YASQ (Yet Another Sendmail Question)

Numberwhun numberwhun at comcast.net
Fri May 20 22:51:01 EDT 2005


Steven W. Orr wrote:

> I recently removed the following line from my sendmail.mc
>
> define(`LUSER_RELAY', `local:steveo')
>
> That line was causing all mail that was coming in to a non-existant 
> address to go to steveo.
>
> Now I'd like to do exactly the opposite:
>
> I want to reject all incoming mail that is coming from a nonexistant 
> account. For example, I have no user called fido. I want to reject 
> (from in sendmail) any mail that says
>
> From: fido at syslang.net
> To: steveo at syslang,net
>
> Any way to do that?
>
Steven,

Ok, I am by no means a sendmail expert.  In fact, I haven't played with 
it in a long time.  But, I did some googling and found the following 
excerpt:


      Sendmail Validation

The check_compat ruleset compares all senders and receiver pairs before 
mail is delivered. It validates the mail based on the results of the 
comparison. It checks to see if host A can legally send a message to 
host B. check_compat is called for all mail deliveries, not just SMTP 
transactions.

It is used in the following situations:

    *

      A set of users who are restricted from sending mail messages to
      external domains need to send mail messages to internal; domains.
      Both the sender and recipient addresses are checked to ensure that
      they are in the local domain.

    *

      A particular user needs to ensure that he or she does not receive
      mail messages from a specific source.

    *

      A particular host needs to ensure that external senders do not use
      that host as a a mail relay. The mail messages are screened based
      on the sender's hostname.


The link to the page is:  http://docs.hp.com/en/B2355-90685/ch04s09.html

The first bullet point, which checks the validity of user accounts 
against the local domain sounds like what you are looking for.

Regards,

Jeff

-------------- next part --------------
Skipped content of type multipart/related


More information about the gnhlug-discuss mailing list