I've seen postings about this before, but I don't recall the solutions (nor can I find them when searching through the list archives). So my question is, plain and simply, how do I stop this from happening?
I don't want this kind of unnecessary activity on the server, nor do I want my logs to be polluted like this anymore.
Mar 26 13:08:27 penguin courierd: completed,id=00107A31.4063826D.00004A7A
Mar 26 13:08:27 penguin courierd: id=00107A31.4063826D.00004A7A,from=<[EMAIL PROTECTED]>,addr=<[EMAIL PROTECTED]>: 450 Service temporarily unavailable.
There are two problems here.
One is that you have the machine name set to "penguin". You must always use a fully-qualified domain name. Adjust your "me" and/or "locals" and/or "default" configuration files accordingly.
Secondly you do need a local postmaster address in every case.
If this is a relay-only box, then just create a postmaster address that bitbuckets all mail. There are many ways to do that. One of them is to create an empty file $sysconfdir/aliasdir/.courier-postmaster, owned by the courier userid and groupid.
A more intelligent way to do this, is to assign an unroutable, fake, but an otherwise valid hostname to the machine. Pick one of your domains, such as example.com. Your DNS zone file most certainly defines localhost.example.com, as 127.0.0.1. Set the machine name, using the "me" configuration file (and locals/default/whatever). From Courier's perspective the domain name is valid, with a local postmaster mailbox, which will bitbucket all doublebounces. Yet you may provide postmaster mailboxes in all of your real domains just like any other address.
pgp00000.pgp
Description: PGP signature
