clewis@eci386.uucp (Chris Lewis) (08/19/89)
I'm not sure if this is the right place, but the alternative was to cross
post all over the place.
I run a small mailing list, about 50 names. I rely heavily on pathalias
on eci386 to route my outgoing articles because paths change quite a bit,
and so far I haven't been able to find enough info to truly understand
how the MTA's work and "proper" ways to steer thru some of the gateways.
I get one or two bounces routinely from each issue, but usually from
MTA's that are capable of telling you what address bounced. Occasionally
not. Sigh. I usually keep bouncing til somebody says something different.
And occasionally, I get one like the one included below that demonstrates
TWO severe problems...
- MTA not telling you what the failing address was.
- rabid-rerouter looping.
The ">" denotes the headers thrown on by the MTA that finally gave up
in disgust. The "<" denotes headers thrown on by the mailers along
the article's wanderings.
And frankly, I don't have a clue which one of my addresses caused this
bounce. I have 5 or so that route to rutgers. None of them are
listed as going to psuvax1. (my route to psuvax1 doesn't go thru rutgers).
There's at least *two* rabid rerouters causing this:
- rutgers for the initial random divergence
- one of psuvax1, burdvax, cbmvax for firing it back at uunet
- one of burdvax, cbmvax or vrdxhq for continuing the problem.
- and possibly uunet
The route that this thing seemed to take was:
telly!attcan!uunet!rutgers!psuvax1!burdvax!cbmvax!uunet!vrdxhq.verdix.com!
burdvax!cbmvax!uunet!vrdxhq.verdix.com!burdvax
(There may be sites in there that don't insert Received headers. Too
bad burdvax didn't keep a copy of the "From_" line and the rmail command
line)
I guess you can see the problem. Someone asked if I had ever seen a loop.
Well I found one.... I just wish I knew what the address was that caused
this. Actually, in a way it's sort of sad that one of the mailers caught
it on the site count - t'would be nice to be remembered in posterity
as a set of bits randomly flying around the country...
These are the "possibles" for the route that I specified (extracts from
my smail log - addresses after pathalias resolution):
telly!attcan!uunet!rutgers!isis.wv.tek.com!<user>
telly!attcan!uunet!rutgers!topaz.rutgers.edu!<user>
telly!attcan!uunet!rutgers!rochester!ur-tut!<user>
telly!attcan!uunet!rutgers!rochester!kodak!aviary!<user>
telly!attcan!uunet!rutgers!pedsga.tinton.ccur.com!<user>
> From lsuc!utai!PRC.Unisys.COM!MAILER-DAEMON Fri Aug 18 13:04:38 1989
> Received: by eci386.UUCP (smail2.5)
> id AA26506; 18 Aug 89 13:04:37 EDT (Fri)
> Received: from burdvax.PRC.Unisys.COM by neat.cs.toronto.edu with SMTP id 10641; Fri, 18 Aug 89 09:26:04 EDT
> Received: from vrdxhq.UUCP by burdvax.PRC.Unisys.COM (5.61/Domain/jpb/2.9)
> id AB05932; Fri, 18 Aug 89 09:25:43 -0400
> Date: Fri, 18 Aug 89 09:25:43 EDT
> From: MAILER-DAEMON@PRC.Unisys.COM (Mail Delivery Subsystem)
> Subject: Returned mail: Unable to deliver mail
> Message-Id: <8908181325.AB05932@burdvax.PRC.Unisys.COM>
> To: Postmaster@cs.toronto.edu
> To: lsuc!eci386!clewis@neat.ai.toronto.edu
> Status: OR
>
> ----- Transcript of session follows -----
