Re: Replies arrive before originals (was:Re: Strange error message when sending)

[Syafril Hermansyah]

Hello Marck D. Pearlstone,

Responding  to  your  article on Sunday, August 13, 2000 at 12:21:19 GMT
+0100 (which was 13/08/2000 18:21 GMT +0700 my Local Time) :

[ ... ]

MDP> I  agree that this is not always the case. The "retry queue" has had a
MDP> mention  in  this thread and I can state what this is here and now. If
MDP> all  MX  servers (see above) fail to respond after a fixed retry count
MDP> (x  times  for  each  server in the MX list for the domain - I have my
MDP> server  set  at  three, I think the default is four) the message fails
MDP> the immediate delivery and is placed in the retry queue.

The  server  will retry sending in first 4 hours, and every attempt will
try  8  times  per  MX.  It depend on the traffic though, sometimes in 4
hours  the server will retry 10 times , sometimes more and sometime less
than 4 (the server will check delivery schedule every 1 minutes).

MDP> This  queue is processed at regular intervals - usually four hourly
MDP> ...  I don't know what Syafril has defaulted ours to - for around 4
MDP> days.

It's  depend  on  error messages when trying to deliver, but normally if
after  4  hours the message can't be deliver it will going to RetryQueue
in  this  list  (will  stack  there for 3-days, trying delivery for next
days) or going to Relay Host for further delivery attempt.

MDP> After  four  days,  the server will inform the list manager (in the
MDP> case  of  list  traffic) of the permanent delivery failure. This is
MDP> the  retry  queue  logic.  It doesn't apply to the example you have
MDP> given  -  there  is no appreciable delay between the dutaint server
MDP> receiving  the  original message and the relay servers on the other
MDP> side accepting the message for forward delivery.

Correct, thank for the explanation.

-- 
Best regards,

- Syafril -


Name  : Syafril Hermansyah  |Company: Duta Integrasi Pratama 
Mailto: <[EMAIL PROTECTED]> |Voice  : (62) (21) 385-1600  
FAXto : (62)(21)351-9241 key:000FAX |URL: 


Using The Bat! 1.46 Beta/3
under Windows NT 5.0 Build 2195 

Created : Monday, August 14, 2000, 13:48:45 GMT +0700

-- 
--
View the TBUDL archive at http://tbudl.thebat.dutaint.com
To send a message to the list moderation team double click here:
   
To Unsubscribe from TBUDL, double click here and send the message:
   
--

You are subscribed as : archive@jab.org

Re: Replies arrive before originals (was:Re: Strange error message when sending)

[Thomas Fernandez]

Hallo Mark,

On Sun, 13 Aug 2000 13:27:23 +0100 GMT (13/08/2000, 20:27 +0800 GMT),
Mark R Harding wrote:

MRH> In case anyone is interested, "Dimension 4" is available here...

MRH>   http://www.accessone.com/~thinkman/dimension4/

I use AboutTime. I used anacron for a while but was unhappy - forgot
the reason. Anyway, there are many free time synchronisers at Yahoo,
category Computing & Internet; search for "time synchro". It'll give
you some choices.

-- 

Cheers,
Thomas mailto:[EMAIL PROTECTED]

Message reply created with The Bat! 1.46 Beta/3
under Chinese Windows 98 4.10 Build 1998 
using an Intel Celeron 366Mhz, 128MB RAM



-- 
--
View the TBUDL archive at http://tbudl.thebat.dutaint.com
To send a message to the list moderation team double click here:
   
To Unsubscribe from TBUDL, double click here and send the message:
   
--

You are subscribed as : archive@jab.org

Re: Replies arrive before originals (was:Re: Strange error message when sending)

[Thomas Fernandez]

Hallo Marck,

On Sun, 13 Aug 2000 12:46:39 +0100 GMT (13/08/2000, 19:46 +0800 GMT),
Marck D. Pearlstone wrote:

[...]
MRH>> Received: from mango.callnet0800.com [212.67.144.19] by smtp.callnet0800.com
MRH>>   (SMTPD32-5.05) id A17A1222021E; Sun, 13 Aug 2000 12:07:38 +0100

MDP> This transit took only two seconds.

