Zombie fglrun processes on AIX
Posted in 1999
Topics: Connectivity: ESQL/C, 4GL & Embedded SQL, Platform-Specific Issues, Versions, Editions & End-of-Life
I'm running the IDS 7.30.UC5 engine on an IBM RS/6000 (AIX 4.3.2) machine. The apps are written in D4GL (4J's) and we're using only shared memory connections. The problem is that when the client workstations exit the apps (running them using the WTK windows xterm tcl/tk service) that the fglrun processes on the AIX box sometimes don't always exit, but instead turn into orphaned zombies and still take up memory. If I shut down the engine, then the zombies begin spinning away in a runaway mode, eating up 100% of available cpu cycles too. Kill in them with "kill -9 <pid>" gets rid of them. Many years ago I used to encounter similar phenominon running Informix Online 5.x on HP-UX systems, but this was due mostly to users killing their unix sessions before exiting the 4GL apps first. I'm seeing this on the AIX box even when the users do a normal exit from the 4GL application's main screen. Could this possible be that the app itself is not really doing a "normal" exit, but instead is aborting abnormally towards the "end" of whatever should be a normal exit? I'm getting no core files or other indications of such,....... just leftover zombies. Anyone else out there seeing this too? -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own