes and once a minute do a TRACE_START/delay/TRACE_STOP. The time you delay is up to you.

According to that question, does the tracelogger work fine? I have tried to use it starting the tracelogger in daemon mode, and then with a program of mine I have set the classes and events I like to filter and I TRACE_START and TRACE_STOP. Then , I process the output file with the traceprint utility (in the help documents it saids specifically that the source code is available but the QNX people told me that this fact is not true, :frowning: ) and the result is always the same: the first eventsfrom the very start-up of the system, but not those belonging to the period of time when I started and stopped the tracelogger by software.

So, are you really sure that the tracelogger works fine in that way?

I think you are being confused by the output - _NTO_TRACE_START will cause the kernel to emit process and thread create events for all current threads. If you don’t want that you can use _NTO_TRACE_STARTNOSTATE