Date: Fri, 28 May 1999 20:12:53 -0700 From: Jerry Preeper <preeper@cts.com> To: Doug White <dwhite@resnet.uoregon.edu> Cc: freebsd-questions@FreeBSD.ORG Subject: Re: swap_pager: out of swap space (-->crash) - procmail is problem Message-ID: <3.0.5.32.19990528201253.041a8a20@crash.cts.com> In-Reply-To: <Pine.BSF.4.03.9905281630160.11808-100000@resnet.uoregon.ed u> References: <3.0.5.32.19990528154716.041b4280@crash.cts.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Well, After crashing a few more times I finally found the culprit. Nothing like someone sending a huge email (at least 50MB of the message appears to get processed before it crashes the web and name servers). Procmail is being used to filter out spam and appears to be killing the system big time when a large attachment goes through it. Granted, I've never seen a 50MB+ email before, but hey, there's a first time for everything I guess. Needless to say, procmail has now been turned off since it's getting the message from the user before sendmail can reject it (added innocent@untilprovenguilty.com 550 Please contact the webmaster immediately to the /etc/mail/access file - real address hidden to protect the innocent) but the sending server just keeps trying to send it over and over again. Has anyone else ran into a problem with procmail crashing the system on large emails being sent to it? Here is the top of the ps aux output when I was finally able to catch this. <fontfamily><param>Times New Roman</param><bigger>su-2.01# ps aux USER PID %CPU %MEM VSZ RSS TT STAT STARTED TIME COMMAND root 6519 91.6 34.6 44140 44608 ?? Rs 6:50PM 3:46.95 procmail -f art root 6518 1.4 0.7 824 916 ?? S 6:50PM 0:06.98 sendmail: NAA14123: fr </bigger></fontfamily> Is there any fix or patch for procmail or sendmail to prevent this (ie, any email over xMB in size gets dropped with a nice message to them and me about what happened and why. Thanks for any assistance someone can provide. Jerry At 04:31 PM 5/28/99 -0700, Doug White wrote: >On Fri, 28 May 1999, Jerry Preeper wrote: > >> I'm hoping someone might be able to provide some insight. When I got back >> from lunch today, I found a few console messages on my connection to my >> server running 2.2.8-Stable (from cvsup about 2 weeks ago), like this: > >(some odd wall messages) > >> and then found my nameserver and webserver dead. I rebooted the machine >> and 10 minutes later, it crashed again. Reboot. Then crashes again. >> Reboot again. >> >> I did get to download the messages file and after the first reboot found >> this in the messages file (which is just filled with May 28 14:49:22 trixie >> /kernel: pid 1302 (httpd), uid 50: exited on signal 11 messages): >> >> May 28 14:35:30 trixie named[675]: Ready to answer queries. >> May 28 14:38:13 trixie proftpd[898]: FTP session closed. >> May 28 14:39:11 trixie su: jerry to root on /dev/ttyp0 >> May 28 14:43:06 trixie /kernel: swap_pager: out of swap space >> May 28 14:43:15 trixie /kernel: pid 1139 (getad.cgi), uid 50: exited on >> signal 11 >> May 28 14:43:16 trixie /kernel: pid 1140 (getad.cgi), uid 50: exited on >> signal 11 >> May 28 14:43:25 trixie /kernel: swap_pager: out of swap space >> May 28 14:48:16 trixie /kernel: pid 1144 (getad.cgi), uid 50: exited on >> signal 11 >> May 28 14:48:40 trixie named[675]: flushset: out of memory - ABORT >> May 28 14:48:40 trixie /kernel: pid 675 (named), uid 0: exited on signal 6 >> May 28 14:48:41 trixie /kernel: pid 1143 (perl5.00404), uid 50: exited on >> signal 11 > >Looks like a runaway process. Perhaps your 'getad.cgi' has a bug? Watch >top for a while. > >> The machine is PII-233, Asus P2L97S MB, 128 MB ram, 141 MB swap space, >> 4.5GB Barracuda HD running basic web server functions. Apache is 1.3.4 and >> sendmail is 8.9.2 > >You should have more swap. > >> Mounted on >> /dev/sd0s1a 4168505 3157871 677154 82% 169750 844008 17% / >> procfs 4 4 0 100% 44 4072 1% /proc >> >> so it seems I have plenty of room for the swap. I checked the archives and >> saw that swap should be slightly more than RAM, which it is, but can't seem >> to find out why this is happening. > >Swap is on a separate partition. What does 'swapinfo' report? > >Doug White >Internet: dwhite@resnet.uoregon.edu | FreeBSD: The Power to Serve >http://gladstone.uoregon.edu/~dwhite | www.freebsd.org > > > >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
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3.0.5.32.19990528201253.041a8a20>