> 554 sendall: too many hops (17 max)
>
> ----- Unsent message follows -----
< Received: from vrdxhq.UUCP by burdvax.PRC.Unisys.COM (5.61/Domain/jpb/2.9)
< id AA05932; Fri, 18 Aug 89 09:25:43 -0400
< Received: by vrdxhq.verdix.com (4.12/01-Aug-89)
< id AA01616; Fri, 18 Aug 89 09:21:48 edt
< Received: from cbmvax.UUCP by uunet.uu.net (5.61/1.14) with UUCP
< id AA22191; Fri, 18 Aug 89 07:58:39 -0400
< Received: by cbmvax.UUCP (5.57/UUCP-Project/Commodore 12/21/87))
< id AA10085; Fri, 18 Aug 89 07:56:30 EDT
< Received: from vrdxhq.UUCP by burdvax.PRC.Unisys.COM (5.61/Domain/jpb/2.9)
< id AA02024; Fri, 18 Aug 89 07:46:50 -0400
< Received: by vrdxhq.verdix.com (4.12/01-Aug-89)
< id AA29332; Fri, 18 Aug 89 07:20:56 edt
< Received: from cbmvax.UUCP by uunet.uu.net (5.61/1.14) with UUCP
< id AA20943; Fri, 18 Aug 89 03:58:39 -0400
< Received: by cbmvax.UUCP (5.57/UUCP-Project/Commodore 12/21/87))
< id AA06995; Fri, 18 Aug 89 03:54:19 EDT
< Received: from vrdxhq.UUCP by burdvax.PRC.Unisys.COM (5.61/Domain/jpb/2.9)
< id AA04955; Fri, 18 Aug 89 03:40:17 -0400
< Received: by vrdxhq.verdix.com (4.12/01-Aug-89)
< id AA21022; Fri, 18 Aug 89 03:35:13 edt
< Received: from cbmvax.UUCP by uunet.uu.net (5.61/1.14) with UUCP
< id AA10017; Fri, 18 Aug 89 02:05:50 -0400
< Received: by cbmvax.UUCP (5.57/UUCP-Project/Commodore 12/21/87))
< id AA05396; Fri, 18 Aug 89 01:54:29 EDT
< Received: from psuvax1.UUCP by burdvax.PRC.Unisys.COM (5.61/Domain/jpb/2.9)
< id AA03524; Fri, 18 Aug 89 00:12:50 -0400
< Received: by psuvax1.cs.psu.edu (5.61/PSUCS-1.0)
< id AA05025; Fri, 18 Aug 89 00:10:54 -0400
< Received: from uunet.UU.NET by rutgers.edu (5.59/SMI4.0/RU1.2/3.04)
< id AA00597; Thu, 17 Aug 89 23:45:41 EDT
< Received: from attcan.UUCP by uunet.uu.net (5.61/1.14) with UUCP
< id AA11829; Thu, 17 Aug 89 23:42:11 -0400
< Received: by telly.on.ca (smail2.5 id AA07512) 17 Aug 89 21:17:11 EDT (Thu)
< Received: by eci386.UUCP (smail2.5)
< id AA15592; 17 Aug 89 14:36:21 EDT (Thu)
< Subject: Ferret Mailing List: Issue 077
< To: ferret-list@neat.ai.toronto.edu
< Message-Id: <8908171434.AA15454@eci386.UUCP>
< Date: 17 Aug 89 14:34:41 EDT (Thu)
< From: lsuc!eci386!clewis@neat.ai.toronto.edu (Chris Lewis)
< Content-Type: text
< Content-Length: 4603
<
< [contents of article]
--
Chris Lewis, R.H. Lathwell & Associates: Elegant Communications Inc.
UUCP: {uunet!mnetor, utcsri!utzoo}!lsuc!eci386!clewis
Phone: (416)-595-5425clewis@ecicrl.UUCP (Chris Lewis) (08/20/89)
In article <1989Aug18.190929.27532@eci386.uucp> clewis@eci386 (Chris Lewis) writes: (eg: myself)... Um, I guess I was a little hard on rutgers, for looking back at the paths that smail generated, rutgers was being explicitly asked to reroute in a couple of cases: > telly!attcan!uunet!rutgers!isis.wv.tek.com!<user> > telly!attcan!uunet!rutgers!topaz.rutgers.edu!<user> > telly!attcan!uunet!rutgers!pedsga.tinton.ccur.com!<user> Sorry Mel. -- Chris Lewis, R.H. Lathwell & Associates: Elegant Communications Inc. UUCP: {uunet!mnetor, utcsri!utzoo}!lsuc!ecicrl!clewis Moderator of the Ferret Mailing List (ferret-request@eci386) Phone: (416)-294-9253
chip@ateng.com (Chip Salzenberg) (08/29/89)
Mail loops. Sigh. Rutgers: Just Say "-d".