G
George Potemkin
Guest
As usually more questions: Signal diagnostic mechanism decribes SIGALRM signal as: (4397) utalsig: Utility Alarm signal handler. (4468) immediate: do not delay handling this signal. Some old knowledgebase articles wrote: "The 4GL PAUSE statement will cause the Progress session to wait for a SIGALRM signal sent by the OS." SIGALRM "should wake up all clients from any PAUSE statements." But it's not true at least for the modern Progress versions. Is SIGALRM completely ignored by Progress? Proshut -C disconnect sends a signal to the self-service clients: It's SIGEMT (Solaris, AIX, HP-UX - Progress 10.2B) or SIGTRAP (HP-UX - Progress 11.7) or SIGUSR2 (Linux) But they do not have any visible effects on Progress sessions if we send the same signals using the kill command. On other hand, Windows does not use the signals and proshut -C disconnect on Windows has the limitations that do not exist on Unix. Do these signals only force a session to check a 'usrtodie' flag in shared memory? There are many knowledgebase articles that discuss the dangerous of the 'trap' command. But Progress sessions seem to ignore the traps. Or did I do it in a wrong way? Example: trap "" SIGHUP mpro sports2000 Progress session will be able to get SIGHUP I tried it with Progress 10.1C, 10.2B and 11.7
Continue reading...
Continue reading...