[and so on]

I got the message from Mark two minutes after he sent it. I was just
typing the analyses when you message come in, so I'll everybody the
repetition. ;-)

MDP> In truth, time difference in seconds can't be really be relied upon. I
MDP> once  saw someone sneering at PC clock synchronizers on this list but,
MDP> IMHO,  such  things  are a necessity for clarity and to ensure that we
MDP> are  all  talking  the same language. I use one which runs on the main
MDP> comms  machine  here  on  my LAN and knows how to get the current real
MDP> time from a time server. It will do this at one minute intervals while
MDP> I  am  connected  to  an ISP and I dial out at least every 15 minutes.

Geez, if a PC clock is so far off real time, I think someone needs to
invent a new oszillator! ;-) I synchronise about once a week.

Anyway, someone's "sent time" being off by six weeks (!) has nothing
to do with synchronisation or lack thereof, he just hasn't figured out
yet how to set the time on his PC. Doesn't anybody know a newsgroup we
can recommend him for that?

(I'll duck for the trouts, which will be coming from four directions
on three continents now. I say, "save the trouts!")

-- 

Cheers,
Thomas mailto:[EMAIL PROTECTED]

Message reply created with The Bat! 1.46 Beta/3
under Chinese Windows 98 4.10 Build 1998 
using an Intel Celeron 366Mhz, 128MB RAM



-- 
--
View the TBUDL archive at http://tbudl.thebat.dutaint.com
To send a message to the list moderation team double click here:
   
To Unsubscribe from TBUDL, double click here and send the message:
   
--

You are subscribed as : archive@jab.org

Re: Replies arrive before originals (was:Re: Strange error message when sending)

[Marck D. Pearlstone]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Hi Mark,

On 13 August 2000 at 12:07:35 GMT +0100 (which was 12:07 where I
live) [EMAIL PROTECTED] wrote and made these points on the subject
of "Replies arrive before originals (was:Re: Strange error message when sending)":

MRH> ... my assumption from examining headers in the past is that some
MRH> ISPs servers are overloaded on occasion and take a while to
MRH> distribute email.

Correct.

MRH> For interests sake - this message is sent from home (via my
MRH> server, the ISP relay and outwards...) so it'll be interesting to
MRH> see whether there are problems in timing with this message too?

MRH> Received: from  by silverstones.com [127.0.0.1]
MRH> with DomainPOP (MDaemon.v3.5.0c.R)
MRH> for <[EMAIL PROTECTED]>; Sun, 13 Aug 2000 12:22:08 +0100

That  took  ten  minutes  to  arrive ... but that's an issue of when I
picked up my domain mail from the relay server.

MRH> Received: from dutaint.com by ...
MRH> for <[EMAIL PROTECTED]>; Sun, 13 Aug 2000 18:12:27 +0700

This  is  some four minutes after it left the originating server ... I
am surprised it took that long!

MRH> Received: from bart.callnet0800.com by dutaint.com
MRH> with SMTP (MDaemon.v3.5.0c.R)
MRH> for <[EMAIL PROTECTED]>; Sun, 13 Aug 2000 18:08:07 +0700

Well,  someone's  clock is wrong! This went back in time by 17 seconds
:-).

MRH> Received: from smtp.callnet0800.com [212.67.128.145] by bart.callnet0800.com with 
ESMTP
MRH>   (SMTPD32-5.05) id A1FE5FEF0140; Sun, 13 Aug 2000 12:09:50 +0100

Two  minutes and twenty second after it was received (but I think that
bart.callnet0800.com has its' clock set slightly fast).

MRH> Received: from mango.callnet0800.com [212.67.144.19] by smtp.callnet0800.com
MRH>   (SMTPD32-5.05) id A17A1222021E; Sun, 13 Aug 2000 12:07:38 +0100

This transit took only two seconds.

