From owner-freebsd-questions Sat Oct 17 06:29:33 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id GAA11474 for freebsd-questions-outgoing; Sat, 17 Oct 1998 06:29:33 -0700 (PDT) (envelope-from owner-freebsd-questions@FreeBSD.ORG) Received: from castor.chuck (lucy.bedford.net [206.99.145.182]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id GAA11468 for ; Sat, 17 Oct 1998 06:29:29 -0700 (PDT) (envelope-from listread@bedford.net) Received: (from listread@localhost) by castor.chuck (8.8.8/8.8.8) id IAA10839; Sat, 17 Oct 1998 08:55:39 -0400 (EDT) (envelope-from listread) Message-Id: <199810171255.IAA10839@castor.chuck> Subject: Re: Dead/zombie'd users ?? In-Reply-To: from Bryce Newall at "Oct 15, 98 00:29:10 am" To: data@dreamhaven.net (Bryce Newall) Date: Sat, 17 Oct 1998 08:55:39 -0400 (EDT) Cc: freebsd-questions@FreeBSD.ORG X-no-archive: yes X-Echelon: aerosol dispersal modalities Reply-to: djv@bedford.net From: "Woodchuck" X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Bryce Newall wrote: > Hi again, > > Somehow, it seems that one of my users must've gotten disconnected, but my > system isn't aware of it. It's showing a 4-day idle time for him: > > [145]root@voyager:/root # w tazq > 12:27AM up 45 days, 15:19, 6 users, load averages: 0.08, 0.05, 0.07 > USER TTY FROM LOGIN@ IDLE WHAT > tazq p6 dial-104.perth.n Sat09PM 4days (tcsh) > tazq p8 dial-104.perth.n Sat04PM 4days (tcsh) > > And when I try to kill processed owned by him on those tty's, with a kill > -9, I don't get an error message, but the processes simply will not die. > Two of his processes don't even seem to exist: > > [149]root@voyager:/root # ps aux | grep tazq > tazq 11534 0.0 0.0 0 0 p8- Z+ - 0:00.00 (pine4) > tazq 21292 0.0 0.0 0 0 p6- Z+ - 0:00.00 (pine4) > tazq 11532 0.0 0.6 552 728 p8- IEs+ - 0:00.00 (tcsh) > tazq 21290 0.0 0.6 552 728 p6- IEs+ - 0:00.00 (tcsh) > > So I try to kill 'em all... ps -jax will show what the parent process is. Kill that. In this case, though, it's highly likely that 11534's parent is 11532. 11532 is I_dle, trying_to_Exit, and is a s_ession leader, and (+) is a foreground process. Why it won't exit is another story. Hangs like this are usually due to some kind of device glitch, like an nfs volume that went away. > > [150]root@voyager:/root # kill -9 11534 21292 11532 21290 > 11534: No such process > 21292: No such process > > Two error messages, two not, but all 4 still exist... > > [151]root@voyager:/root # ps aux | grep tazq > tazq 11534 0.0 0.0 0 0 p8- Z+ - 0:00.00 (pine4) > tazq 21292 0.0 0.0 0 0 p6- Z+ - 0:00.00 (pine4) > tazq 11532 0.0 0.6 552 728 p8- IEs+ - 0:00.00 (tcsh) > tazq 21290 0.0 0.6 552 728 p6- IEs+ - 0:00.00 (tcsh) > > Any suggestions on how to get rid of these ghosties? Rebooting is a sure cure :-( But they are fairly harmless, unless the number of zombies gets large. I would look "upstream", to the tcsh's parent process, trace back to init, and see if it's more obvious what the trouble is. (Man ps to interpret IEs+) Dave -- Strangers know my loved ones' phone numbers!!!! To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message