Community Discussions and Support
Mercury config in webserver sendmail role

It seems to be an error generated directly by mercury.

The message headers indicate:

From: postmaster@[203.26.##.##]

To: admin@[203.26.##.##] 

The attached message has failed delivery and has been referred
to you as postmaster. The following error report or reports
were given to explain the problem:

*** user@address.com.au
Error connecting to primary server '210.9.177.134'.
556 Too many invalid recipient requests. Closing connection.
This is Mercury/32 message back to you reporting it could not deliver the message and the error message sent by the mail.universalmagazines.com.au server.  
 Now if there were only one RCPT TO: in this particular message then I suspect that they are storing the IP addresses of the connecting hosts that send to 
a lot of invalid addresses.
 

<blockquote><div class="ForumReplyToPostArea"> <p>It seems to be an error generated directly by mercury.</p><p>The message headers indicate:</p><p>From: postmaster@[203.26.##.##]</p><p>To: admin@[203.26.##.##]  </p><pre>The attached message has failed delivery and has been referred to you as postmaster. The following error report or reports were given to explain the problem: *** <a href="mailto:kday@universalmagazines.com.au" mce_href="mailto:kday@universalmagazines.com.au" class="moz-txt-link-abbreviated">user@address.com.au</a> Error connecting to primary server '210.9.177.134'. 556 Too many invalid recipient requests. Closing connection.</pre></div></blockquote><div class="ForumReplyToPostArea"><pre>This is Mercury/32 message back to you reporting it could not deliver the message and the error message sent by the mail.universalmagazines.com.au server. </pre><pre> Now if there were only one RCPT TO: in this particular message then I suspect that they are storing the IP addresses of the connecting hosts that send to </pre><pre>a lot of invalid addresses.</pre><pre> </pre></div><blockquote><div class="ForumReplyToPostArea"> </div></blockquote>

Hi,

A client's uses Mercury on W2K3 server to send email to website users.

Mercury does not hosts the clients mail/mailboxes, and only serves a single client, the SQL Server on localhost that creates the emails.

Mercury is working well at the moment, but there are a couple of issues, which include:

 1. The correct handling of hard and soft bounces: (hard bounces do not arrive in postmaster mailbox, soft bounces do not seem to be arriving in offsite reply to or sender mailboxes)

 2. Admin (postmaster) mailbox: (nothing seems to arrive in postmaster mailbox)

I am looking for some advice on how best to configure Mercury in this senario.

At the moment, I am just using the machines ip address rather than the clients website domain info. I have attached the relevant sections of the mercury.ini. Can anyone suggest how I should best configure mercury in this senario and how I might work out issues with admin messages.

TIA

Sean

_____________________________________________________

[General]
myname:          [203.26.xx.xx]    # Canonical name for this server

[Protocols]
MERCURYS.DLL  # SMTP Server, to receive incoming mail
MERCURYE.DLL  # end-to-end SMTP delivery client, to send outgoing mail

[Mercury]
postmaster:    admin    # NetWare UIC of postmaster
PM_notify:     1    # Do/Don't send errors to the postmaster
change_owner:  1

[MercuryE]
HELO : [203.26.xx.xx]

[MercuryS]
Debug : 1
Timeout : 30
Relay : 0
Strict_Relay : 1
Allow_Illegals : 0
SMTP_Authentication : 0
Compliance_Settings : 0
Maximum_Failed_Rcpts : 4
Max_Relay_Attempts : 4
SSL_Mode : 0
ST_Blacklisting : 288
No_VRFY : 0
SMTP_ConnFlags : 0

[Domains]
203: 203
203: 203.26.xx.xx

Hi, A client's uses Mercury on W2K3 server to send email to website users. Mercury does not hosts the clients mail/mailboxes, and only serves a single client, the SQL Server on localhost that creates the emails. Mercury is working well at the moment, but there are a couple of issues, which include:  1. The correct handling of hard and soft bounces: (hard bounces do not arrive in postmaster mailbox, soft bounces do not seem to be arriving in offsite reply to or sender mailboxes)  2. Admin (postmaster) mailbox: (nothing seems to arrive in postmaster mailbox) I am looking for some advice on how best to configure Mercury in this senario. At the moment, I am just using the machines ip address rather than the clients website domain info. I have attached the relevant sections of the mercury.ini. Can anyone suggest how I should best configure mercury in this senario and how I might work out issues with admin messages. TIA Sean _____________________________________________________ [General] myname:          [203.26.xx.xx]    # Canonical name for this server [Protocols] MERCURYS.DLL  # SMTP Server, to receive incoming mail MERCURYE.DLL  # end-to-end SMTP delivery client, to send outgoing mail [Mercury] postmaster:    admin    # NetWare UIC of postmaster PM_notify:     1    # Do/Don't send errors to the postmaster change_owner:  1 [MercuryE] HELO : [203.26.xx.xx] [MercuryS] Debug : 1 Timeout : 30 Relay : 0 Strict_Relay : 1 Allow_Illegals : 0 SMTP_Authentication : 0 Compliance_Settings : 0 Maximum_Failed_Rcpts : 4 Max_Relay_Attempts : 4 SSL_Mode : 0 ST_Blacklisting : 288 No_VRFY : 0 SMTP_ConnFlags : 0 [Domains] 203: 203 203: 203.26.xx.xx

[quote user="Sean"]

[Domains]
203: 203
203: 203.26.xx.xx

[/quote]

Try adding:

203: [203.26.xx.xx]

 

as that should be the local domain part.