MRH> Received: from localhost ([127.0.0.1])
MRH> by mango.callnet0800.com (MERAK 2.10.290) with ESMTP id FE65DCF0
MRH> for <[EMAIL PROTECTED]>; Sun, 13 Aug 2000 12:07:36 +0100

And this one only one second from the time of creation.

MRH> Date: Sun, 13 Aug 2000 12:07:35 +0100

In truth, time difference in seconds can't be really be relied upon. I
once  saw someone sneering at PC clock synchronizers on this list but,
IMHO,  such  things  are a necessity for clarity and to ensure that we
are  all  talking  the same language. I use one which runs on the main
comms  machine  here  on  my LAN and knows how to get the current real
time from a time server. It will do this at one minute intervals while
I  am  connected  to  an ISP and I dial out at least every 15 minutes.
This  software  also provides time server services for the LAN and all
other  machines  here  look  to  it  for an update at around 30 minute
intervals.  A  PC  clock  will  drift  by a few seconds every day. The
practices  I  observe here mean that no PC in the house is out by more
than  a  second  at any time (except in the case of prolonged dial-out
failure).

- --
Cheers,
.\\arck

Marck D. Pearlstone, Consultant Software Engineer
Moderator TBUDL / TBBETA
www: http://www.silverstones.com
PGP key: 
*---
| Using The Bat! 1.46 Beta/3 S/N 14F4B4B2
| under Windows 98 4.10 Build 1998
*---

-BEGIN PGP SIGNATURE-
Version: PGP 6.0.2i

iQA/AwUBOZaKnznkJKuSnc2gEQJskACcDZBj5xIEAVbHwGd4V3YKOJevXiwAn3k2
H4D9ZvCvWT8U2a7uqVQ6AZwT
=O0Wm
-END PGP SIGNATURE-

-- 
--
View the TBUDL archive at http://tbudl.thebat.dutaint.com
To send a message to the list moderation team double click here:
   
To Unsubscribe from TBUDL, double click here and send the message:
   
--

You are subscribed as : archive@jab.org

Re: Replies arrive before originals (was:Re: Strange error message when sending)

[Marck D. Pearlstone]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Hi Gerd,

On 13 August 2000 at 11:45:42 GMT +0200 (which was 10:45 where I
live) [EMAIL PROTECTED] wrote and made these points on the subject
of  "Replies  arrive  before  originals (was:Re: Strange error message
when sending)":

GE> I  can  give  some details to this problem: it seemes as if it has
GE> sometimes  to  do  with Mark R. Harding's msgs (Sorry, no offence,
GE> Mark).

This  has  nothing  to  do  with  Mark's account. It is to do with the
routing  used  to  reach  your  domain,  which  seems to be variable -
possibly  because there are multiple MX records for the domain and the
message  gets  routed  to the first domain MX server which responds in
real  time.  The  server  will traverse the list of MX servers for the
domain  in  order  of priority until it finds one that will accept the
mail. It seems that one of them is more than slightly sluggish.

GE> I  sent  these  examples  OFF-LIST to Curtis (I was't sure whether
GE> this topic is of everyone's interest). He answered:

>>> Any idea ?

ACM>> Nope. Maybe Syafril can help.

GE> Syafril, can you ??

I can tell you from these *exactly* where the blockage occurred:

GE> +++ COPY 1

GE> Received: from mailgate2.cinetic.de by mx03.web.de with smtp
GE> (freemail 4.1.0.0 #62) id m13NWN5-00GZUqa; Sat, 12 Aug 2000 10:10 +0200
^
Right here - this relay point is ten hours *after* it arrived 

GE> Received: from relay.xlink.net (relay.xlink.net [193.141.40.4])
GE> by mailgate2.cinetic.de (8.9.3/8.9.3/SuSE Linux 8.9.3-0.1) with ESMTP id 
AAA02888
GE> for <[EMAIL PROTECTED]>; Sat, 12 Aug 2000 00:12:46 +0200
  
Here  at  relay.xlink.net,  which  had  already sat on the message for
two and a half hours since receiving it from the dutaint server ...

