From owner-freebsd-questions Tue Nov 28 15:30:20 2000 Delivered-To: freebsd-questions@freebsd.org Received: from camelot.bitart.com (BITart-45.BITart.com [206.103.221.45]) by hub.freebsd.org (Postfix) with SMTP id 6A05037B402 for ; Tue, 28 Nov 2000 15:30:17 -0800 (PST) Received: (qmail 2538 invoked by uid 101); 28 Nov 2000 23:30:11 -0000 Message-ID: <20001128233011.2537.qmail@camelot.bitart.com> Content-Type: text/plain MIME-Version: 1.0 (NeXT Mail 4.2mach v148) In-Reply-To: X-Nextstep-Mailer: Mail 4.2mach (Enhance 2.2p1) Received: by NeXT.Mailer (1.148) From: Gerd Knops Date: Tue, 28 Nov 2000 17:30:10 -0600 To: Roy Nasser Subject: Re: Memory utilization rises too much, and then machine crashes... Cc: "'freebsd-questions@FreeBSD.ORG'" Reply-To: gerti-freebsdq@BITart.com References: Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Roy Nasser wrote: > Hi, > > I am migrating my system from a RedHat Linux box, and I am having > some problems... > > I have a perl deamon that needs to run... It serves as a > Search-server. When I run it (I use nice -10), and then go straight > into TOP, i have about 92mb (out of 128mb) free on my Mem. *I have > compiled the source of Perl 5.6.0 to see if it improved on something, > but no luck...* > > After a few seconds, that falls to about 9mb, and the following is > what TOP shows me: > last pid: 13023; load averages: 0.05, 0.10, 0.09 up 0+03:07:06 > 19:07:14 > 58 processes: 1 running, 57 sleeping > CPU states: 0.0% user, 0.0% nice, 0.0% system, 0.8% interrupt, 99.2% > idle Mem: 43M Active, 44M Inact, 21M Wired, 4440K Cache, 22M Buf, > 8964K Free Swap: 258M Total, 258M Free > > I must run 3 similar perl deamons on each server, for diffeent types > of searches (I have two, load balanced, servers). > > After a while, the memory (with 3 servers) falls to literally 0kb, > and the machine stops handling requests, with an error that says: "no > more processes" whenever i try to enter any command... The machine > must be forcefully restarted (as not even reboot or shutdown commands > work), and then brought up again... > 'No more processes' most often is caused by zombie processes, that the perl daemons leave laying around. Most likely the perl daemons fork at some place, but fail to reap the child processes. Either add a 'wait', or use double forking. Memory usage is a second problem, probably caused by a memory leak in your perl daemons. Not easy to debug. Gerd > What can be causing this, how can I debug it? > > Should I recompile the Kernel, if so, can anyone give me some advice, > I am completely new to this? Below is my machine spec... > > It is a DELL, with Pentium III 800 Processor, 128mb Ram, and a 10GB > IDE drive. > > The INACT memory that top shows, what is this? Is this an "eating up" > of memory of some sort (This happened in some weird cases in RedHat, > where "shared" memory was never freed, even after the process went > down...)? What about the SWAP, should it be used? Is there anyway to > force the process to use it after it exceeds Xmb? > > Thanks for any and all help! > > > Roy Nasser > roy@vem.ca > > > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-questions" in the body of the message To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message