From owner-freebsd-questions Tue Aug 21 19:36:25 2001 Delivered-To: freebsd-questions@freebsd.org Received: from mtiwmhc26.worldnet.att.net (mtiwmhc26.worldnet.att.net [204.127.131.51]) by hub.freebsd.org (Postfix) with ESMTP id 4D79337B414 for ; Tue, 21 Aug 2001 19:36:08 -0700 (PDT) (envelope-from parv@worldnet.att.net) Received: from worldnet.att.net ([32.101.235.225]) by mtiwmhc26.worldnet.att.net (InterMail vM.4.01.03.16 201-229-121-116-20010115) with ESMTP id <20010822023606.UVMV26461.mtiwmhc26.worldnet.att.net@worldnet.att.net>; Wed, 22 Aug 2001 02:36:06 +0000 Received: by worldnet.att.net (Postfix, from userid 1001) id 6BC8A50D1E; Tue, 21 Aug 2001 22:36:08 -0400 (EDT) Date: Tue, 21 Aug 2001 22:36:08 -0400 From: parv To: Mike Meyer Cc: questions@freebsd.org Subject: Re: "proc: table is full" while cleaning some ports Message-ID: <20010821223608.B509@moo.holy.cow> Mail-Followup-To: Mike Meyer , questions@freebsd.org References: <88854365@toto.iv> <15235.1141.812127.446128@guru.mired.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <15235.1141.812127.446128@guru.mired.org>; from mwm@mired.org on Tue, Aug 21, 2001 at 08:01:41PM -0500 Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG this was, on the fateful occasion around Aug 21 21:01 -0400, sent by Mike Meyer > > parv types: > > sometimes i get that dreaded message when i run 'make clean' in > > some ports like perl5, gnomevfs, or galeon, etc. i tried w/ > > maxusers=32 in those three ports. > > > > w/ maxusers=128, i still got the same kernel message when make clean > > was going in gnomevfs and galeon. > ... > > nice -(10|20) didn't help either. > > That cuts down on CPU usage, not processes. So it won't help. nice to know & nice to be confirmed. > > another thing, could it have to do me editing the ports/Mk/bsd*mk files > > and/or using ddb kernel option? > > I seriously doubt it. so did/do i. but, sometimes things like these very well may be the cause; so i thought i should mention it anyway. > Personally, I set "NOCLEANDEPENDS=yes" in /etc/make.conf. That solves > that problem, but probably doesn't do what you want. correcto mundo... > I also set WRKDIRPREFIX to a scratch disk, and just blow away the > entire directory every so often. i can do the same (read: have space to spare), but in case of a crash fsck may take too long. WRKDIRPREFIX would be set to the partition where i have sources & ports & docproj & var. next time i partition disk, will keep in mind the need for a scratch (partition). > If you are using the -j option when doing "make clean", try reducing > that. If you aren't, make sure you use -B to disable parallel cleans. never bothered to modify the "default" clean operation. so -B option will be worth investigating. may be due lack of it, while kernel was busy spewing "proc: out of space" messages, on the term where cleaning was going on, i was getting "resources unavailable" & "cannot fork" messages. thanks mike. -- so, do you like word games or scrabble? - parv To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message