GE> Received: from dutaint.com (ns2.dutaint.com [203.130.233.13])
GE> by relay.xlink.net (8.9.3/8.8.7) with ESMTP id VAA01593
GE> for <[EMAIL PROTECTED]>; Fri, 11 Aug 2000 21:38:30 +0200 (MET DST)
  
At  this  time.  Of  course,  this  is assuming that the clocks on the
servers involved are correct.

GE> Received: from postbox.ee.ed.ac.uk by dutaint.com
GE> with SMTP (MDaemon.v3.5.0c.R)
GE> for <[EMAIL PROTECTED]>; Sat, 12 Aug 2000 02:36:02 +0700
  
Take  away 7 and add 2 to give a time of 21:36 in MET DST +0002 format
which states that relay.xlink.net got the message a mere *two minutes*
after it left Mark's site...

GE> Received: from pcv08.ee.ed.ac.uk (pcv08 [129.215.232.118])
GE> by postbox.ee.ed.ac.uk (8.9.1/8.9.1) with ESMTP id UAA05221
GE> for <[EMAIL PROTECTED]>; Fri, 11 Aug 2000 20:35:45 +0100 (BST)

Which was only one minute after he sent it.

GE> ++ COPY 2

Now,  this  message  underwent  a  *completely*  different routing and
relay.xlink.net was not involved giving us the following timings...

GE> Received: from dutaint.com by mx04.web.de with smtp
GE> (freemail 4.1.0.0 #61) id m13NKqh-006qz0a; Fri, 11 Aug 2000 21:52 +0200
^

This is a mere two minutes after it left the dutaint server ...

GE> Received: from teleute.rpglink.com by dutaint.com
GE> with SMTP (MDaemon.v3.5.0c.R)
GE> for <[EMAIL PROTECTED]>; Sat, 12 Aug 2000 02:50:01 +0700
  
Take  away 7 and add 2 to give a time of 21:50 in MET DST +0002 format
as the time that it arrived at dutaint from  rpgliink...

GE> Received: from antelope.it.earthlink.net [207.217.90.52]
GE> by teleute.rpglink.com with esmtp (Exim 3.12 #1 (Debian))
GE> id 13NKjN-0004sG-00; Fri, 11 Aug 2000 12:44:37 -0700

Which is six minutes after rpglink got it in the first place.

IOW: I'd have a word with your ISP.

I  agree that this is not always the case. The "retry queue" has had a
mention  in  this thread and I can state what this is here and now. If
all  MX  servers (see above) fail to respond after a fixed retry count
(x  times  for  each  server in the MX list for the domain - I have my
server  set  at  three, I think the default is four) the message fails
the immediate delivery and is placed in the retry queue. This queue is
processed  at regular intervals - usually four hourly ... I don't know
what  Syafril  has  defaulted  ours to - for around 4 days. After four
days,  the  server  will  inform the list manager (in the case of list
traffic)  of  the  permanent delivery failure. This is the retry queue
logic.  It  doesn't  apply to the example you have given - there is no
appreciable  delay  between  the dutaint server receiving the original
message  and the relay servers on the other side accepting the message
for forw

Re: Replies arrive before originals (was:Re: Strange error message when sending)

[Mark R Harding]

Gerd,

Regarding your message dated: 13 August 2000...

GE> I can give some details to this problem: it seemes as if it has sometimes to do
GE> with Mark R. Harding's msgs (Sorry, no offence, Mark).

No offence taken!  I have an idea though although I'm not sure how it
explains the problem ...

The reply you quoted from me was sent from my machine at work.  Email
from this machine gets routed onto the internet via the SMTP server at
work which is very well run and normally extremely quick.