I'm not sure that this will solve the problem though. [:S]

 

Is there an admin user (in Config - Manage local users)?

Is the mailbox path valid?

 

<p>[quote user="Sean"] [Domains] 203: 203 203: 203.26.xx.xx [/quote]</p><p>Try adding:</p><p>203: <b>[</b>203.26.xx.xx<b>]</b></p><p> </p><p>as that should be the local domain part. </p><p>I'm not sure that this will solve the problem though. [:S]</p><p> </p><p>Is there an admin user (in Config - Manage local users)?</p><p>Is the mailbox path valid?</p><p> </p>

It would have helped if you would have provided a real IP address.    In addition to verifying that the local postmaster account does exist and fixing the DNS entry to add the required brackets yuo should try sending a message from the outside to postmaster@[203.26.xx.xx] to verify the MercuryS can properly receive mail for the postmaster.  

Also not sure what the hard and soft bounces are unless they are refering to the 500 and 400 series messages.  If that is so and the MYSQL application is using a user@[203.26.xx.xx] type address then the fixing of the domain name should fix the bounce problem as well.

<p>It would have helped if you would have provided a real IP address.    In addition to verifying that the local postmaster account does exist and fixing the DNS entry to add the required brackets yuo should try sending a message from the outside to postmaster@[203.26.xx.xx] to verify the MercuryS can properly receive mail for the postmaster.   </p><p>Also not sure what the hard and soft bounces are unless they are refering to the 500 and 400 series messages.  If that is so and the MYSQL application is using a user@[203.26.xx.xx] type address then the fixing of the domain name should fix the bounce problem as well. </p>

Thanks,

I've added the [] sqare brackets around the ip address in the domains section.

Yes, I definitely have an admin user in: [configuration | manage local users] and yes, the mailbox path is valid.

Thomas, The SQL server is addressing mail to 127.0.0.1

Should I also add [127.0.0.1] to the domains / relevant sections of Mercury.ini ?

Finally, is there a limit on the volume of messages that mercurys will accept from a sender. For example, when the SQL scripts run, it can sometimes generate 1000's of emails in a very short space of time and I wonder if Mercury will accept all of these and what it will do if it is not happy ?

I will monitor this evenings mail and hopefully start to see some admin messages begin to turn up.

Thanks

Sean

 

 

 

 

 

<p>Thanks,</p><p>I've added the [] sqare brackets around the ip address in the domains section.</p><p>Yes, I definitely have an admin user in: [configuration | manage local users] and yes, the mailbox path is valid.</p><p>Thomas, The SQL server is addressing mail to 127.0.0.1</p><p>Should I also add [127.0.0.1] to the domains / relevant sections of Mercury.ini ? </p><p>Finally, is there a limit on the volume of messages that mercurys will accept from a sender. For example, when the SQL scripts run, it can sometimes generate 1000's of emails in a very short space of time and I wonder if Mercury will accept all of these and what it will do if it is not happy ?</p><p>I will monitor this evenings mail and hopefully start to see some admin messages begin to turn up.</p><p>Thanks</p><p>Sean </p><p>  </p><p> </p><p> </p><p> </p><p> </p>

Should I also add [127.0.0.1] to the domains / relevant sections of Mercury.ini ?
Not really necessary unless it's using the user@[127.0.0.1] literal address form. 

Finally, is there a limit on the volume of messages that mercurys will

accept from a sender. For example, when the SQL scripts run, it can

sometimes generate 1000's of emails in a very short space of time and I

wonder if Mercury will accept all of these and what it will do if it is

not happy ?

Should handle it without breaking into a sweat unless each of these message is huge.  You can test this via File | Generate test message though.  If these are send via MercuryS then the server can run out of threads but as long as you use a large enough timeout on the client side MercuryS can handle it.
<blockquote>Should I also add [127.0.0.1] to the domains / relevant sections of Mercury.ini ? </blockquote>Not really necessary unless it's using the user@[127.0.0.1] literal address form.  <blockquote>Finally, is there a limit on the volume of messages that mercurys will accept from a sender. For example, when the SQL scripts run, it can sometimes generate 1000's of emails in a very short space of time and I wonder if Mercury will accept all of these and what it will do if it is not happy ?</blockquote>Should handle it without breaking into a sweat unless each of these message is huge.  You can test this via File | Generate test message though.  If these are send via MercuryS then the server can run out of threads but as long as you use a large enough timeout on the client side MercuryS can handle it.

