From owner-freebsd-questions Mon Nov 11 7:46:34 2002 Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 9904237B401 for ; Mon, 11 Nov 2002 07:46:31 -0800 (PST) Received: from mildred.streams.com (streams.com [208.178.202.26]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6B35D43E75 for ; Mon, 11 Nov 2002 07:46:30 -0800 (PST) (envelope-from dlacroix@nospam.suddendeceleration.com) Received: from dlacroix (localhost.streams.com [127.0.0.1]) by mildred.streams.com (8.12.3/8.12.3) with SMTP id gA8Ibwce009158; Fri, 8 Nov 2002 12:37:59 -0600 (CST) (envelope-from dlacroix@nospam.suddendeceleration.com) Message-ID: <00db01c28755$dccf53a0$7d1a910a@dlacroix> From: "David La Croix" To: Cc: Subject: Inetd "Forgets" about one of its services? (Amanda) Date: Fri, 8 Nov 2002 12:37:15 -0600 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1106 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 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 Can somebody tell me why Inetd keeps dropping a UDP service? Anybody know a workaround? My specific problem: I've got an Inetd process (amandad) that periodically just goes into space. This is on a machine with public interfaces, in the DMZ. The nightly amanda backups and checkups periodically fail with errors similar to: WARNING: mildred: selfcheck request timed out. Host down? When this happens -- inetd is no longer listening on port 10080. What puzzles me: A sighup of the inetd service causes it to remember the port -- normal operation resumes. This was happening initially -- and I suspected outside influences, so I moved to running a separate inetd process -- details below. This doesn't seem to make any difference, the UDP port still gets dropped. This feels like an Inetd problem ... but I'm not sure so I'll include the amanda version info too. Details: mildred# ls -ald /var/db/pkg/amanda* drwxr-xr-x 2 root wheel 512 May 31 2001 /var/db/pkg/amanda24-client-2.4.2p2_1 mildred# uname -a FreeBSD mildred.xxxx.com 4.6.1-RELEASE-p10 FreeBSD 4.6.1-RELEASE-p10 #1: Tue Aug 6 17:28:14 CDT 2002 dlacroix@mildred.xxxx.com:/tmp/obj/buroot/src/sys/MILDRED i386 sysctl -a |grep forwarding net.inet.ip.forwarding: 0 net.inet.ip.fastforwarding: 0 mildred# ifconfig -a xl0: flags=8843 mtu 1500 options=3 inet xxx.xxx.xxx.26 netmask 0xffffff00 broadcast xxx.xxx.xxx.255 inet xxx.xxx.xxx.27 netmask 0xffffffff broadcast xxx.xxx.xxx.27 inet xxx.xxx.xxx.29 netmask 0xffffffff broadcast xxx.xxx.xxx.29 inet xxx.xxx.xxx.30 netmask 0xffffffff broadcast xxx.xxx.xxx.30 inet xxx.xxx.xxx.2 netmask 0xffffffff broadcast xxx.xxx.xxx.2 ether 00:xx:xx:xx:xx:xx media: Ethernet autoselect (100baseTX) status: active lo0: flags=8049 mtu 16384 inet 127.0.0.1 netmask 0xff000000 faith0: flags=8002 mtu 1500 tun0: flags=8051 mtu 1500 inet 10.168.5.2 --> 10.168.5.131 netmask 0xffffffff Opened by PID 9015 I've got 2 instances of inetd -- (one for each interface) on the "private" interface (actually a ppp over ssh tunnel) I've got the following as the ONLY non-comment line in /etc/inetd_amanda mildred# egrep -v '^#' /etc/inetd.10.conf amanda dgram udp wait amanda /usr/local/libexec/amanda/amandad amandad This inetd is started from a script in /usr/local/etc/rc.d similar to this: mildred# cat /usr/local/etc/inetd.sh #!/bin/sh if [ -x /usr/sbin/inetd -a -f /etc/inetd.10.conf ]; then /usr/sbin/inetd -wW -a 10.168.5.2 -p /var/run/inetd_amanda /etc/inetd.10.conf && echo -n ' local_inetd' fi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message