NetBSD Problem Report #165

From gnats  Wed Mar  9 21:35:51 1994
Received: from bloom-beacon.mit.edu (BLOOM-BEACON.MIT.EDU [18.70.0.232]) by sun-lamp.cs.berkeley.edu (8.6.5/8.6.5) with SMTP id VAA10921 for <gnats-bugs@sun-lamp.cs.berkeley.edu>; Wed, 9 Mar 1994 21:35:50 -0800
Message-Id: <199403100533.AAA00439@orchard.medford.ma.us>
Date: Thu, 10 Mar 1994 00:33:31 -0500
From: sommerfeld@orchard.medford.ma.us
Reply-To: sommerfeld@orchard.medford.ma.us
To: gnats-bugs@sun-lamp.cs.berkeley.edu
Subject: sendmail could lose mail on hung nameserver.
X-Send-Pr-Version: 3.01.6

>Number:         165
>Category:       bin
>Synopsis:       sendmail could lose mail on hung nameserver.
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    gnats-admin
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Wed Mar 09 21:50:02 +0000 1994
>Closed-Date:    Wed Mar 16 18:37:04 +0000 1994
>Last-Modified:  
>Originator:     Bill Sommerfeld
>Release:        
>Organization:
>Environment:
	seen in sendmail 8.6.4; probably in 8.6.5, too..
System: NetBSD orchard.medford.ma.us 0.9a ORCHARD#3 i386

>Description:
	sendmail apparantly doesn't finish writing the queue file
	(thus committing the message into the queue) until
	after it resolves the names of the recipients.
	if named is hung, this could take a while (several minutes);
	in the interim, the system could crash (or be shut down) and
	the message is effectively lost..

>How-To-Repeat:
	unplug your system from the net
	kill & restart named (to clear the cache).
	mail -v someuser@some.other.host << EOF
	foo
	bar
	EOF 
	# note that it sits there for a while
	^Z
	mailq
	# note that the message shows up as "no queue file".

>Fix:
	Unknown at present time; probably should queue up message
	before attempting to do any action which could trigger a
	name resolution call.
>Release-Note:
>Audit-Trail:

State-Changed-From-To: open->closed
State-Changed-By: jtc
State-Changed-When: Wed Mar 16 18:37:04 1994
State-Changed-Why:
Received following from Eric Allman, sendmail's author:

From: Eric Allman <eric@CS.Berkeley.EDU>
Date: Wed, 09 Mar 1994 22:26:35 -0800

I don't believe this is correct -- or rather, it is only half correct.
It's true that the delay can be long if the nameserver is down, but
sendmail hasn't "accepted" the message until the 250 message at the
end of the DATA transaction (in SMTP) or it returns (in collect-from-
standard-input mode).  The problem is that it is trying to give you
an error message as soon as possible; this is in conflict with trying
to return as soon as possible.  There is no "right" answer.  I'll
consider trying to modify this for a future release, but for the time
being, I don't consider this a "bug" in the truest sense.

eric

Unformatted:
>Unformatted:

NetBSD Home
NetBSD PR Database Search

(Contact us) $NetBSD: query-full-pr,v 1.51 2026/08/10 02:28:17 riastradh Exp $
$NetBSD: gnats_config.sh,v 1.10 2026/05/13 22:00:09 riastradh Exp $
Copyright © 1994-2026 The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.