The admin emails have indeed started to arrive now.

 Most of them seem to be of the following types:

 - 556 Too many invalid recipient requests. Closing connection.
 - 550 Rule imposed mailbox access for user@host.com.au refused: user invalid
 - 550 5.1.1 <user@host.com.au>...  User unknown
 - Error connecting to primary server '210.9.177.134'.

With regard to the 556 errors, should I assume that these messages will not have been received, or is Mercury smart enough to fall back and try again later ?

Thanks again.

Sean

 

&lt;p&gt;The admin emails have indeed started to arrive now.&lt;/p&gt;&lt;p&gt;&amp;nbsp;Most of them seem to be of the following types:&lt;/p&gt;&amp;nbsp;- 556 Too many invalid recipient requests. Closing connection. &amp;nbsp;- 550 Rule imposed mailbox access for user@host.com.au refused: user invalid &amp;nbsp;- 550 5.1.1 &amp;lt;user@host.com.au&amp;gt;...&amp;nbsp; User unknown &amp;nbsp;- Error connecting to primary server &#039;210.9.177.134&#039;.&lt;p&gt;With regard to the 556 errors, should I assume that these messages will not have been received, or is Mercury smart enough to fall back and try again later ? &lt;/p&gt;&lt;p&gt;Thanks again.&lt;/p&gt;&lt;p&gt;Sean &lt;/p&gt;&lt;p&gt;&amp;nbsp;&lt;/p&gt;

AFAIK (reading RFC's makes my brain hurt) 500 series errors are fatal and no retry is made.

A 400 series error will be retried as per the settings in Mercury Core config.

&lt;p&gt;AFAIK (reading RFC&#039;s makes my brain hurt) 500 series errors are fatal and no retry is made.&lt;/p&gt;&lt;p&gt;A 400 series error will be retried as per the settings in Mercury Core config. &lt;/p&gt;

With regard to the 556 errors, should I assume that these messages will

not have been received, or is Mercury smart enough to fall back and try

again later ?

556 is fatal and will not be retried.  This was a single message with a lot of RCPT TO: addresses and the receiving system has set a limit on how many bad addresses it will accept in a single connection.  To solve a problem like this you should use a mailing list to create either a single message for each user ( VERP)  or reduce the number of addresses in each message (Explode).  VERP based is the most effective but it also takes to most resources.  I use it for 1000 member lists though quite effectively with Mercury/32 and MercuryE.

 From the Mercury/32 mailing list help:

2: VERP-based error handling   VERP stands for "Variable Envelope Return-path Processing": when using this method, every recipient in the list gets a separate copy of every message sent to the list, and in that copy of the message, a special version of the Return-path field is created that allows Mercury to work out the individual list and subscriber from any errors that get returned to it. Using this information, Mercury can automatically take certain actions when errors occur, such as setting the subscriber's entry to "NOMAIL" or deleting the subscription. Using VERP allows error handling to be almost entirely automated for a mailing list, but it is very "expensive" in the sense that it generates an individual message for every subscriber when a message is sent to the list. Even given the expense factor, however, VERP is often the only manageable way of handling larger mailing lists. Note: when using VERP error handling, any value you enter in the "Explode submissions into x jobs" field of the Distribution page of the Mailing List editor will be ignored - VERP-based mailing always generates a separate copy of every message for every subscriber on the list.

 Explode submissions  For large lists, it can be significantly more efficient to send the message out to several chunks of the subscription list instead of simply generating one large message, since doing so allows multiple SMTP processes to handle the mail at the same time. If you enter a value here, Mercury will "explode" messages sent to the list into that number of outgoing jobs. This setting can have a dramatic impact on list delivery if you are using the MercuryE SMTP end-to-end delivery protocol module. You cannot explode a submission into more than 20 jobs.

