Hi,
I'm using your configuration without these error messages. Can you help me to reproduce this problem?
Just a theory without having looked at the source code: When the master process collects child processes after receiving a SIG_CHILD it frees the slots on the global score board which afaik also contains the startup time of the process. If the time span between startup and collection of a process exceeds a certain threshold than the message is printed. Because peruser has some very long running child processes (i.e. multiplexers) the warning is printed when these processes exit due to the restart.
Someone shoud grep the apache source code for the error message and confirm this theory. If its true than it might be sufficient to regularly update the time on the score board...
Just mit 2 cents, Stefan _______________________________________________ Peruser mailing list [email protected] http://www.telana.com/mailman/listinfo/peruser
