From owner-freebsd-questions Wed Mar 24 18:49:56 1999 Delivered-To: freebsd-questions@freebsd.org Received: from bolero-x.rahul.net (bolero.rahul.net [192.160.13.1]) by hub.freebsd.org (Postfix) with SMTP id B1EF114CA0 for ; Wed, 24 Mar 1999 18:49:52 -0800 (PST) (envelope-from dhesi@rahul.net) Received: from waltz.rahul.net by bolero-x.rahul.net with SMTP id AA26237 (5.67b8/IDA-1.5 for ); Wed, 24 Mar 1999 18:49:31 -0800 From: Rahul Dhesi Received: by waltz.rahul.net (5.67b8/jive-a2i-1.0) id AA16939; Wed, 24 Mar 1999 18:49:30 -0800 Date: Wed, 24 Mar 1999 18:49:30 -0800 Message-Id: <199903250249.AA16939@waltz.rahul.net> To: freebsd-questions@freebsd.org Subject: Re: Out of Swap Space hangs system Newsgroups: a2i.lists.freebsd-questions References: X-Newsreader: NN version 6.5.1 (NOV) Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Christopher Palmer writes: >On Tue, 23 Mar 1999, Rahul Dhesi wrote: >> >By all means provide an algorithm to solve it, if it concerns >> >you. I think you'll find it's non-trivial. >> >> Here's one possible algorithm, in pseudo-perl: >> >> for $pid (@all_active_pids) { >> $critical_system_process{$pid} && next; >> &kill($pid); >> &got_enough_swap() && last; >> } >That's a workaround, not a solution. Limiting the amount of virtual >memory users are alotted is an already-implemented workaround. I don't mind if you call it a 'workaround'. I only claim that it an 'algorithm'. And it achieves the objective of breaking deadlock and preventing the system from hanging or panicking. -- Rahul Dhesi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message