When I work from home (dial-up), or communicate with friends who use
dial-up connections and their ISP's servers I often notice that
messages start running out-of-sync and my assumption from examining
headers in the past is that some ISPs servers are overloaded on
occasion and take a while to distribute email.

I also have a free
email address at 'email.com' which is setup to directly forward all
email to my work account and if I mail myself (for test purposes only
of course - I promise I'm not that mad!) I notice that it can regularly
take a day or two to get the message back to me. (i.e. - it's hopeless
and I don't use it any more.)

Really, I don't know if factors such as these could be affecting how
you get you mail deliveries out-of-order (logically it doesn't seem to
make sense) but this does happen to a number of people I know and
usually there is a 'free email' service involved somewhere along the
line.

GE> I'm sure there are some more examples (I'm not sure, that /always/ Mark's
GE> message is affected), but I think this will do.

Generally for me, on TBUDL/TBBETA this effect is rare (barring the
other type of out-of-sync messages such as the messages recently from
Starc whose emails all seem to be dated mid-June time and therefore
appear way down near the bottom of the pile due to
sort-by-creation-date settings.) but it does happen a lot on the only
other discussion list I'm on.

There the readership is more diverse in client/platform where some
subscribers use PC's, some macs, some OnMail (digital TV email) and so
on with a whole host of clients on each platform. With clocks not
always accurate and differences in platforms adherring to standards
it's a wonder any of the output is ledgible at all!

For interests sake - this message is sent from home (via my server,
the ISP relay and outwards...) so it'll be interesting to see whether
there are problems in timing with this message too?

Best wishes

Mark


-- 
-
 Using TheBat! 1.45 S/MIME Windows NT 4 0 1381 Service Pack 6
-
 Just4Fun - Freestuff, Humour and More! - http://just4fun.ipfox.com/
-
 Mark R Harding
 The Integrated Systems Group (Vision)
 Department of Electronics & Electrical Engineering
 The University of Edinburgh, King's Buildings, Mayfield Road
 Edinburgh. EH9 3JL. Scotland. U.K.

 Phone:  +44 (0)131 650 5662
 Fax:+44 (0)131 650 6554
 Email:  mailto:[EMAIL PROTECTED]
 URL:http://www.ee.ed.ac.uk/~mrh/
-
 PGP-Key Available at:  http://www.ee.ed.ac.uk/~mrh/pgp.html
-


-- 
--
View the TBUDL archive at http://tbudl.thebat.dutaint.com
To send a message to the list moderation team double click here:
   
To Unsubscribe from TBUDL, double click here and send the message:
   
--

You are subscribed as : archive@jab.org

Re: Replies arrive before originals (was:Re: Strange error message when sending)

[Chuck Mattsen]

On Saturday, August 12, 2000 at 11:45 AM or thereabouts, Thomas
Fernandez wrote the following about Replies arrive before originals
(was:Re: Strange error message when sending):

Thomas> On Sat, 12 Aug 2000 15:53:11 +0200 GMT (12/08/2000, 21:53
Thomas> +0800 GMT), Gerd Ewald wrote:

GE>> I know what you mean: the last few days I receive replies before
GE>> the original msg is transmitted ! Sometimes I have to wait two
GE>> days for the original message.

GE>> Is it only me having that problem or is it a ML-server problem ?

Case in point, Gerd, I only received your message, just now,
second-hand via Thomas' reply; I've yet to receive your original post.
It's been happening to me, only with this list out of many high-volume
lists, for either two or three days. Despite what others may suggest
regarding your ISP, I suspect the problem is *not* on the listmember
end of things.

Chuck
-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Chuck Mattsen[EMAIL PROTECTED]
 http://www.users.uswest.net/~mattsen
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Random Thought/Quote for this Message:
 There is no finish line.

-- 
--
View the TBUDL archive at http://tbudl.thebat.dutaint.com
To send a message to the list moderation team double click here:
   
To Unsubscribe from TBUDL, double click here and send the message:
   
--

You are subscribed as : archive@jab.org