&lt;blockquote&gt;With regard to the 556 errors, should I assume that these messages will not have been received, or is Mercury smart enough to fall back and try again later ?&lt;/blockquote&gt;&lt;p&gt;556 is fatal and will not be retried.&amp;nbsp; This was a single message with a lot of RCPT TO: addresses and the receiving system has set a limit on how many bad addresses it will accept in a single connection.&amp;nbsp; To solve a problem like this you should use a mailing list to create either a single message for each user ( VERP)&amp;nbsp; or reduce the number of addresses in each message (Explode).&amp;nbsp; VERP based is the most effective but it also takes to most resources.&amp;nbsp; I use it for 1000 member lists though quite effectively with Mercury/32 and MercuryE. &lt;/p&gt;&lt;p&gt;&amp;nbsp;From the Mercury/32 mailing list help: &lt;/p&gt;&lt;p&gt;&lt;i&gt;&lt;b&gt;2: VERP-based error handling&lt;/b&gt;&lt;/i&gt;&amp;nbsp;&amp;nbsp; VERP stands for &quot;Variable Envelope Return-path Processing&quot;: when using this method, every recipient in the list gets a separate copy of every message sent to the list, and in that copy of the message, a special version of the Return-path field is created that allows Mercury to work out the individual list and subscriber from any errors that get returned to it. Using this information, Mercury can automatically take certain actions when errors occur, such as setting the subscriber&#039;s entry to &quot;NOMAIL&quot; or deleting the subscription. Using VERP allows error handling to be almost entirely automated for a mailing list, but it is very &quot;expensive&quot; in the sense that it generates an individual message for every subscriber when a message is sent to the list. Even given the expense factor, however, VERP is often the only manageable way of handling larger mailing lists. Note: when using VERP error handling, any value you enter in the &quot;Explode submissions into x jobs&quot; field of the Distribution page of the Mailing List editor will be ignored - VERP-based mailing always generates a separate copy of every message for every subscriber on the list.&lt;/p&gt;&lt;p&gt;&lt;i&gt;&lt;b&gt;&amp;nbsp;Explode submissions&lt;/b&gt;&lt;/i&gt;&amp;nbsp; For large lists, it can be significantly more efficient to send the message out to several chunks of the subscription list instead of simply generating one large message, since doing so allows multiple SMTP processes to handle the mail at the same time. If you enter a value here, Mercury will &quot;explode&quot; messages sent to the list into that number of outgoing jobs. This setting can have a dramatic impact on list delivery if you are using the MercuryE SMTP end-to-end delivery protocol module. You cannot explode a submission into more than 20 jobs.&lt;/p&gt;

That is very strange, as the SQL Server job generates a singe email to each user, for each matching alert in the system.

It does not create messages with multiple recipients ?

 

&lt;p&gt;That is very strange, as the SQL Server job generates a singe email to each user, for each matching alert in the system.&lt;/p&gt;&lt;p&gt;It does not create messages with multiple recipients ?&lt;/p&gt;&lt;p&gt;&amp;nbsp; &lt;/p&gt;

That is very strange, as the SQL Server job generates a singe email to each user, for each matching alert in the system.

It does not create messages with multiple recipients ?

Then I have no idea why you got that particular error message unless that server is recording each message you send to that server to determine if you are sending messages with a lot of bad addresses.  Mercury/32 would never combine messages.  What server sent you that error message? 

&lt;blockquote&gt;&lt;p&gt;That is very strange, as the SQL Server job generates a singe email to each user, for each matching alert in the system.&lt;/p&gt;&lt;/blockquote&gt;&lt;blockquote&gt;&lt;p&gt;It does not create messages with multiple recipients ?&lt;/p&gt;&lt;/blockquote&gt;&lt;p&gt;Then I have no idea why you got that particular error message unless that server is recording each message you send to that server to determine if you are sending messages with a lot of bad addresses.&amp;nbsp; Mercury/32 would never combine messages.&amp;nbsp; What server sent you that error message?&amp;nbsp; &lt;/p&gt;

It seems to be an error generated directly by mercury.

The message headers indicate:

From: postmaster@[203.26.##.##]

To: admin@[203.26.##.##] 

 

The attached message has failed delivery and has been referred
to you as postmaster. The following error report or reports
were given to explain the problem:

*** user@address.com.au
Error connecting to primary server '210.9.177.134'.
556 Too many invalid recipient requests. Closing connection.
&lt;p&gt;It seems to be an error generated directly by mercury.&lt;/p&gt;&lt;p&gt;The message headers indicate:&lt;/p&gt;&lt;p&gt;From: postmaster@[203.26.##.##]&lt;/p&gt;&lt;p&gt;To: admin@[203.26.##.##]&amp;nbsp; &lt;/p&gt;&lt;p&gt;&amp;nbsp; &lt;/p&gt;&lt;pre wrap=&quot;&quot;&gt;The attached message has failed delivery and has been referred to you as postmaster. The following error report or reports were given to explain the problem: *** &lt;a href=&quot;mailto:kday@universalmagazines.com.au&quot; class=&quot;moz-txt-link-abbreviated&quot;&gt;user@address.com.au&lt;/a&gt; Error connecting to primary server &#039;210.9.177.134&#039;. 556 Too many invalid recipient requests. Closing connection.&lt;/pre&gt;
live preview
enter atleast 10 characters
WARNING: You mentioned %MENTIONS%, but they cannot see this message and will not be notified
Saving...
Saved
With selected deselect posts show selected posts
All posts under this topic will be deleted ?
Pending draft ... Click to resume editing
Discard draft