Pegasus Mail & Mercury

Dear all

I have a question about relaying. As I understand it, relaying is achieved by using more than one SMTP server to send/forward mail to its final destination.

We are using the Mercury E module to send mail which tries to connect directly to the destination SMTP server and then delivers the mail.

During the last couple of days, from one address, we are getting 550 relaying denied error messages.

Here is the Mercury E log:

15:31:12.234: --- Thu May 08 15:31:12 2008 ---
15:31:12.234: Connect to '62.172.185.25', timeout 60.
15:32:13.234: 8: Socket read timeout
15:32:13.234: --- Connection closed normally at Thu May 08 15:32:13 2008. ---
15:32:13.234:
15:32:13.250: --- Thu May 08 15:32:13 2008 ---
15:32:13.250: Connect to '62.172.185.42', timeout 60.
15:33:14.250: 8: Socket read timeout
15:33:14.250: --- Connection closed normally at Thu May 08 15:33:14 2008. ---
15:33:14.250:
15:33:14.250: --- Thu May 08 15:33:14 2008 ---
15:33:14.250: Connect to '80.6.93.157', timeout 60.
15:33:15.250: >> 220 THIS IS A MAIL SERVER<cr><lf>
15:33:15.250: << EHLO apsarchaeology.co.uk<cr><lf>
15:33:15.312: >> 250 Requested mail action okay, completed.<cr><lf>
15:33:15.312: << MAIL FROM:<greenman@lincsheritage.org><cr><lf>
15:33:15.359: >> 250 Requested mail action okay, completed.<cr><lf>
15:33:15.359: << RCPT TO:<blueman@lincolnshire.gov.uk><cr><lf>
15:33:15.406: >> 550 Relaying denied.<cr><lf>
15:33:15.406: << QUIT<cr><lf>
15:33:15.437: >> 221 Closing connection.<cr><lf>
15:33:15.437: --- Connection closed normally at Thu May 08 15:33:15 2008. ---
15:33:15.437:

I don't understand where the relaying is coming into it. Is it due to the lincsheritage.org mail being sent from apsarchaeology.co.uk? If so, then I thought that this had been solved when we configured the RDNS to resolve apsarchaeology.co.uk to our permanent IP address and back again.

Previously, mail to this address has been accepted (for years). Can anyone shed any light on what is happening here please?

Many thanks.

The domain lincolnshire.gov.uk has 3 MX records. The first 2 failed to answer. The 3rd did answer, although with a somewhat unorthodox greeting phrase. It did however not think mail to lincolnshire.gov.uk should be handled by the server, and issued an error message saying relaying denied.

I get exactly the same results when telneting from here to these 3 servers. I think lincolnshire.gov.uk may have a serious problem with their mail servers at present.

/Rolf 

Rolf, thanks a lot for your reply. Thanks for the clarification.

When I saw the Greeting Message, I thought it a little weird.

So, is the 'relaying' derived from the fact that the first two connections failed, and it regards this 'trail' as a 'source route'?

Cheers!

This is caused by that server treating 'lincolnshire.gov.uk' as a non-local domain even though it is listed in the MX records for that domain.

In other words their setup is borked.

It has nothing to do with where you tried to connect beforehand. 

dilberts_left_nut:

This is caused by that server treating 'lincolnshire.gov.uk' as a non-local domain even though it is listed in the MX records for that domain.

In other words their setup is borked.

It has nothing to do with where you tried to connect beforehand. 

Thanks. I understand in principle what you are saying.

So, is the server that is handling the request misconfigured? Or, is the SMTP server in the wrong domain/sub-domain within their forest (assuming they use WS2k3 AD)? Or can it be caused by misconfiguration of software?

I'm wondering how this could be done. If I was to try to generate the same error message from our copy of Mercury32 so that anyone trying to send mail to our domains got a 550 Relaying Denied response, how would I achieve it?

I'm still learning about this so if anyone would like to help out by explaining this I'd appreciate it. I've worked my way through the first part of RFC 2821 and now understand most of the elementary commands used by senders/receivers, but the finer points are still a black art :)

Cheers!

I'm wondering how this could be done. If I was to try to generate the same error message from our copy of Mercury32 so that anyone trying to send mail to our domains got a 550 Relaying Denied response, how would I achieve it?

Simply remove the domain from the domains list.  Could not get any mail for that domain though since it would all be bounced. 

Ah! The penny has dropped :)

So, as said, the setup is wrong and their mail server does not recognise the RCPT data as being local. I understand.

Thanks very much!