severity 41554 minor tags 41554 patch thanks The complaint here was that when the pidfile method is used to check whether an instance of a daemon is running, it is the existence of the pidfile and not the existence of the process named in the pidfile that is checked.
This is reasonable behavior. start-stop-daemon is designed to start a daemon unless an *instance* of that daemon is already running. Thus, it needs to check for processes of a given type, not for a process with a specific process i.d. The type of the process is specified either by the name of the executable or by the name of the pidfile it creates. It could be done the other way (i.e., start-stop-daemon could check for the actual process named in the pidfile) but I don't think it would be very safe to make that change: it might cause some serious breakage. So I guess the manual could be clarified a bit. Here's a suggestion. --- start-stop-daemon.8_ORIG Tue Oct 1 11:00:55 2002 +++ start-stop-daemon.8 Tue Oct 1 11:01:58 2002 @@ -73,7 +73,7 @@ ). .TP \fB-p\fP|\fB--pidfile\fP \fIpid-file\fP -Check for processes whose process-id is specified in +Check for processes that create the file .IR pid-file . .TP \fB-u\fP|\fB--user\fP \fIusername\fP|\fIuid\fP
