illegal Via: field
Adam Bentley
are coming back with the error below. It seems that for some reason a
blank 'Via:' field is creeping in somewhere.....
I don't know whether this is at our end or not....
fyi,
our janet link is on system k (cc_sysk) so all external mail
is passed there for processing.
We are running uk-sendmail 2.1 on a Sequent Symmetry S81/Dynix 3.1.2
From ro...@uk.ac.coventry.cck Tue Nov 5 14:35:21 1991
Received: from cc_sysk by cch.coventry.ac.uk; Tue, 5 Nov 91 14:25:01 GMT
From: Admin <ro...@uk.ac.coventry.cck>
Date: Tue, 5 Nov 91 14:26:11 GMT
Message-Id: <8715.91...@cck.coventry.ac.uk>
To: ro...@uk.ac.coventry.cck
Subject: X25 Xfer status of /usr/spool/x25/M007359
Status: OR
Mail (id 093027) to nsfnet-relay (000040010153/FTP.MAIL)
was aborted after 5 attempts
file called '/usr/spool/x25/M007359'
because 'PP hard error message: Illegal Via: field, No ';' ""'
info 'UNIX-NIFTP $Revision: 5.5 $'
info 'Mail not delivered, bad remote user name or host name'
Full text of rejected letter now follows:
|>>> listserv%uga.cc....@uk.ac.nsfnet-relay
|>>>
|>>> From: Shaozhong Shi <gey...@uk.ac.coventry.cck>
|>>> Date: Tue, 5 Nov 91 13:55:10 GMT
|>>> Message-Id: <24031.91...@cck.coventry.ac.uk>
|>>> Received:
|>>> Via:
|>>> To: list...@edu.uga.cc.uga
|>>>
|>>> SIGNOFF MAPS-L
--
_
/-\dam
FLESH: Adam Bentley (Fraggle), Programmer/Adviser, Systems, Coventry Poly. UK
INET : ad...@cck.cov.ac.uk
Ian Crorie
| Can anyone give any pointers as to why some messages sent via nsfnet-relay
| are coming back with the error below. It seems that for some reason a
| blank 'Via:' field is creeping in somewhere.....
| I don't know whether this is at our end or not....
Well I can't help but I sure would like to know the answer since it has
been happening at hw.cs for a couple of weeks now as well (and once or
twice in the 2 months before that but the number of bounced mail
messages is increasing).
The problem is intermittent in that when resent the mail gets through.
I've looked through our sendmail 2.1 stuff and through our niftp code
and I can't find any sign that anything at our site adds the Via:
field.
--- Support Desk:
Two, four, six, eight
Who do we appreciate?
DESK!
---
The Postmaster
I _believe_ it's caused by a known bug in PP, for which XTEL have a patch.
Ask Julian Onions (J.On...@uk.co.xtel)
Alan
--
+----------------------------------------------------------------------------+
| LANCASTER UNIVERSITY POSTMASTER Alan Phillips |
|----------------------------------------------------------------------------|
| Computer Centre, Lancaster University, Lancaster LA1 4YW, UK |
| Phone: 0524-65201 x 3672 |
| |
| "Where am I?" |
| "In The Village...." |
|----------------------------------------------------------------------------|
|JANET :postm...@uk.ac.lancaster | INTERNET: postm...@lancaster.ac.uk |
|BITNET:postm...@lancaster.ac.uk | UUCP : postmaster%lancaster@uknet |
+----------------------------------------------------------------------------+
Andrew J Cole
|The easiest workaround for this problem is to persuade nsfnet-relay
|to stop complaining about blank Via fields. I assume they have only
|recently started rejecting mail like this as sendmail has had this
|feature for years.
Long ago I fixed this one using a conditional on 's'. We saw the problem
mostly with empty Received: lines and very infrequently with Via: (too
heavily loaded relaying incoming traffic to accept incoming SMTP!) its
not down to a recent change at nsfnet-relay.
#H?R?Received: $?sfrom $s by $j; $b$.
#H?V?Via: $?s$s; $b$.
H?R?Received: $?sfrom $s by $j; $b$|from $j by $j; $b$.
H?V?Via: $?s$s; $b$|$j; $b$.
Andrew
Jim Crammond
a user mail program) and gets queued on the local machine rather than
immediately sent (and thus it is sent when the sendmail daemon processes the
mail queue).
Mail gets queued because (a) the load average was too high, or
(b) if forwarding with SMTP (ether/tcp channels) sendmail was unable
to connect to the remote host on the first attempt.
The easiest workaround for this problem is to persuade nsfnet-relay
to stop complaining about blank Via fields. I assume they have only
recently started rejecting mail like this as sendmail has had this
feature for years.
-Jim.
Khalid Sattar
|>>>> Jim Crammond <ji...@quintus.com> writes:
> The easiest workaround for this problem is to persuade nsfnet-relay
> to stop complaining about blank Via fields. I assume they have only
> recently started rejecting mail like this as sendmail has had this
> feature for years.
I have seen this problem for atleast a year, maybe more, but could not
accurately determine the cause or recreate the problem.
In our case we only have a single uucp connection, would removing the
HVia: $?S$S; $b$.
line from our sendmail configuration help? since Via field is set by
janet mailers only???
----------------------------------------------------------------------
Khalid Sattar
Systems Manager
Harlequin Ltd, Barrington Hall, Barington, Cambridge CB2 5RG
Email: kha...@harlqn.co.uk, kha...@harlqn.uucp
Richard Hopkins
very much nsfnet-relay problems, introduced with the changeover to PP.
(This is in a purely Grey Book context, not SMTP.) I've examined the
outbound messages locally and found that messages are rejected in this
way, from time to time (not all the time) when the Via: field is absent.
Richard Hopkins
Andrew Findlay
|very much nsfnet-relay problems, introduced with the changeover to PP.
I believe this is now fixed. Certainly a bug in PP has been found that
caused exactly this problem, and a patch was issued a few weeks back.
The problem seemed to be that PP was remembering a fault condition
from one message to the next, so several messages got rejected as a
result of one bad one on busy machines.
|Richard Hopkins
Andrew
--
---------------------------------------------------------------------
| From Andrew Findlay at Brunel University, Uxbridge, UB8 3PH, UK |
| Andrew....@brunel.ac.uk phone: +44 895 274000 x2512 |
---------------------------------------------------------------------
Ian Dickinson
|Jim writes:
|>I've satisfied myself that the complaints about illegal Via: fields are
|>very much nsfnet-relay problems, introduced with the changeover to PP.
It occurs to other PP sites over Grey Book too.
|I believe this is now fixed. Certainly a bug in PP has been found that
|caused exactly this problem, and a patch was issued a few weeks back.
|The problem seemed to be that PP was remembering a fault condition
|from one message to the next, so several messages got rejected as a
|result of one bad one on busy machines.
Yes, this has solved the unnecessary multiple rejection problem.
But it hasn't solved the illegal Via: rejections entirely.
The Via fields are illegal, and PP is quite a pedantic mailer.
How do I make sure that sendmail doesn't output bad Via: headers
in the first place, when local queuing occurs?
(The real solution was to move to SMTP, pending X.400)
Cheers,
--
\/ato
Ian Dickinson - NIC handle: ID17
va...@csv.warwick.ac.uk ...!mcsun!uknet!warwick!vato
@c=GB@o=University of Warwick@ou=Computing Services@cn=Ian Dickinson