Date: Fri, 6 Sep 1996 12:23:35 +0200 (MET DST) From: grog@lemis.de (Greg Lehey) To: klam@awod.com (Ken Lam) Cc: freebsd-isp@freebsd.org Subject: Re: Unusual sendmail behavior ? Message-ID: <199609061023.MAA13964@allegro.lemis.de> In-Reply-To: <1.5.4.32.19960905233944.0094aa94@awod.com> from "Ken Lam" at Sep 5, 96 07:39:44 pm
next in thread | previous in thread | raw e-mail | index | archive | help
Ken Lam writes: > > At 05:36 PM 9/5/96 -0400, Kurt Schafer wrote: >> >> Can anybody comment on their experiences with the sendmail daemon. I've noticed >> that whenever I do a 'ps' I see upwards of 10 sendmail processes going at any >> given time, and of those, there are always a good number of them that seem >> to last for several minutes. > > nothing wrong with 10 active processes, I've seen averages of 25 mail processes. > >> The /var/spool/mqueue directory yields several 1k message sizes, which I would >> think sendmail could transfer over a T1 line in a considerrably shorter >> period of time than 10 minutes or more. > > check connectivity to the destinations (ie. port 25 usage via netstat, then do > a traceroute or ping). If there is a saturated link...... > > what is the amount of mail received, perhaps some of your customers have > subscribed to very active lists or to every list in the "internet yellow pages". To amplify on this statement: to send a message, sendmail spawns a child process, which then goes through all recipients and tries to send the message to them. This can take quite some time if the list is long and contains destinations which are hard to reach, so it's not unusual to find a number of processes active. I just checked freefall and found: === grog@freefall (/dev/ttype) /a/grog 11 -> ps aux|grep sendmail root 12204 2.0 1.9 952 1212 ?? S 3:20AM 0:01.21 sendmail: DAA11645 ft-mailhost.brunel.ac.uk.: client M root 161 0.0 0.2 520 140 ?? Is 25Aug96 1:29.18 sendmail: accepting connections (sendmail) root 4110 0.0 0.0 556 0 ?? IW 1:00AM 0:00.08 /usr/sbin/sendmail -FCronDaemon -odi -oem -t root 8253 0.0 1.6 952 1000 ?? I 2:40AM 0:01.27 sendmail: WAA24903 real.za.freebsd.org.: client DATA 3 root 8519 0.0 2.6 1396 1616 ?? I 2:44AM 0:05.20 sendmail: CAA08366 sel1.zit.th-darmstadt.de.: client R root 8922 0.0 2.3 1244 1468 ?? I 2:48AM 0:03.82 sendmail: CAA08594 sel1.zit.th-darmstadt.de.: client R root 9153 0.0 1.8 816 1116 ?? S 2:50AM 0:00.74 sendmail: WAA24796 real.za.freebsd.org.: client DATA 3 root 9422 0.0 2.1 1080 1308 ?? I 2:52AM 0:02.78 sendmail: CAA08590 mail.jp.freebsd.org.: client RCPT ( root 9837 0.0 2.2 1116 1380 ?? S 2:56AM 0:02.76 sendmail: CAA09584 mail.jp.freebsd.org.: client RCPT ( root 10051 0.4 2.2 1116 1356 ?? S 3:00AM 0:02.53 sendmail: CAA09910 ktnet.ktnet.co.kr.: user open (send root 10210 0.0 2.0 1008 1268 ?? S 3:04AM 0:01.99 sendmail: CAA09567 sel1.zit.th-darmstadt.de.: client D root 10337 0.0 1.8 912 1136 ?? S 3:08AM 0:00.99 sendmail: CAA09905 mail.jp.freebsd.org.: client RCPT ( root 10407 0.0 1.7 776 1084 ?? I 3:10AM 0:00.52 sendmail: OAA19036 gledek.binus.ac.id.: client DATA 35 root 10525 0.0 1.9 964 1204 ?? I 3:12AM 0:01.51 sendmail: CAA09839 sel1.zit.th-darmstadt.de.: client D root 11381 0.0 1.7 776 1080 ?? I 3:16AM 0:00.50 sendmail: DAA11277 sel1.zit.th-darmstadt.de.: client R root 12147 0.0 1.6 636 984 ?? I 3:20AM 0:00.18 sendmail: DAA04692 jojo.kotel.co.kr.: user open (sendm root 12163 0.0 1.6 676 1004 ?? S 3:20AM 0:00.17 sendmail: TAA08041 gledek.binus.ac.id.: client DATA 25 grog 12251 0.0 0.9 288 584 pe S+ 3:20AM 0:00.03 grep sendmail As you can see, sendmail changes its argv values to show who it is trying to contact. This particular display suggests that there are currently network problems to Japan, Korea, and Germany. Greg
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199609061023.MAA13964>